From 509a05200f190b1317c4387fd6fdb260dd9f8433 Mon Sep 17 00:00:00 2001 From: Anshuman Chhabra Date: Wed, 25 Jul 2018 16:22:18 +0530 Subject: [PATCH 1/5] Added documentation --- lib/graph.ex | 3 +- lib/matrix.ex | 3 +- lib/nifs.ex | 2 + lib/tensor.ex | 3 +- lib/tensorflex.ex | 883 ++++++++++++++++++++++++++++++++++++++++++++++ mix.exs | 3 +- 6 files changed, 893 insertions(+), 4 deletions(-) diff --git a/lib/graph.ex b/lib/graph.ex index 6efa374..4620c11 100644 --- a/lib/graph.ex +++ b/lib/graph.ex @@ -1,5 +1,6 @@ defmodule Tensorflex.Graph do - + @moduledoc false + defstruct [:def, :name] end diff --git a/lib/matrix.ex b/lib/matrix.ex index f1c97b5..c6c3d7b 100644 --- a/lib/matrix.ex +++ b/lib/matrix.ex @@ -1,5 +1,6 @@ defmodule Tensorflex.Matrix do - + @moduledoc false + defstruct [:nrows, :ncols, :data] end diff --git a/lib/nifs.ex b/lib/nifs.ex index 2aab31f..ce72f73 100644 --- a/lib/nifs.ex +++ b/lib/nifs.ex @@ -1,4 +1,6 @@ defmodule Tensorflex.NIFs do + @moduledoc false + @on_load :load_nifs def load_nifs do diff --git a/lib/tensor.ex b/lib/tensor.ex index 8046776..d314e4f 100644 --- a/lib/tensor.ex +++ b/lib/tensor.ex @@ -1,5 +1,6 @@ defmodule Tensorflex.Tensor do - + @moduledoc false + defstruct [:datatype, :tensor] end diff --git a/lib/tensorflex.ex b/lib/tensorflex.ex index 03ac0c9..b5a0302 100644 --- a/lib/tensorflex.ex +++ b/lib/tensorflex.ex @@ -1,11 +1,81 @@ defmodule Tensorflex do + @moduledoc """ + A simple and fast library for running Tensorflow graph models in Elixir. Tensorflex is written around the [Tensorflow C API](https://www.tensorflow.org/install/install_c), and allows Elixir developers to leverage Machine Learning and Deep Learning solutions in their projects. + __NOTE__: + + - Make sure that the C API version and Python API version (assuming you are using the Python API for first training your models) are the latest. As of July 2018, the latest version is `r1.9`. + + - Since Tensorflex provides Inference capability for pre-trained graph models, it is assumed you have adequate knowledge of the pre-trained models you are using (such as the input data type/dimensions, input and output operation names, etc.). Some basic understanding of the [Tensorflow Python API](https://www.tensorflow.org/api_docs/python/) can come in very handy. + + - Tensorflex consists of multiple NIFs, so exercise caution while using it-- providing incorrect operation names for running sessions, incorrect dimensions of tensors than the actual pre-trained graph requires, providing different tensor datatypes than the ones required by the graph can all lead to failure. While these are not easy errors to make, do ensure that you test your solution well before deployment. + """ + alias Tensorflex.{NIFs, Graph, Tensor, Matrix} defp empty_list?([[]]), do: true defp empty_list?(list) when is_list(list) do false end + + @doc """ + Used for loading a Tensorflow `.pb` graph model in Tensorflex. + + Reads in a pre-trained Tensorflow protobuf (`.pb`) Graph model binary file. + + Returns a tuple `{:ok, %Graph}`. + + `%Graph` is an internal Tensorflex struct which holds the name of the graph file and the binary definition data that is read in via the `.pb` file. + + ## Examples: + + _Reading in a graph_ + + As an example, we can try reading in the [Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) convolutional neural network based image classification graph model by Google. The graph file is named `classify_image_graph_def.pb`: + ```elixir + iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb" + 2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization(). + {:ok, + %Tensorflex.Graph{ + def: #Reference<0.3018278404.759824385.5268>, + name: "classify_image_graph_def.pb" + }} + ``` + Generally to check that the loaded graph model is correct and contains computational operations, the `get_graph_ops/1` function is useful: + ```elixir + iex(2)> Tensorflex.get_graph_ops graph + ["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims", + "ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul", + "conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta", + "conv/batchnorm/gamma", "conv/batchnorm/moving_mean", + "conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics", + "conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D", + "conv_1/batchnorm/beta", "conv_1/batchnorm/gamma", + "conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance", + "conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency", + "conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta", + "conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean", + "conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics", + "conv_2/control_dependency", "conv_2", "pool/CheckNumerics", + "pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D", + "conv_3/batchnorm/beta", "conv_3/batchnorm/gamma", + "conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...] + + ``` + + _Incorrect usage will `raise`_: + + ```elixir + iex(3)> {:ok, graph} = Tensorflex.read_graph "Makefile" + ** (ArgumentError) file is not a protobuf .pb file + (tensorflex) lib/tensorflex.ex:27: Tensorflex.read_graph/1 + + iex(3)> {:ok, graph} = Tensorflex.read_graph "Makefile.pb" + ** (ArgumentError) graph definition file does not exist + (tensorflex) lib/tensorflex.ex:23: Tensorflex.read_graph/1 + + ``` + """ def read_graph(filepath) do unless File.exists?(filepath) do @@ -20,10 +90,190 @@ defmodule Tensorflex do {:ok, %Graph{def: ref, name: filepath}} end + @doc """ + Used for listing all the operations in a Tensorflow `.pb` graph. + + Reads in a Tensorflex ```%Graph``` struct obtained from `read_graph/1`. + + Returns a list of all the operation names (as strings) that populate the graph model. + + ## Examples + + - _Google Inception CNN Model_ ([source](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz)) + + ```elixir + iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb" + 2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization(). + {:ok, + %Tensorflex.Graph{ + def: #Reference<0.3018278404.759824385.5268>, + name: "classify_image_graph_def.pb" + }} + + iex(2)> Tensorflex.get_graph_ops graph + ["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims", + "ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul", + "conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta", + "conv/batchnorm/gamma", "conv/batchnorm/moving_mean", + "conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics", + "conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D", + "conv_1/batchnorm/beta", "conv_1/batchnorm/gamma", + "conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance", + "conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency", + "conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta", + "conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean", + "conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics", + "conv_2/control_dependency", "conv_2", "pool/CheckNumerics", + "pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D", + "conv_3/batchnorm/beta", "conv_3/batchnorm/gamma", + "conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...] + ``` + + - _Iris Dataset MLP Model_ ([source](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/)) + + ```elixir + iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_iris.pb" + {:ok, + %Tensorflex.Graph{ + def: #Reference<0.4109712726.1847984130.24506>, + name: "graphdef_iris.pb" + }} + + iex(2)> Tensorflex.get_graph_ops graph + ["input", "weights1", "weights1/read", "biases1", "biases1/read", "weights2", "weights2/read", "biases2", "biases2/read", "MatMul", "Add", "Relu", "MatMul_1", "Add_1", "output"] + + ``` + + - _Toy Computational Graph Model_ ([source](https://github.com/anshuman23/tensorflex/tree/master/examples/toy-example)) + + ```elixir + iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_toy.pb" + {:ok, + %Tensorflex.Graph{ + def: #Reference<0.1274892327.1580335105.235135>, + name: "graphdef_toy.pb" + }} + + iex(2)> Tensorflex.get_graph_ops graph + ["input", "weights", "weights/read", "biases", "biases/read", "MatMul", "add", "output"] + ``` + + - _RNN LSTM Sentiment Analysis Model_ ([source](https://github.com/anshuman23/tensorflex/pull/25)) + + ```elixir + iex(1)> {:ok, graph} = Tensorflex.read_graph "frozen_model_lstm.pb" + {:ok, + %Tensorflex.Graph{ + def: #Reference<0.713975820.1050542081.11558>, + name: "frozen_model_lstm.pb" + }} + + iex(2)> Tensorflex.get_graph_ops graph + ["Placeholder_1", "embedding_lookup/params_0", "embedding_lookup", + "transpose/perm", "transpose", "rnn/Shape", "rnn/strided_slice/stack", + "rnn/strided_slice/stack_1", "rnn/strided_slice/stack_2", "rnn/strided_slice", + "rnn/stack/1", "rnn/stack", "rnn/zeros/Const", "rnn/zeros", "rnn/stack_1/1", + "rnn/stack_1", "rnn/zeros_1/Const", "rnn/zeros_1", "rnn/Shape_1", + "rnn/strided_slice_2/stack", "rnn/strided_slice_2/stack_1", + "rnn/strided_slice_2/stack_2", "rnn/strided_slice_2", "rnn/time", + "rnn/TensorArray", "rnn/TensorArray_1", "rnn/TensorArrayUnstack/Shape", + "rnn/TensorArrayUnstack/strided_slice/stack", + "rnn/TensorArrayUnstack/strided_slice/stack_1", + "rnn/TensorArrayUnstack/strided_slice/stack_2", + "rnn/TensorArrayUnstack/strided_slice", "rnn/TensorArrayUnstack/range/start", + "rnn/TensorArrayUnstack/range/delta", "rnn/TensorArrayUnstack/range", + "rnn/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3", + "rnn/while/Enter", "rnn/while/Enter_1", "rnn/while/Enter_2", + "rnn/while/Enter_3", "rnn/while/Merge", "rnn/while/Merge_1", + "rnn/while/Merge_2", "rnn/while/Merge_3", "rnn/while/Less/Enter", + "rnn/while/Less", "rnn/while/LoopCond", "rnn/while/Switch", + "rnn/while/Switch_1", "rnn/while/Switch_2", "rnn/while/Switch_3", ...] + ``` + """ + def get_graph_ops(%Graph{def: ref, name: filepath}) do NIFs.get_graph_ops(ref) end + @doc """ + Creates a 2-D Tensorflex matrix from custom input specifications. + + Takes three input arguments: number of rows in matrix (`nrows`), number of columns in matrix (`ncols`), and a list of lists of the data that will form the matrix (`datalist`). + + Returns a `%Matrix` Tensorflex struct type. + + ## Examples: + + _Creating a new matrix_ + + ```elixir + iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]]) %Tensorflex.Matrix{ + data: #Reference<0.759278808.823525378.128525>, + ncols: 3, + nrows: 2 + } + ``` + + All `%Matrix` Tensorflex matrices can be passed in to the other matrix inspection and manipulation functions-- `matrix_pos/3`,`size_of_matrix/1`, `matrix_to_lists/1`, and `append_to_matrix/2`: + + ```elixir + iex(1)> mat = Tensorflex.create_matrix(4,4,[[123,431,23,1],[1,2,3,4],[5,6,7,8],[768,564,44,5]]) + %Tensorflex.Matrix{ + data: #Reference<0.878138179.2435973124.131489>, + ncols: 4, + nrows: 4 + } + + iex(2)> mat = Tensorflex.append_to_matrix(mat, [[1,1,1,1]]) + %Tensorflex.Matrix{ + data: #Reference<0.878138179.2435973124.131489>, + ncols: 4, + nrows: 5 + } + + iex(3)> Tensorflex.matrix_to_lists mat + [ + [123.0, 431.0, 23.0, 1.0], + [1.0, 2.0, 3.0, 4.0], + [5.0, 6.0, 7.0, 8.0], + [768.0, 564.0, 44.0, 5.0], + [1.0, 1.0, 1.0, 1.0] + ] + + iex(4)> Tensorflex.matrix_pos(mat,5,3) + 1.0 + + iex(5)> Tensorflex.size_of_matrix mat + {5, 4} + ``` + + _Incorrect usage will `raise`_: + + ```elixir + iex(1)> Tensorflex.create_matrix(1,2,[[1,2,3]]) + ** (ArgumentError) argument error + (tensorflex) Tensorflex.NIFs.create_matrix(1, 2, [[1, 2, 3]]) + (tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3 + + iex(1)> Tensorflex.create_matrix(2,1,[[1,2,3]]) + ** (ArgumentError) argument error + (tensorflex) Tensorflex.NIFs.create_matrix(2, 1, [[1, 2, 3]]) + (tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3 + + iex(1)> Tensorflex.create_matrix(2,3,[[1.1,23,3.4], []]) + ** (ArgumentError) argument error + (tensorflex) Tensorflex.NIFs.create_matrix(2, 3, [[1.1, 23, 3.4], []]) + (tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3 + + iex(1)> Tensorflex.create_matrix(1,2,[[]]) + ** (ArgumentError) data provided cannot be an empty list + (tensorflex) lib/tensorflex.ex:243: Tensorflex.create_matrix/3 + + iex(1)> Tensorflex.create_matrix(-1,2,[[3,4]]) + ** (FunctionClauseError) no function clause matching in Tensorflex.create_matrix/3 + ``` + """ + def create_matrix(nrows, ncols, datalist) when nrows > 0 and ncols > 0 do if(empty_list? datalist) do raise ArgumentError, "data provided cannot be an empty list" @@ -33,14 +283,112 @@ defmodule Tensorflex do %Matrix{nrows: nrows, ncols: ncols, data: ref} end + @doc """ + + Used for accessing an element of a Tensorflex matrix. + + Takes in three input arguments: a Tensorflex `%Matrix` struct matrix, and the row (`row`) and column (`col`) values of the required element in the matrix. Both `row` and `col` here are __NOT__ zero indexed. + + Returns the value as float. + + ## Examples + + ```elixir + iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]]) + %Tensorflex.Matrix{ + data: #Reference<0.759278808.823525378.128525>, + ncols: 3, + nrows: 2 + } + + iex(2)> Tensorflex.matrix_pos(mat,2,1) + 5.5 + + iex(3)> Tensorflex.matrix_pos(mat,1,3) + 44.5 + + ``` + """ + def matrix_pos(%Matrix{nrows: nrows, ncols: ncols, data: ref}, row, col) when row > 0 and col > 0 do NIFs.matrix_pos(ref, row, col) end + @doc """ + Used for obtaining the size of a Tensorflex matrix. + + Takes a Tensorflex `%Matrix` struct matrix as input. + +Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of the matrix and `ncols` represents the number of columns of the matrix. + + ## Examples + + ```elixir + iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]]) + %Tensorflex.Matrix{ + data: #Reference<0.759278808.823525378.128525>, + ncols: 3, + nrows: 2 + } + + iex(2)> Tensorflex.size_of_matrix mat + {2, 3} + ``` + """ + def size_of_matrix(%Matrix{nrows: nrows, ncols: ncols, data: ref}) do {nrows, ncols} end + @doc """ + Appends a single row to the back of a Tensorflex matrix. + + Takes a Tensorflex `%Matrix` matrix as input and a single row of data (with the same number of columns as the original matrix) as a list of lists (`datalist`) to append to the original matrix. + + Returns the extended and modified `%Matrix` struct matrix. + + ## Examples + + ```elixir + iex(1)> m = Tensorflex.create_matrix(2,3,[[23,23,23],[32,32,32]]) + %Tensorflex.Matrix{ + data: #Reference<0.153563642.2042232833.193025>, + ncols: 3, + nrows: 2 + } + + iex(2)> m = Tensorflex.append_to_matrix(m,[[2,2,2]]) + %Tensorflex.Matrix{ + data: #Reference<0.153563642.2042232833.193025>, + ncols: 3, + nrows: 3 + } + + iex(3)> m = Tensorflex.append_to_matrix(m,[[3,3,3]]) + %Tensorflex.Matrix{ + data: #Reference<0.153563642.2042232833.193025>, + ncols: 3, + nrows: 4 + } + + iex(4)> m |> Tensorflex.matrix_to_lists + [[23.0, 23.0, 23.0], [32.0, 32.0, 32.0], [2.0, 2.0, 2.0], [3.0, 3.0, 3.0]] + + ``` + _Incorrect usage will `raise`_: + + ```elixir + iex(5)> m = Tensorflex.append_to_matrix(m,[[2,2,2],[3,3,3]]) + ** (ArgumentError) data columns must be same as matrix and number of rows must be 1 + (tensorflex) lib/tensorflex.ex:345: Tensorflex.append_to_matrix/2 + + iex(5)> m = Tensorflex.append_to_matrix(m,[[2,2,2,2]]) + ** (ArgumentError) data columns must be same as matrix and number of rows must be 1 + (tensorflex) lib/tensorflex.ex:345: Tensorflex.append_to_matrix/2 + ``` + + """ + def append_to_matrix(%Matrix{nrows: nrows, ncols: ncols, data: ref}, datalist) do unless (datalist |> List.flatten |> Kernel.length) == ncols do raise ArgumentError, "data columns must be same as matrix and number of rows must be 1" @@ -49,64 +397,482 @@ defmodule Tensorflex do %Matrix{nrows: nrows+1, ncols: ncols, data: new_ref} end + @doc """ + Converts a Tensorflex matrix (back) to a list of lists format. + + Takes a Tensorflex `%Matrix` struct matrix as input. + + Returns a list of lists representing the data stored in the matrix. + + __NOTE__: If the matrix contains very high dimensional data, typically obtained from a function like `load_csv_as_matrix/2`, then it is not recommended to convert the matrix back to a list of lists format due to a possibility of memory errors. + + ## Examples + + ```elixir + iex(1)> Tensorflex.create_matrix(2,3,[[23,23,23],[32,32,32]]) |> Tensorflex.matrix_to_lists + [[23.0, 23.0, 23.0], [32.0, 32.0, 32.0]] + ``` + """ + def matrix_to_lists(%Matrix{nrows: nrows, ncols: ncols, data: ref}) do NIFs.matrix_to_lists(ref) end + @doc """ + Creates a `TF_DOUBLE` tensor from Tensorflex matrices containing the values and dimensions specified. + + Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the tensor should have and another `%Matrix` matrix (`matrix2`) containing the dimensions of the required tensor. + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + + ## Examples: + + ```elixir + iex(1)> vals = Tensorflex.create_matrix(2,3,[[12.0,45.2,2.11],[36.7,8.09,9.81]]) + %Tensorflex.Matrix{ + data: #Reference<0.1251941183.3671982081.254268>, + ncols: 3, + nrows: 2 + } + + iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]]) + %Tensorflex.Matrix{ + data: #Reference<0.1251941183.3671982081.254723>, + ncols: 2, + nrows: 1 + } + + iex(3)> {:ok, tensor} = Tensorflex.float64_tensor vals,dims + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_double, + tensor: #Reference<0.1251941183.3671982081.255216> + }} + + ``` + """ + def float64_tensor(%Matrix{nrows: val_rows, ncols: val_cols, data: val_ref}, %Matrix{nrows: dim_rows, ncols: dim_cols, data: dim_ref}) do {:ok, ref} = NIFs.float64_tensor(val_ref, dim_ref) {:ok, %Tensor{datatype: :tf_double, tensor: ref}} end + @doc """ + Creates a `TF_DOUBLE` constant value one-dimensional tensor from the floating point value specified. + + Takes in a float value as input. + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + + ## Examples + + ```elixir + iex(1)> {:ok, tensor} = Tensorflex.float64_tensor 123.123 + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_double, + tensor: #Reference<0.2778616536.4219338753.155412> + }} + + ``` + + _Incorrect usage will `raise`_: + + ```elixir + iex(2)> {:ok, tensor} = Tensorflex.float64_tensor "123.123" + ** (FunctionClauseError) no function clause matching in Tensorflex.float64_tensor/1 + + iex(2)> {:ok, tensor} = Tensorflex.float64_tensor 123 + ** (FunctionClauseError) no function clause matching in Tensorflex.float64_tensor/1 + ``` + """ + def float64_tensor(floatval) when is_float(floatval) do {:ok, ref} = NIFs.float64_tensor(floatval) {:ok, %Tensor{datatype: :tf_double, tensor: ref}} end + @doc """ + Creates a `TF_FLOAT` tensor from Tensorflex matrices containing the values and dimensions specified. + + Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the tensor should have and another `%Matrix` matrix (`matrix2`) containing the dimensions of the required tensor. + +Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + + ## Examples: + + ```elixir + iex(1)> vals = Tensorflex.create_matrix(2,3,[[12.0,45.2,2.11],[36.7,8.09,9.81]]) + %Tensorflex.Matrix{ + data: #Reference<0.1251941183.3671982081.254268>, + ncols: 3, + nrows: 2 + } + + iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]]) + %Tensorflex.Matrix{ + data: #Reference<0.1251941183.3671982081.254723>, + ncols: 2, + nrows: 1 + } + + iex(3)> {:ok, tensor} = Tensorflex.float32_tensor vals,dims + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_float, + tensor: #Reference<0.1251941183.3671982081.255228> + }} + + ``` + """ + def float32_tensor(%Matrix{nrows: val_rows, ncols: val_cols, data: val_ref}, %Matrix{nrows: dim_rows, ncols: dim_cols, data: dim_ref}) do {:ok, ref} = NIFs.float32_tensor(val_ref, dim_ref) {:ok, %Tensor{datatype: :tf_float, tensor: ref}} end + @doc """ + Creates a `TF_FLOAT` constant value one-dimensional tensor from the floating point value specified. + + Takes in a float value as input. + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + + ## Examples + + ```elixir + iex(1)> {:ok, tensor} = Tensorflex.float32_tensor 123.123 + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_float, + tensor: #Reference<0.2011963375.1804468228.236110> + }} + + ``` + + _Incorrect usage will `raise`_: + + ```elixir + iex(2)> {:ok, tensor} = Tensorflex.float32_tensor "123.123" + ** (FunctionClauseError) no function clause matching in Tensorflex.float32_tensor/1 + + iex(2)> {:ok, tensor} = Tensorflex.float32_tensor 123 + ** (FunctionClauseError) no function clause matching in Tensorflex.float32_tensor/1 + ``` + """ + def float32_tensor(floatval) when is_float(floatval) do {:ok, ref} = NIFs.float32_tensor(floatval) {:ok, %Tensor{datatype: :tf_float, tensor: ref}} end + @doc """ + Creates a `TF_INT32` tensor from Tensorflex matrices containing the values and dimensions specified. + + Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the tensor should have and another `%Matrix` matrix (`matrix2`) containing the dimensions of the required tensor. + +Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + + __NOTE__: In case floating point values are passed in the values matrix (`matrix1`) as arguments for this function, the tensor will still be created and all the float values will be typecast to integers. + + ## Examples: + + ```elixir + iex(1)> vals = Tensorflex.create_matrix(2,3,[[123,45,333],[2,2,899]]) + %Tensorflex.Matrix{ + data: #Reference<0.1256144000.2868510721.170449>, + ncols: 3, + nrows: 2 + } + + iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]]) + %Tensorflex.Matrix{ + data: #Reference<0.1256144000.2868510721.170894>, + ncols: 2, + nrows: 1 + } + + iex(3)> {:ok, tensor} = Tensorflex.int32_tensor vals,dims + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_int32, + tensor: #Reference<0.1256144000.2868510721.171357> + }} + + ``` + """ + def int32_tensor(%Matrix{nrows: val_rows, ncols: val_cols, data: val_ref}, %Matrix{nrows: dim_rows, ncols: dim_cols, data: dim_ref}) do {:ok, ref} = NIFs.int32_tensor(val_ref, dim_ref) {:ok, %Tensor{datatype: :tf_int32, tensor: ref}} end + @doc """ + Creates a `TF_INT32` constant value one-dimensional tensor from the integer value specified. + + Takes in an integer value as input. + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + + ## Examples + + ```elixir + iex(1)> {:ok, tensor} = Tensorflex.int32_tensor 123 + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_int32, + tensor: #Reference<0.1927663658.3415343105.162588> + }} + ``` + + _Incorrect usage will `raise`_: + + ```elixir + iex(2)> {:ok, tensor} = Tensorflex.int32_tensor 123.123 + ** (FunctionClauseError) no function clause matching in Tensorflex.int32_tensor/1 + + iex(2)> {:ok, tensor} = Tensorflex.int32_tensor "123.123" + ** (FunctionClauseError) no function clause matching in Tensorflex.int32_tensor/1 + + ``` + """ + def int32_tensor(intval) when is_integer(intval) do {:ok, ref} = NIFs.int32_tensor(intval) {:ok, %Tensor{datatype: :tf_int32, tensor: ref}} end + @doc """ + Creates a `TF_STRING` constant value string tensor from the string value specified. + + Takes in a string value as input. + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + + ## Examples + + ```elixir + iex(1)> {:ok, tensor} = Tensorflex.string_tensor "123.123" + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_string, + tensor: #Reference<0.2069282048.194904065.41126> + }} + + ``` + + _Incorrect usage will `raise`_: + + ```elixir + iex(2)> {:ok, tensor} = Tensorflex.string_tensor 123.123 + ** (FunctionClauseError) no function clause matching in Tensorflex.string_tensor/1 + + iex(2)> {:ok, tensor} = Tensorflex.string_tensor 123 + ** (FunctionClauseError) no function clause matching in Tensorflex.string_tensor/1 + ``` + """ + def string_tensor(stringval) when is_binary(stringval) do {:ok, ref} = NIFs.string_tensor(stringval) {:ok, %Tensor{datatype: :tf_string, tensor: ref}} end + @doc """ + Allocates a `TF_INT32` tensor of specified dimensions. + + This function is generally used to allocate output tensors that do not hold any value data yet, but _will_ after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the `run_session/5` function to hold the output values generated as predictions. + + Takes a Tensorflex `%Matrix` struct matrix as input. + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding the potential tensor data and type. + + ## Examples + + As an example, we can allocate an `int32` output tensor that will be a vector of 250 values (`1x250` matrix). Therefore, after the session is run, the output will be an `integer` vector containing 250 values: + + ```elixir + iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.int32_tensor_alloc + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_int32, + tensor: #Reference<0.961157994.2087059457.18950> + }} + + ``` + """ + def int32_tensor_alloc(%Matrix{nrows: dim_rows, ncols: dim_cols, data: dim_ref}) do {:ok, ref} = NIFs.int32_tensor_alloc(dim_ref) {:ok, %Tensor{datatype: :tf_int32, tensor: ref}} end + + @doc """ + Allocates a `TF_FLOAT` tensor of specified dimensions. + + This function is generally used to allocate output tensors that do not hold any value data yet, but _will_ after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the `run_session/5` function to hold the output values generated as predictions. + + Takes a Tensorflex `%Matrix` struct matrix as input. + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding the potential tensor data and type. + + ## Examples + + As an example, we can allocate a `float32` output tensor that will be a vector of 250 values (`1x250` matrix). Therefore, after the session is run, the output will be a `float` vector containing 250 values: + + ```elixir + iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float32_tensor_alloc + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_float, + tensor: #Reference<0.961157994.2087059457.19014> + }} + + ``` + """ def float32_tensor_alloc(%Matrix{nrows: dim_rows, ncols: dim_cols, data: dim_ref}) do {:ok, ref} = NIFs.float32_tensor_alloc(dim_ref) {:ok, %Tensor{datatype: :tf_float, tensor: ref}} end + @doc """ + Allocates a `TF_DOUBLE` tensor of specified dimensions. + + This function is generally used to allocate output tensors that do not hold any value data yet, but _will_ after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the `run_session/5` function to hold the output values generated as predictions. + + Takes a Tensorflex `%Matrix` struct matrix as input. + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding the potential tensor data and type. + + ## Examples + + As an example, we can allocate a `float64` output tensor that will be a vector of 250 values (`1x250` matrix). Therefore, after the session is run, the output will be a `double` vector containing 250 values: + + ```elixir + iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float64_tensor_alloc + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_double, + tensor: #Reference<0.961157994.2087059457.19025> + }} + + ``` + """ + def float64_tensor_alloc(%Matrix{nrows: dim_rows, ncols: dim_cols, data: dim_ref}) do {:ok, ref} = NIFs.float64_tensor_alloc(dim_ref) {:ok, %Tensor{datatype: :tf_double, tensor: ref}} end + @doc """ + Used to get the datatype of a created tensor. + + Takes in a `%Tensor` struct tensor as input. + + Returns a tuple `{:ok, datatype}` where `datatype` is an atom representing the list of Tensorflow `TF_DataType` tensor datatypes. Click [here](https://github.com/anshuman23/tensorflex/blob/master/c_src/c_api.h#L98-L122) to view a list of all possible datatypes. + + ## Examples + + ```elixir + iex(1)> {:ok, tensor} = Tensorflex.string_tensor "example" + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_string, + tensor: #Reference<0.4132928949.2894987267.194583> + }} + + iex(2)> Tensorflex.tensor_datatype tensor + {:ok, :tf_string} + ``` + """ + def tensor_datatype(%Tensor{datatype: datatype, tensor: ref}) do {:ok, datatype} end + @doc """ + Loads `JPEG` images into Tensorflex directly as a `TF_UINT8` tensor of dimensions `image height x image width x number of color channels`. + + This function is very useful if you wish to do image classification using Convolutional Neural Networks, or other Deep Learning Models. One of the most widely adopted and robust image classification models is the [Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) model by Google. It makes classifications on images from over a 1000 classes with highly accurate results. The `load_image_as_tensor/1` function is an essential component for the prediction pipeline of the Inception model (and for other similar image classification models) to work in Tensorflex. + + Reads in the path to a `JPEG` image file (`.jpg` or `.jpeg`). + + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding the tensor data and type. Here the created Tensor is a `uint8` tensor (`TF_UINT8`). + + __NOTE__: For now, only 3 channel RGB `JPEG` color images can be passed as arguments. Support for grayscale images and other image formats such as `PNG` will be added in the future. + + ## Examples + + To exemplify the working of the `load_image_as_tensor/1` function we will cover the entire prediction pipeline for the Inception model. However, this makes use of many other Tensorflex functions such as `run_session/5` and the other tensor functions so it would be advisable to go through them first. Also, the Inception model can be downloaded [here](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz). We will be making use of the `cropped_panda.jpg` image file that comes along with the model to test out the model in Tensorflex. + + First the graph is loaded: + + ```elixir + iex(1)> {:ok, graph} = Tensorflex.read_graph("classify_image_graph_def.pb") + 2018-07-25 14:20:29.079139: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization(). + {:ok, + %Tensorflex.Graph{ + def: #Reference<0.542869014.389152771.105680>, + name: "classify_image_graph_def.pb" + }} + ``` + Then we load the image as a `uint8` tensor: + + ```elixir + iex(2)> {:ok, input_tensor} = Tensorflex.load_image_as_tensor("cropped_panda.jpg") + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_uint8, + tensor: #Reference<0.1203951739.122552322.52747> + }} + ``` + Then we create the output tensor which will hold out output vector values. For the Inception model, the output is received as a `1008x1 float32` tensor, as there are 1008 classes in the model: + + ```elixir + iex(3)> {:ok, output_tensor} = Tensorflex.create_matrix(1,2,[[1008,1]]) |> Tensorflex.float32_tensor_alloc + {:ok, + %Tensorflex.Tensor{ + datatype: :tf_float, + tensor: #Reference<0.1203951739.122552322.52794> + }} + ``` + Next, we obtain the results by running the session: + + ```elixir + iex(4)> results = Tensorflex.run_session(graph, input_tensor, output_tensor, "DecodeJpeg", "softmax") + 2018-07-25 14:33:40.992813: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA + [ + [1.059142014128156e-4, 2.8240500250831246e-4, 8.30648496048525e-5, + 1.2982363114133477e-4, 7.32232874725014e-5, 8.014426566660404e-5, + 6.63459359202534e-5, 0.003170756157487631, 7.931600703159347e-5, + 3.707312498590909e-5, 3.0997329304227605e-5, 1.4232713147066534e-4, + 1.0381334868725389e-4, 1.1057958181481808e-4, 1.4321311027742922e-4, + 1.203602587338537e-4, 1.3130248407833278e-4, 5.850398520124145e-5, + 2.641105093061924e-4, 3.1629020668333396e-5, 3.906813799403608e-5, + 2.8646905775531195e-5, 2.2863158665131778e-4, 1.2222197256051004e-4, + 5.956588938715868e-5, 5.421260357252322e-5, 5.996063555357978e-5, + 4.867801326327026e-4, 1.1005574924638495e-4, 2.3433618480339646e-4, + 1.3062104699201882e-4, 1.317620772169903e-4, 9.388553007738665e-5, + 7.076268957462162e-5, 4.281177825760096e-5, 1.6863139171618968e-4, + 9.093972039408982e-5, 2.611844101920724e-4, 2.7584232157096267e-4, + 5.157176201464608e-5, 2.144951868103817e-4, 1.3628098531626165e-4, + 8.007588621694595e-5, 1.7929042223840952e-4, 2.2831936075817794e-4, + 6.216531619429588e-5, 3.736453436431475e-5, 6.782123091397807e-5, + 1.1538144462974742e-4, ...] + ] + + ``` + Finally, we need to find which class has the maximum probability and identify it's label. Since `results` is a List of Lists, it's better to read in the flattened list. Then we need to find the index of the element in the new list which as the maximum value. Therefore: + ```elixir + iex(5)> max_prob = List.flatten(results) |> Enum.max + 0.8849328756332397 + + iex(6)> Enum.find_index(results |> List.flatten, fn(x) -> x == max_prob end) + 169 + ``` + We can thus see that the class with the maximum probability predicted (`0.8849328756332397`) for the image is `169`. We will now find what the `169` label corresponds to. For this we can look back into the unzipped Inception folder, where there is a file called `imagenet_2012_challenge_label_map_proto.pbtxt`. On opening this file, we can find the string class identifier for the `169` class index. This is `n02510455` and is present on Line 1556 in the file. Finally, we need to match this string identifier to a set of identification labels by referring to the file `imagenet_synset_to_human_label_map.txt` file. Here we can see that corresponding to the string class `n02510455` the human labels are `giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca` (Line 3691 in the file). Thus, we have correctly identified the animal in the image as a panda using Tensorflex. + """ + def load_image_as_tensor(imagepath) do unless File.exists?(imagepath) do raise ArgumentError, "image file does not exist" @@ -120,6 +886,105 @@ defmodule Tensorflex do {:ok, %Tensor{datatype: :tf_uint8, tensor: ref}} end + @doc """ + Loads high-dimensional data from a `CSV` file as a Tensorflex 2-D matrix in a super-fast manner. + + The `load_csv_as_matrix/2` function is very fast-- when compared with the Python based `pandas` library for data science and analysis' function `read_csv` on the `test.csv` file from MNIST Kaggle data ([source](https://www.kaggle.com/c/digit-recognizer/data)), the following execution times were obtained: + - `read_csv`: `2.549233` seconds + - `load_csv_as_matrix/2`: `1.711494` seconds + + This function takes in 2 arguments: a path to a valid CSV file (`filepath`) and other optional arguments `opts`. These include whether or not a header needs to be discarded in the CSV, and what the delimiter type is. These are specified by passing in an atom `:true` or `:false` to the `header:` key, and setting a string value for the `delimiter:` key. By default, the header is considered to be present (`:true`) and the delimiter is set to `,`. + + Returns a `%Matrix` Tensorflex struct type. + + ## Examples: + We first exemplify the working with the `test.csv` file which belongs to the MNIST Kaggle CSV data ([source](https://www.kaggle.com/c/digit-recognizer/data)), which contains `28000` rows and `784` columns (without the header). It is comma delimited and also contains a header. From the `test.csv` file, we also create a custom file withou the header present which we refer to as `test_without_header.csv` in the examples below: + + ```elixir + iex(1)> mat = Tensorflex.load_csv_as_matrix("test.csv") + %Tensorflex.Matrix{ + data: #Reference<0.4024686574.590479361.258459>, + ncols: 784, + nrows: 28000 + } + + iex(2)> Tensorflex.matrix_pos mat, 5,97 + 80.0 + + iex(3)> Tensorflex.matrix_pos mat, 5,96 + 13.0 + ``` + + On a visual inspection of the very large `test.csv` file, one can see that the values in these particular positions are correct. Now we show usage for the same file but without header, `test_without_header.csv`: + ```elixir + iex(1)> no_header = Tensorflex.load_csv_as_matrix("test/test_without_header.csv", header: :false) + %Tensorflex.Matrix{ + data: #Reference<0.4024686574.590479364.257078>, + ncols: 784, + nrows: 28000 + } + + iex(2)> Tensorflex.matrix_pos no_header,5,97 + 80.0 + + iex(3)> Tensorflex.matrix_pos no_header,5,96 + 13.0 + ``` + + Next we see the delimiter functionalities. First, assuming we have two simple `CSV` files, `sample1.csv` and `sample2.csv` + + _sample1.csv_: + + ```elixir + 1,2,3,4,5 + 6,7,8,9,10 + 11,12,13,14,15 + ``` + + _sample2.csv_: + + ```elixir + col1-col2-col3-col4 + 1-2-3-4 + 5-6-7-8 + 9-10-11-12 + ``` + + The examples are as follows: + ```elixir + iex(1)> m1 = Tensorflex.load_csv_as_matrix("sample1.csv", header: :false) + %Tensorflex.Matrix{ + data: #Reference<0.3878093040.3013214209.247502>, + ncols: 5, + nrows: 3 + } + + iex(2)> Tensorflex.matrix_to_lists m1 + [ + [1.0, 2.0, 3.0, 4.0, 5.0], + [6.0, 7.0, 8.0, 9.0, 10.0], + [11.0, 12.0, 13.0, 14.0, 15.0] + ] + + iex(3)> m2 = Tensorflex.load_csv_as_matrix("sample2.csv", header: :true, delimiter: "-") + %Tensorflex.Matrix{ + data: #Reference<0.4024686574.590479361.258952>, + ncols: 4, + nrows: 3 + } + + iex(4)> Tensorflex.matrix_to_lists m2 + [[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0], [9.0, 10.0, 11.0, 12.0]] + ``` + + _Incorrect usage will `raise`_: + ```elixir + iex(1)> not_working = Tensorflex.load_csv_as_matrix("test.csv", header: :no_header, delimiter: ",") + ** (ArgumentError) header indicator atom must be either :true or :false + (tensorflex) lib/tensorflex.ex:122: Tensorflex.load_csv_as_matrix/2 + ``` + """ + def load_csv_as_matrix(filepath, opts \\ []) do unless File.exists?(filepath) do raise ArgumentError, "csv file does not exist" @@ -142,6 +1007,24 @@ defmodule Tensorflex do %Matrix{nrows: nrows, ncols: ncols, data: ref} end + @doc """ + Runs a Tensorflow session to generate predictions for a given graph, input data, and required input/output operations. + + This function is the final step of the Inference (prediction) pipeline and generates output for a given set of input data, a pre-trained graph model, and the specified input and output operations of the graph. + + Takes in five arguments: a pre-trained Tensorflow graph `.pb` model read in from the `read_graph/1` function (`graph`), an input tensor with the dimensions and data required for the input operation of the graph to run (`tensor1`), an output tensor allocated with the right dimensions (`tensor2`), the name of the input operation of the graph that needs where the input data is fed (`input_opname`), and the output operation name in the graph where the outputs are obtained (`output_opname`). The input tensor is generally created from the matrices manually or using the `load_csv_as_matrix/2` function, and then passed through to one of the tensor creation functions. For image classification the `load_image_as_tensor/1` can also be used to create the input tensor from an image. The output tensor is created using the tensor allocation functions (generally containing `alloc` at the end of the function name). + + Returns a List of Lists (similar to the `matrix_to_lists/1` function) containing the generated predictions as per the output tensor dimensions. + + ## Examples + + - A blog post [here](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/) covers generating predictions and running sessions using an MLP model on the Iris Dataset + + - Generating predictions from the Inception model by Google is covered in the `load_image_as_tensor/1` function examples. + + - Working with an RNN-LSTM example for sentiment analysis is covered [here](https://github.com/anshuman23/tensorflex/pull/25). + """ + def run_session(%Graph{def: graphdef, name: filepath}, %Tensor{datatype: input_datatype, tensor: input_ref}, %Tensor{datatype: output_datatype, tensor: output_ref}, input_opname, output_opname) do NIFs.run_session(graphdef, input_ref, output_ref, input_opname, output_opname) end diff --git a/mix.exs b/mix.exs index 3261696..4d440b1 100644 --- a/mix.exs +++ b/mix.exs @@ -25,7 +25,8 @@ defmodule Tensorflex.MixProject do [ # {:dep_from_hexpm, "~> 0.3.0"}, # {:dep_from_git, git: "https://github.com/elixir-lang/my_dep.git", tag: "0.1.0"}, - {:elixir_make, "~> 0.4", runtime: false} + {:elixir_make, "~> 0.4", runtime: false}, + {:ex_doc, "~> 0.18.0", only: :dev, runtime: false} ] end end From 4dbc8bfa7dd745de28b7fbdfc71ed5634ab15a2f Mon Sep 17 00:00:00 2001 From: Anshuman Chhabra Date: Wed, 25 Jul 2018 16:23:42 +0530 Subject: [PATCH 2/5] Added doc/ folder --- doc/.build | 12 + doc/404.html | 101 ++ doc/Tensorflex.html | 1399 ++++++++++++++++++++++++++ doc/api-reference.html | 123 +++ doc/dist/app-480ffdc169.css | 1 + doc/dist/app-9bd040e5e5.js | 8 + doc/dist/sidebar_items-3a30d3745e.js | 1 + doc/fonts/icomoon.eot | Bin 0 -> 3096 bytes doc/fonts/icomoon.svg | 18 + doc/fonts/icomoon.ttf | Bin 0 -> 2932 bytes doc/fonts/icomoon.woff | Bin 0 -> 3008 bytes doc/index.html | 11 + doc/search.html | 96 ++ 13 files changed, 1770 insertions(+) create mode 100644 doc/.build create mode 100644 doc/404.html create mode 100644 doc/Tensorflex.html create mode 100644 doc/api-reference.html create mode 100644 doc/dist/app-480ffdc169.css create mode 100644 doc/dist/app-9bd040e5e5.js create mode 100644 doc/dist/sidebar_items-3a30d3745e.js create mode 100644 doc/fonts/icomoon.eot create mode 100644 doc/fonts/icomoon.svg create mode 100644 doc/fonts/icomoon.ttf create mode 100644 doc/fonts/icomoon.woff create mode 100644 doc/index.html create mode 100644 doc/search.html diff --git a/doc/.build b/doc/.build new file mode 100644 index 0000000..149b262 --- /dev/null +++ b/doc/.build @@ -0,0 +1,12 @@ +dist/app-480ffdc169.css +dist/app-9bd040e5e5.js +fonts/icomoon.eot +fonts/icomoon.svg +fonts/icomoon.ttf +fonts/icomoon.woff +dist/sidebar_items-3a30d3745e.js +api-reference.html +search.html +404.html +Tensorflex.html +index.html diff --git a/doc/404.html b/doc/404.html new file mode 100644 index 0000000..409102f --- /dev/null +++ b/doc/404.html @@ -0,0 +1,101 @@ + + + + + + + + 404 – tensorflex v0.1.0 + + + + + + + + + + + +
+ + +
+ + + +
+

+ tensorflex +

+

+ v0.1.0 +

+
+ + + +
+ + +
+ + +
+ +
+ +
+
+
+ + +

Page not found

+ +

Sorry, but the page you were trying to get to, does not exist. You +may want to try searching this site using the sidebar or using our +API Reference page to find what +you were looking for.

+ + +
+
+
+
+ + + + + + + + diff --git a/doc/Tensorflex.html b/doc/Tensorflex.html new file mode 100644 index 0000000..ebbd702 --- /dev/null +++ b/doc/Tensorflex.html @@ -0,0 +1,1399 @@ + + + + + + + + Tensorflex – tensorflex v0.1.0 + + + + + + + + + + + +
+ + +
+ + + +
+

+ tensorflex +

+

+ v0.1.0 +

+
+ + + +
+ + +
+ + +
+ +
+ +
+
+
+ + +

+ tensorflex v0.1.0 + Tensorflex + +

+ + +
+

A simple and fast library for running Tensorflow graph models in Elixir. Tensorflex is written around the Tensorflow C API, and allows Elixir developers to leverage Machine Learning and Deep Learning solutions in their projects.

+

NOTE:

+
    +

  • Make sure that the C API version and Python API version (assuming you are using the Python API for first training your models) are the latest. As of July 2018, the latest version is r1.9.

    +
    • +

  • Since Tensorflex provides Inference capability for pre-trained graph models, it is assumed you have adequate knowledge of the pre-trained models you are using (such as the input data type/dimensions, input and output operation names, etc.). Some basic understanding of the Tensorflow Python API can come in very handy.

    +
    • +

  • Tensorflex consists of multiple NIFs, so exercise caution while using it— providing incorrect operation names for running sessions, incorrect dimensions of tensors than the actual pre-trained graph requires, providing different tensor datatypes than the ones required by the graph can all lead to failure. While these are not easy errors to make, do ensure that you test your solution well before deployment.

    +
    • +
+ +
+ + + +
+

+ + + Link to this section + + Summary +

+ + + +
+

+ Functions +

+
+
+ append_to_matrix(matrix, datalist) +
+ +

Appends a single row to the back of a Tensorflex matrix

+
+ +
+
+
+ create_matrix(nrows, ncols, datalist) +
+ +

Creates a 2-D Tensorflex matrix from custom input specifications

+
+ +
+
+
+ float32_tensor(floatval) +
+ +

Creates a TF_FLOAT constant value one-dimensional tensor from the floating point value specified

+
+ +
+
+
+ float32_tensor(matrix1, matrix2) +
+ +

Creates a TF_FLOAT tensor from Tensorflex matrices containing the values and dimensions specified

+
+ +
+
+
+ float32_tensor_alloc(matrix) +
+ +

Allocates a TF_FLOAT tensor of specified dimensions

+
+ +
+
+
+ float64_tensor(floatval) +
+ +

Creates a TF_DOUBLE constant value one-dimensional tensor from the floating point value specified

+
+ +
+
+
+ float64_tensor(matrix1, matrix2) +
+ +

Creates a TF_DOUBLE tensor from Tensorflex matrices containing the values and dimensions specified

+
+ +
+
+
+ float64_tensor_alloc(matrix) +
+ +

Allocates a TF_DOUBLE tensor of specified dimensions

+
+ +
+
+
+ get_graph_ops(graph) +
+ +

Used for listing all the operations in a Tensorflow .pb graph

+
+ +
+
+
+ int32_tensor(intval) +
+ +

Creates a TF_INT32 constant value one-dimensional tensor from the integer value specified

+
+ +
+
+
+ int32_tensor(matrix1, matrix2) +
+ +

Creates a TF_INT32 tensor from Tensorflex matrices containing the values and dimensions specified

+
+ +
+
+
+ int32_tensor_alloc(matrix) +
+ +

Allocates a TF_INT32 tensor of specified dimensions

+
+ +
+
+
+ load_csv_as_matrix(filepath, opts \\ []) +
+ +

Loads high-dimensional data from a CSV file as a Tensorflex 2-D matrix in a super-fast manner

+
+ +
+
+
+ load_image_as_tensor(imagepath) +
+ +

Loads JPEG images into Tensorflex directly as a TF_UINT8 tensor of dimensions image height x image width x number of color channels

+
+ +
+
+
+ matrix_pos(matrix, row, col) +
+ +

Used for accessing an element of a Tensorflex matrix

+
+ +
+
+
+ matrix_to_lists(matrix) +
+ +

Converts a Tensorflex matrix (back) to a list of lists format

+
+ +
+
+
+ read_graph(filepath) +
+ +

Used for loading a Tensorflow .pb graph model in Tensorflex

+
+ +
+
+
+ run_session(graph, tensor1, tensor2, input_opname, output_opname) +
+ +

Runs a Tensorflow session to generate predictions for a given graph, input data, and required input/output operations

+
+ +
+
+
+ size_of_matrix(matrix) +
+ +

Used for obtaining the size of a Tensorflex matrix

+
+ +
+
+
+ string_tensor(stringval) +
+ +

Creates a TF_STRING constant value string tensor from the string value specified

+
+ +
+
+
+ tensor_datatype(tensor) +
+ +

Used to get the datatype of a created tensor

+
+ +
+ +
+ + + + +
+ + + + + +
+

+ + + Link to this section + + Functions +

+
+ + +
+ + + Link to this function + + append_to_matrix(matrix, datalist) + + + +
+
+

Appends a single row to the back of a Tensorflex matrix.

+

Takes a Tensorflex %Matrix matrix as input and a single row of data (with the same number of columns as the original matrix) as a list of lists (datalist) to append to the original matrix.

+

Returns the extended and modified %Matrix struct matrix.

+

+ + Examples +

+ +
iex(1)> m = Tensorflex.create_matrix(2,3,[[23,23,23],[32,32,32]])
+%Tensorflex.Matrix{
+data: #Reference<0.153563642.2042232833.193025>,
+ncols: 3,
+nrows: 2
+}
+
+iex(2)> m = Tensorflex.append_to_matrix(m,[[2,2,2]])
+%Tensorflex.Matrix{
+data: #Reference<0.153563642.2042232833.193025>,
+ncols: 3,
+nrows: 3
+}
+
+iex(3)> m = Tensorflex.append_to_matrix(m,[[3,3,3]])
+%Tensorflex.Matrix{
+data: #Reference<0.153563642.2042232833.193025>,
+ncols: 3,
+nrows: 4
+}
+
+iex(4)> m |> Tensorflex.matrix_to_lists
+[[23.0, 23.0, 23.0], [32.0, 32.0, 32.0], [2.0, 2.0, 2.0], [3.0, 3.0, 3.0]]
+
+

Incorrect usage will raise:

+
iex(5)> m = Tensorflex.append_to_matrix(m,[[2,2,2],[3,3,3]])
+** (ArgumentError) data columns must be same as matrix and number of rows must be 1
+(tensorflex) lib/tensorflex.ex:345: Tensorflex.append_to_matrix/2
+
+iex(5)> m = Tensorflex.append_to_matrix(m,[[2,2,2,2]])      
+** (ArgumentError) data columns must be same as matrix and number of rows must be 1
+(tensorflex) lib/tensorflex.ex:345: Tensorflex.append_to_matrix/2
+ +
+
+
+ + +
+ + + Link to this function + + create_matrix(nrows, ncols, datalist) + + + +
+
+

Creates a 2-D Tensorflex matrix from custom input specifications.

+

Takes three input arguments: number of rows in matrix (nrows), number of columns in matrix (ncols), and a list of lists of the data that will form the matrix (datalist).

+

Returns a %Matrix Tensorflex struct type.

+

+ + Examples: +

+ +

Creating a new matrix

+
iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]])    %Tensorflex.Matrix{
+data: #Reference<0.759278808.823525378.128525>,
+ncols: 3,
+nrows: 2
+}
+

All %Matrix Tensorflex matrices can be passed in to the other matrix inspection and manipulation functions— matrix_pos/3,size_of_matrix/1, matrix_to_lists/1, and append_to_matrix/2:

+
iex(1)> mat = Tensorflex.create_matrix(4,4,[[123,431,23,1],[1,2,3,4],[5,6,7,8],[768,564,44,5]])
+%Tensorflex.Matrix{
+data: #Reference<0.878138179.2435973124.131489>,
+ncols: 4,
+nrows: 4
+}
+
+iex(2)> mat = Tensorflex.append_to_matrix(mat, [[1,1,1,1]])
+%Tensorflex.Matrix{
+data: #Reference<0.878138179.2435973124.131489>,
+ncols: 4,
+nrows: 5
+}
+
+iex(3)> Tensorflex.matrix_to_lists mat
+[
+[123.0, 431.0, 23.0, 1.0],
+[1.0, 2.0, 3.0, 4.0],
+[5.0, 6.0, 7.0, 8.0],
+[768.0, 564.0, 44.0, 5.0],
+[1.0, 1.0, 1.0, 1.0]
+]
+
+iex(4)> Tensorflex.matrix_pos(mat,5,3)
+1.0
+
+iex(5)> Tensorflex.size_of_matrix mat
+{5, 4}
+

Incorrect usage will raise:

+
iex(1)> Tensorflex.create_matrix(1,2,[[1,2,3]])
+** (ArgumentError) argument error
+(tensorflex) Tensorflex.NIFs.create_matrix(1, 2, [[1, 2, 3]])
+(tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3
+
+iex(1)> Tensorflex.create_matrix(2,1,[[1,2,3]])
+** (ArgumentError) argument error
+(tensorflex) Tensorflex.NIFs.create_matrix(2, 1, [[1, 2, 3]])
+(tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3
+
+iex(1)> Tensorflex.create_matrix(2,3,[[1.1,23,3.4], []])
+** (ArgumentError) argument error
+  (tensorflex) Tensorflex.NIFs.create_matrix(2, 3, [[1.1, 23, 3.4], []])
+  (tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3
+  
+iex(1)> Tensorflex.create_matrix(1,2,[[]])              
+** (ArgumentError) data provided cannot be an empty list
+(tensorflex) lib/tensorflex.ex:243: Tensorflex.create_matrix/3
+
+iex(1)> Tensorflex.create_matrix(-1,2,[[3,4]])
+** (FunctionClauseError) no function clause matching in Tensorflex.create_matrix/3
+ +
+
+
+ + +
+ + + Link to this function + + float32_tensor(floatval) + + + +
+
+

Creates a TF_FLOAT constant value one-dimensional tensor from the floating point value specified.

+

Takes in a float value as input.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

+

+ + Examples +

+ +
iex(1)> {:ok, tensor} = Tensorflex.float32_tensor 123.123
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_float,
+tensor: #Reference<0.2011963375.1804468228.236110>
+}}
+
+

Incorrect usage will raise:

+
iex(2)> {:ok, tensor} = Tensorflex.float32_tensor "123.123"
+** (FunctionClauseError) no function clause matching in Tensorflex.float32_tensor/1 
+
+iex(2)> {:ok, tensor} = Tensorflex.float32_tensor 123      
+** (FunctionClauseError) no function clause matching in Tensorflex.float32_tensor/1
+ +
+
+
+ + +
+ + + Link to this function + + float32_tensor(matrix1, matrix2) + + + +
+
+

Creates a TF_FLOAT tensor from Tensorflex matrices containing the values and dimensions specified.

+

Takes two arguments: a %Matrix matrix (matrix1) containing the values the tensor should have and another %Matrix matrix (matrix2) containing the dimensions of the required tensor.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

+

+ + Examples: +

+ +
iex(1)> vals = Tensorflex.create_matrix(2,3,[[12.0,45.2,2.11],[36.7,8.09,9.81]])
+%Tensorflex.Matrix{
+data: #Reference<0.1251941183.3671982081.254268>,
+ncols: 3,
+nrows: 2
+}
+
+iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]])
+%Tensorflex.Matrix{
+data: #Reference<0.1251941183.3671982081.254723>,
+ncols: 2,
+nrows: 1
+}
+  
+iex(3)> {:ok, tensor} = Tensorflex.float32_tensor vals,dims
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_float,
+tensor: #Reference<0.1251941183.3671982081.255228>
+}}
+
+ +
+
+
+ + +
+ + + Link to this function + + float32_tensor_alloc(matrix) + + + +
+
+

Allocates a TF_FLOAT tensor of specified dimensions.

+

This function is generally used to allocate output tensors that do not hold any value data yet, but will after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the run_session/5 function to hold the output values generated as predictions.

+

Takes a Tensorflex %Matrix struct matrix as input.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding the potential tensor data and type.

+

+ + Examples +

+ +

As an example, we can allocate a float32 output tensor that will be a vector of 250 values (1x250 matrix). Therefore, after the session is run, the output will be a float vector containing 250 values:

+
iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float32_tensor_alloc
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_float,
+tensor: #Reference<0.961157994.2087059457.19014>
+}}
+
+ +
+
+
+ + +
+ + + Link to this function + + float64_tensor(floatval) + + + +
+
+

Creates a TF_DOUBLE constant value one-dimensional tensor from the floating point value specified.

+

Takes in a float value as input.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

+

+ + Examples +

+ +
iex(1)> {:ok, tensor} = Tensorflex.float64_tensor 123.123
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_double,
+tensor: #Reference<0.2778616536.4219338753.155412>
+}}
+
+

Incorrect usage will raise:

+
iex(2)> {:ok, tensor} = Tensorflex.float64_tensor "123.123"
+** (FunctionClauseError) no function clause matching in Tensorflex.float64_tensor/1
+
+iex(2)> {:ok, tensor} = Tensorflex.float64_tensor 123      
+** (FunctionClauseError) no function clause matching in Tensorflex.float64_tensor/1
+ +
+
+
+ + +
+ + + Link to this function + + float64_tensor(matrix1, matrix2) + + + +
+
+

Creates a TF_DOUBLE tensor from Tensorflex matrices containing the values and dimensions specified.

+

Takes two arguments: a %Matrix matrix (matrix1) containing the values the tensor should have and another %Matrix matrix (matrix2) containing the dimensions of the required tensor.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

+

+ + Examples: +

+ +
iex(1)> vals = Tensorflex.create_matrix(2,3,[[12.0,45.2,2.11],[36.7,8.09,9.81]])
+%Tensorflex.Matrix{
+data: #Reference<0.1251941183.3671982081.254268>,
+ncols: 3,
+nrows: 2
+}
+
+iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]])
+%Tensorflex.Matrix{
+data: #Reference<0.1251941183.3671982081.254723>,
+ncols: 2,
+nrows: 1
+}
+
+iex(3)> {:ok, tensor} = Tensorflex.float64_tensor vals,dims
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_double,
+tensor: #Reference<0.1251941183.3671982081.255216>
+}}
+
+ +
+
+
+ + +
+ + + Link to this function + + float64_tensor_alloc(matrix) + + + +
+
+

Allocates a TF_DOUBLE tensor of specified dimensions.

+

This function is generally used to allocate output tensors that do not hold any value data yet, but will after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the run_session/5 function to hold the output values generated as predictions.

+

Takes a Tensorflex %Matrix struct matrix as input.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding the potential tensor data and type.

+

+ + Examples +

+ +

As an example, we can allocate a float64 output tensor that will be a vector of 250 values (1x250 matrix). Therefore, after the session is run, the output will be a double vector containing 250 values:

+
iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float64_tensor_alloc
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_double,
+tensor: #Reference<0.961157994.2087059457.19025>
+}}
+
+ +
+
+
+ + +
+ + + Link to this function + + get_graph_ops(graph) + + + +
+
+

Used for listing all the operations in a Tensorflow .pb graph.

+

Reads in a Tensorflex %Graph struct obtained from read_graph/1.

+

Returns a list of all the operation names (as strings) that populate the graph model.

+

+ + Examples +

+ +
    +
  • Google Inception CNN Model (source) +
    • +
+
iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb"
+2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
+{:ok,
+%Tensorflex.Graph{
+def: #Reference<0.3018278404.759824385.5268>,
+name: "classify_image_graph_def.pb"
+}}
+
+iex(2)> Tensorflex.get_graph_ops graph
+["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims",
+"ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul",
+"conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta",
+"conv/batchnorm/gamma", "conv/batchnorm/moving_mean",
+"conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics",
+"conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D",
+"conv_1/batchnorm/beta", "conv_1/batchnorm/gamma",
+"conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance",
+"conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency",
+"conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta",
+"conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean",
+"conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics",
+"conv_2/control_dependency", "conv_2", "pool/CheckNumerics",
+"pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D",
+"conv_3/batchnorm/beta", "conv_3/batchnorm/gamma",
+"conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...]
+
    +
  • Iris Dataset MLP Model (source) +
    • +
+
iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_iris.pb"
+{:ok,
+%Tensorflex.Graph{
+def: #Reference<0.4109712726.1847984130.24506>,
+name: "graphdef_iris.pb"
+}}
+
+iex(2)> Tensorflex.get_graph_ops graph
+["input", "weights1", "weights1/read", "biases1", "biases1/read", "weights2", "weights2/read", "biases2", "biases2/read", "MatMul", "Add", "Relu", "MatMul_1", "Add_1", "output"]
+
+
    +
  • Toy Computational Graph Model (source) +
    • +
+
iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_toy.pb"
+{:ok,
+%Tensorflex.Graph{
+def: #Reference<0.1274892327.1580335105.235135>,
+name: "graphdef_toy.pb"
+}}
+
+iex(2)> Tensorflex.get_graph_ops graph
+["input", "weights", "weights/read", "biases", "biases/read", "MatMul", "add", "output"]
+
    +
  • RNN LSTM Sentiment Analysis Model (source) +
    • +
+
iex(1)> {:ok, graph} = Tensorflex.read_graph "frozen_model_lstm.pb"
+{:ok,
+%Tensorflex.Graph{
+def: #Reference<0.713975820.1050542081.11558>,
+name: "frozen_model_lstm.pb"
+}}
+
+iex(2)> Tensorflex.get_graph_ops graph
+["Placeholder_1", "embedding_lookup/params_0", "embedding_lookup",
+"transpose/perm", "transpose", "rnn/Shape", "rnn/strided_slice/stack",
+"rnn/strided_slice/stack_1", "rnn/strided_slice/stack_2", "rnn/strided_slice",
+"rnn/stack/1", "rnn/stack", "rnn/zeros/Const", "rnn/zeros", "rnn/stack_1/1",
+"rnn/stack_1", "rnn/zeros_1/Const", "rnn/zeros_1", "rnn/Shape_1",
+"rnn/strided_slice_2/stack", "rnn/strided_slice_2/stack_1",
+"rnn/strided_slice_2/stack_2", "rnn/strided_slice_2", "rnn/time",
+"rnn/TensorArray", "rnn/TensorArray_1", "rnn/TensorArrayUnstack/Shape",
+"rnn/TensorArrayUnstack/strided_slice/stack",
+"rnn/TensorArrayUnstack/strided_slice/stack_1",
+"rnn/TensorArrayUnstack/strided_slice/stack_2",
+"rnn/TensorArrayUnstack/strided_slice", "rnn/TensorArrayUnstack/range/start",
+"rnn/TensorArrayUnstack/range/delta", "rnn/TensorArrayUnstack/range",
+"rnn/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3",
+"rnn/while/Enter", "rnn/while/Enter_1", "rnn/while/Enter_2",
+"rnn/while/Enter_3", "rnn/while/Merge", "rnn/while/Merge_1",
+"rnn/while/Merge_2", "rnn/while/Merge_3", "rnn/while/Less/Enter",
+"rnn/while/Less", "rnn/while/LoopCond", "rnn/while/Switch",
+"rnn/while/Switch_1", "rnn/while/Switch_2", "rnn/while/Switch_3", ...]
+ +
+
+
+ + +
+ + + Link to this function + + int32_tensor(intval) + + + +
+
+

Creates a TF_INT32 constant value one-dimensional tensor from the integer value specified.

+

Takes in an integer value as input.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

+

+ + Examples +

+ +
iex(1)> {:ok, tensor} = Tensorflex.int32_tensor 123
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_int32,
+tensor: #Reference<0.1927663658.3415343105.162588>
+}}
+

Incorrect usage will raise:

+
iex(2)> {:ok, tensor} = Tensorflex.int32_tensor 123.123
+** (FunctionClauseError) no function clause matching in Tensorflex.int32_tensor/1 
+
+iex(2)> {:ok, tensor} = Tensorflex.int32_tensor "123.123"
+** (FunctionClauseError) no function clause matching in Tensorflex.int32_tensor/1  
+
+ +
+
+
+ + +
+ + + Link to this function + + int32_tensor(matrix1, matrix2) + + + +
+
+

Creates a TF_INT32 tensor from Tensorflex matrices containing the values and dimensions specified.

+

Takes two arguments: a %Matrix matrix (matrix1) containing the values the tensor should have and another %Matrix matrix (matrix2) containing the dimensions of the required tensor.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

+

NOTE: In case floating point values are passed in the values matrix (matrix1) as arguments for this function, the tensor will still be created and all the float values will be typecast to integers.

+

+ + Examples: +

+ +
iex(1)> vals = Tensorflex.create_matrix(2,3,[[123,45,333],[2,2,899]]) 
+%Tensorflex.Matrix{
+data: #Reference<0.1256144000.2868510721.170449>,
+ncols: 3,
+nrows: 2
+}
+
+iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]])
+%Tensorflex.Matrix{
+data: #Reference<0.1256144000.2868510721.170894>,
+ncols: 2,
+nrows: 1
+}
+
+iex(3)> {:ok, tensor} = Tensorflex.int32_tensor vals,dims
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_int32,
+tensor: #Reference<0.1256144000.2868510721.171357>
+}}
+
+ +
+
+
+ + +
+ + + Link to this function + + int32_tensor_alloc(matrix) + + + +
+
+

Allocates a TF_INT32 tensor of specified dimensions.

+

This function is generally used to allocate output tensors that do not hold any value data yet, but will after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the run_session/5 function to hold the output values generated as predictions.

+

Takes a Tensorflex %Matrix struct matrix as input.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding the potential tensor data and type.

+

+ + Examples +

+ +

As an example, we can allocate an int32 output tensor that will be a vector of 250 values (1x250 matrix). Therefore, after the session is run, the output will be an integer vector containing 250 values:

+
iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.int32_tensor_alloc
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_int32,
+tensor: #Reference<0.961157994.2087059457.18950>
+}}
+
+ +
+
+
+ + + + +
+ + + Link to this function + + load_csv_as_matrix(filepath, opts \\ []) + + + +
+
+

Loads high-dimensional data from a CSV file as a Tensorflex 2-D matrix in a super-fast manner.

+

The load_csv_as_matrix/2 function is very fast— when compared with the Python based pandas library for data science and analysis’ function read_csv on the test.csv file from MNIST Kaggle data (source), the following execution times were obtained:

+ +

This function takes in 2 arguments: a path to a valid CSV file (filepath) and other optional arguments opts. These include whether or not a header needs to be discarded in the CSV, and what the delimiter type is. These are specified by passing in an atom :true or :false to the header: key, and setting a string value for the delimiter: key. By default, the header is considered to be present (:true) and the delimiter is set to ,.

+

Returns a %Matrix Tensorflex struct type.

+

+ + Examples: +

+ +

We first exemplify the working with the test.csv file which belongs to the MNIST Kaggle CSV data (source), which contains 28000 rows and 784 columns (without the header). It is comma delimited and also contains a header. From the test.csv file, we also create a custom file withou the header present which we refer to as test_without_header.csv in the examples below:

+
iex(1)> mat = Tensorflex.load_csv_as_matrix("test.csv")
+%Tensorflex.Matrix{
+data: #Reference<0.4024686574.590479361.258459>,
+ncols: 784,
+nrows: 28000
+}
+
+iex(2)> Tensorflex.matrix_pos mat, 5,97
+80.0
+
+iex(3)> Tensorflex.matrix_pos mat, 5,96
+13.0
+

On a visual inspection of the very large test.csv file, one can see that the values in these particular positions are correct. Now we show usage for the same file but without header, test_without_header.csv:

+
iex(1)> no_header = Tensorflex.load_csv_as_matrix("test/test_without_header.csv", header: :false)    
+%Tensorflex.Matrix{
+data: #Reference<0.4024686574.590479364.257078>,
+ncols: 784,
+nrows: 28000
+}
+
+iex(2)> Tensorflex.matrix_pos no_header,5,97
+80.0
+
+iex(3)> Tensorflex.matrix_pos no_header,5,96
+13.0
+

Next we see the delimiter functionalities. First, assuming we have two simple CSV files, sample1.csv and sample2.csv

+

sample1.csv:

+
1,2,3,4,5
+6,7,8,9,10
+11,12,13,14,15
+

sample2.csv:

+
col1-col2-col3-col4
+1-2-3-4
+5-6-7-8
+9-10-11-12
+

The examples are as follows:

+
iex(1)> m1 = Tensorflex.load_csv_as_matrix("sample1.csv", header: :false) 
+%Tensorflex.Matrix{
+data: #Reference<0.3878093040.3013214209.247502>,
+ncols: 5,
+nrows: 3
+}
+
+iex(2)> Tensorflex.matrix_to_lists m1
+[
+[1.0, 2.0, 3.0, 4.0, 5.0],
+[6.0, 7.0, 8.0, 9.0, 10.0],
+[11.0, 12.0, 13.0, 14.0, 15.0]
+]
+
+iex(3)> m2 = Tensorflex.load_csv_as_matrix("sample2.csv", header: :true, delimiter: "-")
+%Tensorflex.Matrix{
+data: #Reference<0.4024686574.590479361.258952>,
+ncols: 4,
+nrows: 3
+}
+
+iex(4)> Tensorflex.matrix_to_lists m2
+[[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0], [9.0, 10.0, 11.0, 12.0]]
+

Incorrect usage will raise:

+
iex(1)> not_working = Tensorflex.load_csv_as_matrix("test.csv", header: :no_header, delimiter: ",")
+** (ArgumentError) header indicator atom must be either :true or :false 
+(tensorflex) lib/tensorflex.ex:122: Tensorflex.load_csv_as_matrix/2
+ +
+
+
+ + +
+ + + Link to this function + + load_image_as_tensor(imagepath) + + + +
+
+

Loads JPEG images into Tensorflex directly as a TF_UINT8 tensor of dimensions image height x image width x number of color channels.

+

This function is very useful if you wish to do image classification using Convolutional Neural Networks, or other Deep Learning Models. One of the most widely adopted and robust image classification models is the Inception model by Google. It makes classifications on images from over a 1000 classes with highly accurate results. The load_image_as_tensor/1 function is an essential component for the prediction pipeline of the Inception model (and for other similar image classification models) to work in Tensorflex.

+

Reads in the path to a JPEG image file (.jpg or .jpeg).

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding the tensor data and type. Here the created Tensor is a uint8 tensor (TF_UINT8).

+

NOTE: For now, only 3 channel RGB JPEG color images can be passed as arguments. Support for grayscale images and other image formats such as PNG will be added in the future.

+

+ + Examples +

+ +

To exemplify the working of the load_image_as_tensor/1 function we will cover the entire prediction pipeline for the Inception model. However, this makes use of many other Tensorflex functions such as run_session/5 and the other tensor functions so it would be advisable to go through them first. Also, the Inception model can be downloaded here. We will be making use of the cropped_panda.jpg image file that comes along with the model to test out the model in Tensorflex.

+

First the graph is loaded:

+
iex(1)> {:ok, graph} = Tensorflex.read_graph("classify_image_graph_def.pb")
+2018-07-25 14:20:29.079139: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
+{:ok,
+%Tensorflex.Graph{
+def: #Reference<0.542869014.389152771.105680>,
+name: "classify_image_graph_def.pb"
+}}
+

Then we load the image as a uint8 tensor:

+
iex(2)> {:ok, input_tensor} = Tensorflex.load_image_as_tensor("cropped_panda.jpg")
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_uint8,
+tensor: #Reference<0.1203951739.122552322.52747>
+}}
+

Then we create the output tensor which will hold out output vector values. For the Inception model, the output is received as a 1008x1 float32 tensor, as there are 1008 classes in the model:

+
iex(3)> {:ok, output_tensor} = Tensorflex.create_matrix(1,2,[[1008,1]]) |> Tensorflex.float32_tensor_alloc
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_float,
+tensor: #Reference<0.1203951739.122552322.52794>
+}}
+

Next, we obtain the results by running the session:

+
iex(4)> results = Tensorflex.run_session(graph, input_tensor, output_tensor, "DecodeJpeg", "softmax")
+2018-07-25 14:33:40.992813: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA
+[
+[1.059142014128156e-4, 2.8240500250831246e-4, 8.30648496048525e-5,
+1.2982363114133477e-4, 7.32232874725014e-5, 8.014426566660404e-5,
+6.63459359202534e-5, 0.003170756157487631, 7.931600703159347e-5,
+3.707312498590909e-5, 3.0997329304227605e-5, 1.4232713147066534e-4,
+1.0381334868725389e-4, 1.1057958181481808e-4, 1.4321311027742922e-4,
+1.203602587338537e-4, 1.3130248407833278e-4, 5.850398520124145e-5,
+2.641105093061924e-4, 3.1629020668333396e-5, 3.906813799403608e-5,
+2.8646905775531195e-5, 2.2863158665131778e-4, 1.2222197256051004e-4,
+5.956588938715868e-5, 5.421260357252322e-5, 5.996063555357978e-5,
+4.867801326327026e-4, 1.1005574924638495e-4, 2.3433618480339646e-4,
+1.3062104699201882e-4, 1.317620772169903e-4, 9.388553007738665e-5,
+7.076268957462162e-5, 4.281177825760096e-5, 1.6863139171618968e-4,
+9.093972039408982e-5, 2.611844101920724e-4, 2.7584232157096267e-4,
+5.157176201464608e-5, 2.144951868103817e-4, 1.3628098531626165e-4,
+8.007588621694595e-5, 1.7929042223840952e-4, 2.2831936075817794e-4,
+6.216531619429588e-5, 3.736453436431475e-5, 6.782123091397807e-5,
+1.1538144462974742e-4, ...]
+]
+
+

Finally, we need to find which class has the maximum probability and identify it’s label. Since results is a List of Lists, it’s better to read in the flattened list. Then we need to find the index of the element in the new list which as the maximum value. Therefore:

+
iex(5)> max_prob = List.flatten(results) |> Enum.max
+0.8849328756332397
+
+iex(6)> Enum.find_index(results |> List.flatten, fn(x) -> x == max_prob end)
+169
+

We can thus see that the class with the maximum probability predicted (0.8849328756332397) for the image is 169. We will now find what the 169 label corresponds to. For this we can look back into the unzipped Inception folder, where there is a file called imagenet_2012_challenge_label_map_proto.pbtxt. On opening this file, we can find the string class identifier for the 169 class index. This is n02510455 and is present on Line 1556 in the file. Finally, we need to match this string identifier to a set of identification labels by referring to the file imagenet_synset_to_human_label_map.txt file. Here we can see that corresponding to the string class n02510455 the human labels are giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca (Line 3691 in the file). Thus, we have correctly identified the animal in the image as a panda using Tensorflex.

+ +
+
+
+ + +
+ + + Link to this function + + matrix_pos(matrix, row, col) + + + +
+
+

Used for accessing an element of a Tensorflex matrix.

+

Takes in three input arguments: a Tensorflex %Matrix struct matrix, and the row (row) and column (col) values of the required element in the matrix. Both row and col here are NOT zero indexed.

+

Returns the value as float.

+

+ + Examples +

+ +
iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]])
+%Tensorflex.Matrix{
+data: #Reference<0.759278808.823525378.128525>,
+ncols: 3,
+nrows: 2
+}
+
+iex(2)> Tensorflex.matrix_pos(mat,2,1)
+5.5
+
+iex(3)> Tensorflex.matrix_pos(mat,1,3)
+44.5
+
+ +
+
+
+ + +
+ + + Link to this function + + matrix_to_lists(matrix) + + + +
+
+

Converts a Tensorflex matrix (back) to a list of lists format.

+

Takes a Tensorflex %Matrix struct matrix as input.

+

Returns a list of lists representing the data stored in the matrix.

+

NOTE: If the matrix contains very high dimensional data, typically obtained from a function like load_csv_as_matrix/2, then it is not recommended to convert the matrix back to a list of lists format due to a possibility of memory errors.

+

+ + Examples +

+ +
iex(1)> Tensorflex.create_matrix(2,3,[[23,23,23],[32,32,32]]) |> Tensorflex.matrix_to_lists
+[[23.0, 23.0, 23.0], [32.0, 32.0, 32.0]]
+ +
+
+
+ + +
+ + + Link to this function + + read_graph(filepath) + + + +
+
+

Used for loading a Tensorflow .pb graph model in Tensorflex.

+

Reads in a pre-trained Tensorflow protobuf (.pb) Graph model binary file.

+

Returns a tuple {:ok, %Graph}.

+

%Graph is an internal Tensorflex struct which holds the name of the graph file and the binary definition data that is read in via the .pb file.

+

+ + Examples: +

+ +

Reading in a graph

+

As an example, we can try reading in the Inception convolutional neural network based image classification graph model by Google. The graph file is named classify_image_graph_def.pb:

+
iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb"
+2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
+{:ok,
+%Tensorflex.Graph{
+def: #Reference<0.3018278404.759824385.5268>,
+name: "classify_image_graph_def.pb"
+}}
+

Generally to check that the loaded graph model is correct and contains computational operations, the get_graph_ops/1 function is useful:

+
iex(2)> Tensorflex.get_graph_ops graph
+["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims",
+"ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul",
+"conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta",
+"conv/batchnorm/gamma", "conv/batchnorm/moving_mean",
+"conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics",
+"conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D",
+"conv_1/batchnorm/beta", "conv_1/batchnorm/gamma",
+"conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance",
+"conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency",
+"conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta",
+"conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean",
+"conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics",
+"conv_2/control_dependency", "conv_2", "pool/CheckNumerics",
+"pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D",
+"conv_3/batchnorm/beta", "conv_3/batchnorm/gamma",
+"conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...]
+
+

Incorrect usage will raise:

+
iex(3)> {:ok, graph} = Tensorflex.read_graph "Makefile"
+** (ArgumentError) file is not a protobuf .pb file
+(tensorflex) lib/tensorflex.ex:27: Tensorflex.read_graph/1
+
+iex(3)> {:ok, graph} = Tensorflex.read_graph "Makefile.pb"
+** (ArgumentError) graph definition file does not exist
+(tensorflex) lib/tensorflex.ex:23: Tensorflex.read_graph/1
+
+ +
+
+
+ + +
+ + + Link to this function + + run_session(graph, tensor1, tensor2, input_opname, output_opname) + + + +
+
+

Runs a Tensorflow session to generate predictions for a given graph, input data, and required input/output operations.

+

This function is the final step of the Inference (prediction) pipeline and generates output for a given set of input data, a pre-trained graph model, and the specified input and output operations of the graph.

+

Takes in five arguments: a pre-trained Tensorflow graph .pb model read in from the read_graph/1 function (graph), an input tensor with the dimensions and data required for the input operation of the graph to run (tensor1), an output tensor allocated with the right dimensions (tensor2), the name of the input operation of the graph that needs where the input data is fed (input_opname), and the output operation name in the graph where the outputs are obtained (output_opname). The input tensor is generally created from the matrices manually or using the load_csv_as_matrix/2 function, and then passed through to one of the tensor creation functions. For image classification the load_image_as_tensor/1 can also be used to create the input tensor from an image. The output tensor is created using the tensor allocation functions (generally containing alloc at the end of the function name).

+

Returns a List of Lists (similar to the matrix_to_lists/1 function) containing the generated predictions as per the output tensor dimensions.

+

+ + Examples +

+ +
    +

  • A blog post here covers generating predictions and running sessions using an MLP model on the Iris Dataset

    +
    • +

  • Generating predictions from the Inception model by Google is covered in the load_image_as_tensor/1 function examples.

    +
    • +

  • Working with an RNN-LSTM example for sentiment analysis is covered here.

    +
    • +
+ +
+
+
+ + +
+ + + Link to this function + + size_of_matrix(matrix) + + + +
+
+

Used for obtaining the size of a Tensorflex matrix.

+

Takes a Tensorflex %Matrix struct matrix as input.

+

Returns a tuple {nrows, ncols} where nrows represents the number of rows of the matrix and ncols represents the number of columns of the matrix.

+

+ + Examples +

+ +
iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]])
+%Tensorflex.Matrix{
+data: #Reference<0.759278808.823525378.128525>,
+ncols: 3,
+nrows: 2
+}
+
+iex(2)> Tensorflex.size_of_matrix mat
+{2, 3}
+ +
+
+
+ + +
+ + + Link to this function + + string_tensor(stringval) + + + +
+
+

Creates a TF_STRING constant value string tensor from the string value specified.

+

Takes in a string value as input.

+

Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

+

+ + Examples +

+ +
iex(1)> {:ok, tensor} = Tensorflex.string_tensor "123.123"
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_string,
+tensor: #Reference<0.2069282048.194904065.41126>
+}}
+
+

Incorrect usage will raise:

+
iex(2)> {:ok, tensor} = Tensorflex.string_tensor 123.123  
+** (FunctionClauseError) no function clause matching in Tensorflex.string_tensor/1 
+
+iex(2)> {:ok, tensor} = Tensorflex.string_tensor 123    
+** (FunctionClauseError) no function clause matching in Tensorflex.string_tensor/1
+ +
+
+
+ + +
+ + + Link to this function + + tensor_datatype(tensor) + + + +
+
+

Used to get the datatype of a created tensor.

+

Takes in a %Tensor struct tensor as input.

+

Returns a tuple {:ok, datatype} where datatype is an atom representing the list of Tensorflow TF_DataType tensor datatypes. Click here to view a list of all possible datatypes.

+

+ + Examples +

+ +
iex(1)> {:ok, tensor} = Tensorflex.string_tensor "example"
+{:ok,
+%Tensorflex.Tensor{
+datatype: :tf_string,
+tensor: #Reference<0.4132928949.2894987267.194583>
+}}
+
+iex(2)> Tensorflex.tensor_datatype tensor
+{:ok, :tf_string}
+ +
+
+ +
+ + + + +
+
+
+
+ + + + + + + + diff --git a/doc/api-reference.html b/doc/api-reference.html new file mode 100644 index 0000000..ce885f9 --- /dev/null +++ b/doc/api-reference.html @@ -0,0 +1,123 @@ + + + + + + + + API Reference – tensorflex v0.1.0 + + + + + + + + + + + +
+ + +
+ + + +
+

+ tensorflex +

+

+ v0.1.0 +

+
+ + + +
+ + +
+ + +
+ +
+ +
+
+
+ + +

+ tensorflex v0.1.0 + API Reference +

+ + +
+

+ + Modules +

+ +
+
+
Tensorflex
+ +

A simple and fast library for running Tensorflow graph models in Elixir. Tensorflex is written around the Tensorflow C API, and allows Elixir developers to leverage Machine Learning and Deep Learning solutions in their projects

+
+ +
+ +
+
+ + + + + + + +
+
+
+
+ + + + + + + + diff --git a/doc/dist/app-480ffdc169.css b/doc/dist/app-480ffdc169.css new file mode 100644 index 0000000..c30bd32 --- /dev/null +++ b/doc/dist/app-480ffdc169.css @@ -0,0 +1 @@ +@import url(https://fonts.googleapis.com/css?family=Lato:300,700|Merriweather:300italic,300|Inconsolata:400,700);.hljs,article,aside,details,figcaption,figure,footer,header,hgroup,main,menu,nav,section,summary{display:block}.hljs-strong,b,optgroup,strong{font-weight:700}.hljs-emphasis,dfn{font-style:italic}img,legend{border:0}#search ul,.sidebar ul{list-style:none}.night-mode-toggle:focus,.sidebar .sidebar-search .search-input:focus,.sidebar .sidebar-search .search-input:hover,.sidebar-button:active,.sidebar-button:focus,.sidebar-button:hover,a:active,a:hover{outline:0}.content-inner .footer a,.content-inner a,body.night-mode .content-inner a{-webkit-text-decoration-skip:ink;text-decoration-skip:ink}.hljs-comment,.hljs-quote{color:#8e908c}.hljs-deletion,.hljs-name,.hljs-regexp,.hljs-selector-class,.hljs-selector-id,.hljs-tag,.hljs-template-variable,.hljs-variable{color:#c82829}.hljs-built_in,.hljs-builtin-name,.hljs-link,.hljs-literal,.hljs-meta,.hljs-number,.hljs-params,.hljs-type{color:#f5871f}.hljs-attribute{color:#eab700}.hljs-addition,.hljs-bullet,.hljs-string,.hljs-symbol{color:#718c00}.hljs-section,.hljs-title{color:#4271ae}.hljs-keyword,.hljs-selector-tag{color:#8959a8}.hljs{overflow-x:auto;background:#fff;color:#4d4d4c;padding:.5em}legend,td,th{padding:0}/*! normalize.css v3.0.3 | MIT License | github.com/necolas/normalize.css */html{font-family:sans-serif;-ms-text-size-adjust:100%;-webkit-text-size-adjust:100%}audio,canvas,progress,video{display:inline-block;vertical-align:baseline}audio:not([controls]){display:none;height:0}[hidden],template{display:none}a{background-color:transparent}abbr[title]{border-bottom:1px dotted}h1{font-size:2em;margin:.67em 0}mark{background:#ff0;color:#000}.content-outer,body{background-color:#fff}small{font-size:80%}sub,sup{font-size:75%;line-height:0;position:relative;vertical-align:baseline}sup{top:-.5em}sub{bottom:-.25em}svg:not(:root){overflow:hidden}figure{margin:1em 40px}hr{box-sizing:content-box;height:0}pre,textarea{overflow:auto}code,kbd,pre,samp{font-family:monospace,monospace;font-size:1em}button,input,optgroup,select,textarea{color:inherit;font:inherit;margin:0}button{overflow:visible}button,select{text-transform:none}button,html input[type=button],input[type=reset],input[type=submit]{-webkit-appearance:button;cursor:pointer}button[disabled],html input[disabled]{cursor:default}button::-moz-focus-inner,input::-moz-focus-inner{border:0;padding:0}input{line-height:normal}input[type=checkbox],input[type=radio]{box-sizing:border-box;padding:0}input[type=number]::-webkit-inner-spin-button,input[type=number]::-webkit-outer-spin-button{height:auto}.content,.main,.sidebar,body,html{height:100%}input[type=search]{-webkit-appearance:textfield;box-sizing:content-box}input[type=search]::-webkit-search-cancel-button,input[type=search]::-webkit-search-decoration{-webkit-appearance:none}fieldset{border:1px solid silver;margin:0 2px;padding:.35em .625em .75em}table{border-collapse:collapse;border-spacing:0}@font-face{font-family:icomoon;src:url(../fonts/icomoon.eot?h5z89e);src:url(../fonts/icomoon.eot?#iefixh5z89e) format('embedded-opentype'),url(../fonts/icomoon.ttf?h5z89e) format('truetype'),url(../fonts/icomoon.woff?h5z89e) format('woff'),url(../fonts/icomoon.svg?h5z89e#icomoon) format('svg');font-weight:400;font-style:normal}.icon-elem,[class*=" icon-"],[class^=icon-]{font-family:icomoon;speak:none;font-style:normal;font-weight:400;font-variant:normal;text-transform:none;line-height:1;-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale}.sidebar,body{font-family:Lato,sans-serif}.icon-link:before{content:"\e005"}.icon-search:before{content:"\e036"}.icon-cross:before{content:"\e117"}@media screen and (max-width:768px){.icon-menu{font-size:1em}}@media screen and (min-width:769px){.icon-menu{font-size:1.25em}}@media screen and (min-width:1281px){.icon-menu{font-size:1.5em}}.icon-menu:before{content:"\e120"}.icon-angle-right:before{content:"\f105"}.icon-code:before{content:"\f121"}body,html{box-sizing:border-box;width:100%}body{margin:0;font-size:16px;line-height:1.6875em}*,:after,:before{box-sizing:inherit}.main{display:-ms-flexbox;display:-ms-flex;display:flex;-ms-flex-pack:end;justify-content:flex-end}.sidebar{display:-ms-flexbox;display:-ms-flex;display:flex;min-height:0;-moz-flex-direction:column;-ms-flex-direction:column;flex-direction:column;width:300px;position:fixed;top:0;left:0;z-index:4}.content{width:100%;padding-left:300px;overflow-y:auto;-webkit-overflow-scrolling:touch;position:relative;z-index:3}@media screen and (max-width:768px){body .content{z-index:0;padding-left:0}body .sidebar{z-index:3;transform:translateX(-102%);will-change:transform}}body.sidebar-closed .sidebar,body.sidebar-closing .sidebar,body.sidebar-opening .sidebar{z-index:0}body.sidebar-opened .sidebar-button,body.sidebar-opening .sidebar-button{transform:translateX(250px)}@media screen and (max-width:768px){body.sidebar-opened .sidebar,body.sidebar-opening .sidebar{transform:translateX(0)}}body.sidebar-closed .content,body.sidebar-closing .content{padding-left:0}body.sidebar-closed .sidebar-button,body.sidebar-closing .sidebar-button{transform:none}body.sidebar-closed .sidebar-button{color:#000}body.sidebar-opening .sidebar-button{transition:transform .3s ease-in-out}body.sidebar-opening .content{padding-left:300px;transition:padding-left .3s ease-in-out}@media screen and (max-width:768px){body.sidebar-opening .content{padding-left:0}body.sidebar-opening .sidebar{transition:transform .3s ease-in-out;z-index:3}}body.sidebar-closing .sidebar-button{transition:transform .3s ease-in-out}body.sidebar-closing .content{transition:padding-left .3s ease-in-out}@media screen and (max-width:768px){body.sidebar-closing .sidebar{z-index:3;transition:transform .3s ease-in-out;transform:translateX(-102%)}}.sidebar a,.sidebar-button{transition:color .3s ease-in-out}body.sidebar-closed .sidebar{visibility:hidden}.content-inner{max-width:949px;margin:0 auto;padding:3px 60px}.content-outer{min-height:100%}@media screen and (max-width:768px){.content-inner{padding:27px 20px 27px 40px}}.sidebar-button{position:fixed;z-index:99;left:18px;color:#e1e1e1;background-color:transparent;border:none;padding:0;font-size:16px;will-change:transform;transform:translateX(250px)}.sidebar-button:hover{color:#e1e1e1}.sidebar-button .sidebar-toggle{top:8px}@media screen and (max-width:768px){.sidebar-button{transform:translateX(0);left:5px;top:5px}.sidebar-opened .sidebar-button{left:18px}}.sidebar{font-size:15px;line-height:18px;background:#373f52;color:#d5dae6;overflow:hidden}.sidebar .gradient{background:linear-gradient(#373f52,rgba(55,63,82,0));height:20px;margin-top:-20px;pointer-events:none;position:relative;top:20px;z-index:100}.sidebar ul li{margin:0;padding:0 10px}.sidebar a{color:#d5dae6;text-decoration:none}.sidebar a:hover{color:#fff}.sidebar .sidebar-projectLink{margin:18px 30px 0}.sidebar .sidebar-projectDetails{display:inline-block;text-align:right;vertical-align:top;margin-top:6px}.sidebar .sidebar-projectImage{display:inline-block;max-width:64px;max-height:64px;margin-left:15px;vertical-align:bottom}.sidebar .sidebar-projectName{font-weight:700;font-size:24px;line-height:30px;color:#fff;margin:0;padding:0;max-width:230px;word-wrap:break-word}.sidebar .sidebar-projectVersion{margin:0;padding:0;font-weight:300;font-size:16px;line-height:20px;color:#fff}.sidebar .sidebar-listNav{padding:10px 30px 20px;margin:0}.sidebar .sidebar-listNav li,.sidebar .sidebar-listNav li a{text-transform:uppercase;font-weight:300;font-size:14px}.sidebar .sidebar-listNav li{padding-left:17px;border-left:3px solid transparent;transition:all .3s linear;line-height:27px}.sidebar .sidebar-listNav li.selected,.sidebar .sidebar-listNav li.selected a,.sidebar .sidebar-listNav li:hover,.sidebar .sidebar-listNav li:hover a{border-color:#9768d1;color:#fff}.sidebar .sidebar-search{margin:18px 30px;display:-ms-flexbox;display:-ms-flex;display:flex}.sidebar .sidebar-search .search-button:hover,.sidebar .sidebar-search.selected .search-button{color:#9768d1}.sidebar .sidebar-search .search-button{font-size:14px;color:#d5dae6;background-color:transparent;border:none;padding:2px 0 0;margin:0}.sidebar .sidebar-search .search-input{background-color:transparent;border:none;border-radius:0;border-bottom:1px solid #959595;margin-left:5px;height:20px}.sidebar .sidebar-search .icon-search{font-weight:700}.sidebar #full-list{margin:0 0 0 30px;padding:10px 20px 40px;overflow-y:auto;-webkit-overflow-scrolling:touch;-moz-flex:1 1 .01%;-ms-flex:1 1 .01%;flex:1 1 .01%;-ms-flex-positive:1;-ms-flex-negative:1;-ms-flex-preferred-size:.01%}.sidebar #full-list ul{display:none;margin:9px 15px;padding:0}.sidebar #full-list ul li{font-weight:300;line-height:18px;padding:2px 10px}.sidebar #full-list ul li a.expand:before{content:"+";font-family:monospaced;font-size:15px;float:left;width:13px;margin-left:-13px}.sidebar #full-list ul li.open a.expand:before{content:"−"}.sidebar #full-list ul li ul{display:none;margin:9px 6px}.sidebar #full-list li.open>ul,.sidebar #full-list ul li.open>ul{display:block}.sidebar #full-list ul li ul li{border-left:1px solid #959595;padding:0 10px}.sidebar #full-list ul li ul li.active:before{font-family:icomoon;speak:none;font-style:normal;font-weight:400;font-variant:normal;text-transform:none;line-height:1;-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale;content:"\f105";margin-left:-10px;font-size:16px;margin-right:5px}.sidebar #full-list li{padding:0;line-height:27px}.sidebar #full-list li.active{border-left:none}.sidebar #full-list li.active>a,.sidebar #full-list li.clicked>a{color:#fff}.sidebar #full-list li.group{text-transform:uppercase;font-weight:700;font-size:.8em;margin:2em 0 0;line-height:1.8em;color:#ddd}@media screen and (max-height:500px){.sidebar{overflow-y:auto}.sidebar #full-list{overflow:visible}}.content-inner{font-family:Merriweather,'Book Antiqua',Georgia,'Century Schoolbook',serif;font-size:1em;line-height:1.6875em}.content-inner h1,.content-inner h2,.content-inner h3,.content-inner h4,.content-inner h5,.content-inner h6{font-family:Lato,sans-serif;font-weight:700;line-height:1.5em;word-wrap:break-word}.content-inner h1{font-size:2em;margin:1em 0 .5em}.content-inner h1.section-heading{margin:1.5em 0 .5em}.content-inner h1 small{font-weight:300}.content-inner h1 a.view-source{font-size:1.2rem}.content-inner h2{font-size:1.6em;margin:1em 0 .5em;font-weight:700}.content-inner h3{font-size:1.375em;margin:1em 0 .5em;font-weight:700}.content-inner a{color:#000;text-decoration:underline}.content-inner a:visited{color:#000}.content-inner ul li{line-height:1.5em}.content-inner ul li>p{margin:0}.content-inner a.view-source{float:right;color:#959595;text-decoration:none;border:none;transition:color .3s ease-in-out;margin-top:1px}.content-inner a.view-source:hover{color:#373f52}.content-inner .note{color:#959595;margin:0 5px;font-size:14px;font-weight:400}.content-inner blockquote{font-style:italic;margin:.5em 0;padding:.25em 1.5em;border-left:3px solid #e1e1e1;display:inline-block}.content-inner blockquote :first-child{padding-top:0;margin-top:0}.content-inner blockquote :last-child{padding-bottom:0;margin-bottom:0}.content-inner table{margin:2em 0}.content-inner th{text-align:left;font-family:Lato,sans-serif;text-transform:uppercase;font-weight:700;padding-bottom:.5em}.content-inner tr{border-bottom:1px solid #d5dae6;vertical-align:bottom;height:2.5em}.content-inner .summary .summary-row .summary-signature a,.content-inner .summary h2 a{border:none;text-decoration:none}.content-inner td,.content-inner th{padding-left:1em;line-height:2em;vertical-align:top}.content-inner .section-heading:hover a.hover-link{opacity:1;text-decoration:none}.content-inner .section-heading a.hover-link{transition:opacity .3s ease-in-out;display:inline-block;opacity:0;padding:.3em .6em .6em;line-height:1em;margin-left:-2.7em;text-decoration:none;border:none;font-size:16px;vertical-align:middle}.content-inner .detail h2.section-heading{margin-left:.3em}.content-inner .visible-xs{display:none!important}@media screen and (max-width:768px){.content-inner .visible-xs{display:block!important}}.content-inner img{max-width:100%}.content-inner .summary h2{font-weight:700}.content-inner .summary .summary-row .summary-signature{font-family:Inconsolata,Menlo,Courier,monospace;font-weight:700}.content-inner .summary .summary-row .summary-synopsis{font-family:Merriweather,'Book Antiqua',Georgia,'Century Schoolbook',serif;font-style:italic;padding:0 1.2em;margin:0 0 .5em}.content-inner .summary .summary-row .summary-synopsis p{margin:0;padding:0}@keyframes blink-background{from{background-color:#f7f7f7}to{background-color:#ff9}}.content-inner .detail:target .detail-header{animation-duration:.55s;animation-name:blink-background;animation-iteration-count:1;animation-timing-function:ease-in-out}.content-inner .detail-header{margin:2em 0 1em;padding:.5em 1em;background:#f7f7f7;border-left:3px solid #9768d1;font-size:1em;font-family:Inconsolata,Menlo,Courier,monospace;position:relative}.content-inner .detail-header .note{float:right}.content-inner .detail-header .signature{font-size:1rem;font-weight:700}.content-inner .detail-header:hover a.detail-link{opacity:1;text-decoration:none}.content-inner .detail-header a.detail-link{transition:opacity .3s ease-in-out;position:absolute;top:0;left:0;display:block;opacity:0;padding:.6em;line-height:1.5em;margin-left:-2.5em;text-decoration:none;border:none}#search h1,.content-inner .footer .line{display:inline-block}.content-inner .specs pre,.content-inner code{font-family:Inconsolata,Menlo,Courier,monospace;font-style:normal;line-height:24px}.content-inner .specs{opacity:.7;padding-bottom:.05em}.content-inner .specs pre{font-size:.9em;white-space:pre-wrap;margin:0;padding:0}.content-inner .docstring{margin:1.2em 0 2.1em 1.2em}.content-inner .docstring h2,.content-inner .docstring h3,.content-inner .docstring h4,.content-inner .docstring h5{font-weight:700}.content-inner .docstring h2{font-size:1.1em}.content-inner .docstring h3{font-size:1em}.content-inner .docstring h4{font-size:.95em}.content-inner .docstring h5{font-size:.9em}.content-inner a.no-underline,.content-inner pre a{color:#9768d1;text-shadow:none;text-decoration:none;background-image:none}.content-inner a.no-underline:active,.content-inner a.no-underline:focus,.content-inner a.no-underline:hover,.content-inner a.no-underline:visited,.content-inner pre a:active,.content-inner pre a:focus,.content-inner pre a:hover,.content-inner pre a:visited{color:#9768d1;text-decoration:none}.content-inner code{font-weight:400;background-color:#f7f9fc;vertical-align:baseline;border-radius:2px;padding:.1em .2em;border:1px solid #d2ddee}.content-inner pre{margin:1.5em 0}.content-inner pre.spec{margin:0}.content-inner pre.spec code{padding:0}.content-inner pre code.hljs{white-space:inherit;padding:.5em 1em;background-color:#f7f9fc}.content-inner .footer{margin:4em auto 1em;text-align:center;font-style:italic;font-size:14px;color:#959595}.content-inner .footer a{color:#959595;text-decoration:underline}.content-inner .footer a:visited{color:#959595}#search{min-height:200px}#search .result-id{font-size:1.2em}#search .result-id a{text-decoration:none;transition:color .3s ease-in-out}#search .result-id a:active,#search .result-id a:focus,#search .result-id a:visited{color:#000}#search .result-id a:hover{color:#9768d1}#search .result-elem em,#search .result-id em{font-style:normal;color:#9768d1}#search ul{margin:0;padding:0}.night-mode-toggle{top:1.5em}.night-mode-toggle .icon-theme::before{content:"\e900"}@media screen and (max-width:768px){.night-mode-toggle .icon-theme::before{font-size:1em}.night-mode-toggle{transform:translateX(0);left:5px;top:1.5em}.sidebar-opened .night-mode-toggle{left:18px}}@media screen and (min-width:769px){.night-mode-toggle .icon-theme::before{font-size:1.25em}}@media screen and (min-width:1281px){.night-mode-toggle .icon-theme::before{font-size:1.5em}}body.night-mode{background:#212127}body.night-mode .hljs-comment,body.night-mode .hljs-quote{color:#969896}body.night-mode .hljs-deletion,body.night-mode .hljs-name,body.night-mode .hljs-regexp,body.night-mode .hljs-selector-class,body.night-mode .hljs-selector-id,body.night-mode .hljs-tag,body.night-mode .hljs-template-variable,body.night-mode .hljs-variable{color:#c66}body.night-mode .hljs-built_in,body.night-mode .hljs-builtin-name,body.night-mode .hljs-link,body.night-mode .hljs-literal,body.night-mode .hljs-meta,body.night-mode .hljs-number,body.night-mode .hljs-params,body.night-mode .hljs-type{color:#de935f}body.night-mode .hljs-attribute{color:#f0c674}body.night-mode .hljs-addition,body.night-mode .hljs-bullet,body.night-mode .hljs-string,body.night-mode .hljs-symbol{color:#b5bd68}body.night-mode .hljs-section,body.night-mode .hljs-title{color:#81a2be}body.night-mode .hljs-keyword,body.night-mode .hljs-selector-tag{color:#b294bb}body.night-mode .hljs{display:block;overflow-x:auto;background:#1d1f21;color:#c5c8c6;padding:.5em}body.night-mode .hljs-emphasis{font-style:italic}body.night-mode .hljs-strong{font-weight:700}body.night-mode .content-outer{background:#212127}body.night-mode .night-mode-toggle .icon-theme::before{content:"\e901"}body.night-mode #search .result-id a:active,body.night-mode #search .result-id a:focus,body.night-mode #search .result-id a:visited{color:#D2D2D2}body.night-mode #search .result-id a:hover{color:#9768d1}body.night-mode .content-inner{color:#B4B4B4}body.night-mode .content-inner a:visited,body.night-mode .content-inner h1,body.night-mode .content-inner h2,body.night-mode .content-inner h3,body.night-mode .content-inner h4,body.night-mode .content-inner h5,body.night-mode .content-inner h6{color:#D2D2D2}body.night-mode .content-inner a{color:#D2D2D2;text-decoration:underline}body.night-mode .content-inner .summary h2 a,body.night-mode .content-inner a.no-underline,body.night-mode .content-inner a.view-source,body.night-mode .content-inner pre a{text-decoration:none}body.night-mode .content-inner a.view-source:hover{color:#fff}@keyframes night-blink-background{from{background-color:#3A4152}to{background-color:#660}}body.night-mode .content-inner .detail:target .detail-header{animation-name:night-blink-background}body.night-mode .content-inner .detail-header{background:#3A4152;color:#D2D2D2}body.night-mode .content-inner .footer,body.night-mode .content-inner .footer a{color:#959595}body.night-mode .content-inner code{background-color:#2C2C31;border-color:#44444c}body.night-mode .content-inner pre code.hljs{background-color:#2C2C31}body.night-mode .content-inner .footer .line{display:inline-block}body.night-mode .sidebar-button,body.night-mode .sidebar-closed .sidebar-button{color:#d5dae6}.sr-only{position:absolute;width:1px;height:1px;padding:0;margin:-1px;overflow:hidden;clip:rect(0,0,0,0);border:0}@media print{.main{display:block}.sidebar,.sidebar-button{display:none}.content{padding-left:0;overflow:visible}.summary-row{page-break-inside:avoid}} \ No newline at end of file diff --git a/doc/dist/app-9bd040e5e5.js b/doc/dist/app-9bd040e5e5.js new file mode 100644 index 0000000..e4a0ccf --- /dev/null +++ b/doc/dist/app-9bd040e5e5.js @@ -0,0 +1,8 @@ +!function(e){function t(r){if(n[r])return n[r].exports;var i=n[r]={exports:{},id:r,loaded:!1};return e[r].call(i.exports,i,i.exports,t),i.loaded=!0,i.exports}var n={};return t.m=e,t.c=n,t.p="",t(0)}([function(e,t,n){"use strict";var r=n(1)["default"],i=n(2),o=r(i),a=n(3),u=r(a),s=n(4),l=n(86),c=n(89);window.$=o["default"],(0,o["default"])(function(){u["default"].configure({tabReplace:" ",languages:[]}),(0,c.initialize)(),(0,l.initialize)(),(0,s.initialize)(),u["default"].initHighlighting()})},function(e,t){"use strict";t["default"]=function(e){return e&&e.__esModule?e:{"default":e}},t.__esModule=!0},function(e,t,n){var r,i;!function(t,n){"use strict";"object"==typeof e&&"object"==typeof e.exports?e.exports=t.document?n(t,!0):function(e){if(!e.document)throw new Error("jQuery requires a window with a document");return n(e)}:n(t)}("undefined"!=typeof window?window:this,function(n,o){"use strict";function a(e,t,n){t=t||ce;var r,i=t.createElement("script");if(i.text=e,n)for(r in ke)n[r]&&(i[r]=n[r]);t.head.appendChild(i).parentNode.removeChild(i)}function u(e){return null==e?e+"":"object"==typeof e||"function"==typeof e?ve[me.call(e)]||"object":typeof e}function s(e){var t=!!e&&"length"in e&&e.length,n=u(e);return!we(e)&&!Ee(e)&&("array"===n||0===t||"number"==typeof t&&t>0&&t-1 in e)}function l(e,t){return e.nodeName&&e.nodeName.toLowerCase()===t.toLowerCase()}function c(e,t,n){return we(t)?Ne.grep(e,function(e,r){return!!t.call(e,r,e)!==n}):t.nodeType?Ne.grep(e,function(e){return e===t!==n}):"string"!=typeof t?Ne.grep(e,function(e){return ge.call(t,e)>-1!==n}):Ne.filter(t,e,n)}function f(e,t){for(;(e=e[t])&&1!==e.nodeType;);return e}function p(e){var t={};return Ne.each(e.match(qe)||[],function(e,n){t[n]=!0}),t}function d(e){return e}function h(e){throw e}function g(e,t,n,r){var i;try{e&&we(i=e.promise)?i.call(e).done(t).fail(n):e&&we(i=e.then)?i.call(e,t,n):t.apply(void 0,[e].slice(r))}catch(e){n.apply(void 0,[e])}}function v(){ce.removeEventListener("DOMContentLoaded",v),n.removeEventListener("load",v),Ne.ready()}function m(e,t){return t.toUpperCase()}function y(e){return e.replace(ze,"ms-").replace(Fe,m)}function b(){this.expando=Ne.expando+b.uid++}function _(e){return"true"===e||"false"!==e&&("null"===e?null:e===+e+""?+e:Ge.test(e)?JSON.parse(e):e)}function x(e,t,n){var r;if(void 0===n&&1===e.nodeType)if(r="data-"+t.replace(Ve,"-$&").toLowerCase(),n=e.getAttribute(r),"string"==typeof n){try{n=_(n)}catch(i){}Ke.set(e,t,n)}else n=void 0;return n}function w(e,t,n,r){var i,o,a=20,u=r?function(){return r.cur()}:function(){return Ne.css(e,t,"")},s=u(),l=n&&n[3]||(Ne.cssNumber[t]?"":"px"),c=(Ne.cssNumber[t]||"px"!==l&&+s)&&Ze.exec(Ne.css(e,t));if(c&&c[3]!==l){for(s/=2,l=l||c[3],c=+s||1;a--;)Ne.style(e,t,c+l),(1-o)*(1-(o=u()/s||.5))<=0&&(a=0),c/=o;c=2*c,Ne.style(e,t,c+l),n=n||[]}return n&&(c=+c||+s||0,i=n[1]?c+(n[1]+1)*n[2]:+n[2],r&&(r.unit=l,r.start=c,r.end=i)),i}function E(e){var t,n=e.ownerDocument,r=e.nodeName,i=et[r];return i?i:(t=n.body.appendChild(n.createElement(r)),i=Ne.css(t,"display"),t.parentNode.removeChild(t),"none"===i&&(i="block"),et[r]=i,i)}function k(e,t){for(var n,r,i=[],o=0,a=e.length;o-1)i&&i.push(o);else if(c=Ne.contains(o.ownerDocument,o),a=C(p.appendChild(o),"script"),c&&N(a),n)for(f=0;o=a[f++];)rt.test(o.type||"")&&n.push(o);return p}function A(){return!0}function S(){return!1}function j(){try{return ce.activeElement}catch(e){}}function O(e,t,n,r,i,o){var a,u;if("object"==typeof t){"string"!=typeof n&&(r=r||n,n=void 0);for(u in t)O(e,u,n,r,t[u],o);return e}if(null==r&&null==i?(i=n,r=n=void 0):null==i&&("string"==typeof n?(i=r,r=void 0):(i=r,r=n,n=void 0)),i===!1)i=S;else if(!i)return e;return 1===o&&(a=i,i=function(e){return Ne().off(e),a.apply(this,arguments)},i.guid=a.guid||(a.guid=Ne.guid++)),e.each(function(){Ne.event.add(this,t,i,r,n)})}function M(e,t){return l(e,"table")&&l(11!==t.nodeType?t:t.firstChild,"tr")?Ne(e).children("tbody")[0]||e:e}function D(e){return e.type=(null!==e.getAttribute("type"))+"/"+e.type,e}function R(e){return"true/"===(e.type||"").slice(0,5)?e.type=e.type.slice(5):e.removeAttribute("type"),e}function P(e,t){var n,r,i,o,a,u,s,l;if(1===t.nodeType){if(Ue.hasData(e)&&(o=Ue.access(e),a=Ue.set(t,o),l=o.events)){delete a.handle,a.events={};for(i in l)for(n=0,r=l[i].length;n1&&"string"==typeof h&&!xe.checkClone&&pt.test(h))return e.each(function(i){var o=e.eq(i);g&&(t[0]=h.call(this,i,o.html())),I(o,t,n,r)});if(p&&(i=T(t,e[0].ownerDocument,!1,e,r),o=i.firstChild,1===i.childNodes.length&&(i=o),o||r)){for(u=Ne.map(C(i,"script"),D),s=u.length;f=0&&(s+=Math.max(0,Math.ceil(e["offset"+t[0].toUpperCase()+t.slice(1)]-o-s-u-.5))),s}function U(e,t,n){var r=gt(e),i=H(e,t,r),o="border-box"===Ne.css(e,"boxSizing",!1,r),a=o;if(ht.test(i)){if(!n)return i;i="auto"}return a=a&&(xe.boxSizingReliable()||i===e.style[t]),("auto"===i||!parseFloat(i)&&"inline"===Ne.css(e,"display",!1,r))&&(i=e["offset"+t[0].toUpperCase()+t.slice(1)],a=!0),i=parseFloat(i)||0,i+W(e,t,n||(o?"border":"content"),a,r,i)+"px"}function K(e,t,n,r,i){return new K.prototype.init(e,t,n,r,i)}function G(){kt&&(ce.hidden===!1&&n.requestAnimationFrame?n.requestAnimationFrame(G):n.setTimeout(G,Ne.fx.interval),Ne.fx.tick())}function V(){return n.setTimeout(function(){Et=void 0}),Et=Date.now()}function X(e,t){var n,r=0,i={height:e};for(t=t?1:0;r<4;r+=2-t)n=Qe[r],i["margin"+n]=i["padding"+n]=e;return t&&(i.opacity=i.width=e),i}function Z(e,t,n){for(var r,i=(J.tweeners[t]||[]).concat(J.tweeners["*"]),o=0,a=i.length;o=0&&nE.cacheLength&&delete e[t.shift()],e[n+" "]=r}var t=[];return e}function r(e){return e[B]=!0,e}function i(e){var t=D.createElement("fieldset");try{return!!e(t)}catch(n){return!1}finally{t.parentNode&&t.parentNode.removeChild(t),t=null}}function o(e,t){for(var n=e.split("|"),r=n.length;r--;)E.attrHandle[n[r]]=t}function a(e,t){var n=t&&e,r=n&&1===e.nodeType&&1===t.nodeType&&e.sourceIndex-t.sourceIndex;if(r)return r;if(n)for(;n=n.nextSibling;)if(n===t)return-1;return e?1:-1}function u(e){return function(t){var n=t.nodeName.toLowerCase();return"input"===n&&t.type===e}}function s(e){return function(t){var n=t.nodeName.toLowerCase();return("input"===n||"button"===n)&&t.type===e}}function l(e){return function(t){return"form"in t?t.parentNode&&t.disabled===!1?"label"in t?"label"in t.parentNode?t.parentNode.disabled===e:t.disabled===e:t.isDisabled===e||t.isDisabled!==!e&&ke(t)===e:t.disabled===e:"label"in t&&t.disabled===e}}function c(e){return r(function(t){return t=+t,r(function(n,r){for(var i,o=e([],n.length,t),a=o.length;a--;)n[i=o[a]]&&(n[i]=!(r[i]=n[i]))})})}function f(e){return e&&"undefined"!=typeof e.getElementsByTagName&&e}function p(){}function d(e){for(var t=0,n=e.length,r="";t1?function(t,n,r){for(var i=e.length;i--;)if(!e[i](t,n,r))return!1;return!0}:e[0]}function v(e,n,r){for(var i=0,o=n.length;i-1&&(r[l]=!(a[l]=f))}}else b=m(b===a?b.splice(h,b.length):b),o?o(null,a,b,s):Y.apply(a,b)})}function b(e){for(var t,n,r,i=e.length,o=E.relative[e[0].type],a=o||E.relative[" "],u=o?1:0,s=h(function(e){return e===t},a,!0),l=h(function(e){return ee(t,e)>-1},a,!0),c=[function(e,n,r){var i=!o&&(r||n!==S)||((t=n).nodeType?s(e,n,r):l(e,n,r));return t=null,i}];u1&&g(c),u>1&&d(e.slice(0,u-1).concat({value:" "===e[u-2].type?"*":""})).replace(ue,"$1"),n,u0,o=e.length>0,a=function(r,a,u,s,l){var c,f,p,d=0,h="0",g=r&&[],v=[],y=S,b=r||o&&E.find.TAG("*",l),_=z+=null==y?1:Math.random()||.1,x=b.length;for(l&&(S=a===D||a||l);h!==x&&null!=(c=b[h]);h++){if(o&&c){for(f=0,a||c.ownerDocument===D||(M(c),u=!P);p=e[f++];)if(p(c,a||D,u)){s.push(c);break}l&&(z=_)}i&&((c=!p&&c)&&d--,r&&g.push(c))}if(d+=h,i&&h!==d){for(f=0;p=n[f++];)p(g,v,a,u);if(r){if(d>0)for(;h--;)g[h]||v[h]||(v[h]=Z.call(s));v=m(v)}Y.apply(s,v),l&&!r&&v.length>0&&d+n.length>1&&t.uniqueSort(s)}return l&&(z=_,S=y),g};return i?r(a):a}var x,w,E,k,C,N,T,A,S,j,O,M,D,R,P,L,I,q,H,B="sizzle"+1*new Date,$=e.document,z=0,F=0,W=n(),U=n(),K=n(),G=function(e,t){return e===t&&(O=!0),0},V={}.hasOwnProperty,X=[],Z=X.pop,Q=X.push,Y=X.push,J=X.slice,ee=function(e,t){for(var n=0,r=e.length;n+~]|"+ne+")"+ne+"*"),ce=new RegExp("="+ne+"*([^\\]'\"]*?)"+ne+"*\\]","g"),fe=new RegExp(oe),pe=new RegExp("^"+re+"$"),de={ID:new RegExp("^#("+re+")"),CLASS:new RegExp("^\\.("+re+")"),TAG:new RegExp("^("+re+"|[*])"),ATTR:new RegExp("^"+ie),PSEUDO:new RegExp("^"+oe),CHILD:new RegExp("^:(only|first|last|nth|nth-last)-(child|of-type)(?:\\("+ne+"*(even|odd|(([+-]|)(\\d*)n|)"+ne+"*(?:([+-]|)"+ne+"*(\\d+)|))"+ne+"*\\)|)","i"),bool:new RegExp("^(?:"+te+")$","i"),needsContext:new RegExp("^"+ne+"*[>+~]|:(even|odd|eq|gt|lt|nth|first|last)(?:\\("+ne+"*((?:-\\d)?\\d*)"+ne+"*\\)|)(?=[^-]|$)","i")},he=/^(?:input|select|textarea|button)$/i,ge=/^h\d$/i,ve=/^[^{]+\{\s*\[native \w/,me=/^(?:#([\w-]+)|(\w+)|\.([\w-]+))$/,ye=/[+~]/,be=new RegExp("\\\\([\\da-f]{1,6}"+ne+"?|("+ne+")|.)","ig"),_e=function(e,t,n){var r="0x"+t-65536;return r!==r||n?t:r<0?String.fromCharCode(r+65536):String.fromCharCode(r>>10|55296,1023&r|56320)},xe=/([\0-\x1f\x7f]|^-?\d)|^-$|[^\0-\x1f\x7f-\uFFFF\w-]/g,we=function(e,t){return t?"\0"===e?"�":e.slice(0,-1)+"\\"+e.charCodeAt(e.length-1).toString(16)+" ":"\\"+e},Ee=function(){M()},ke=h(function(e){return e.disabled===!0&&("form"in e||"label"in e)},{dir:"parentNode",next:"legend"});try{Y.apply(X=J.call($.childNodes),$.childNodes),X[$.childNodes.length].nodeType}catch(Ce){Y={apply:X.length?function(e,t){Q.apply(e,J.call(t))}:function(e,t){for(var n=e.length,r=0;e[n++]=t[r++];);e.length=n-1}}}w=t.support={},C=t.isXML=function(e){var t=e&&(e.ownerDocument||e).documentElement;return!!t&&"HTML"!==t.nodeName},M=t.setDocument=function(e){var t,n,r=e?e.ownerDocument||e:$;return r!==D&&9===r.nodeType&&r.documentElement?(D=r,R=D.documentElement,P=!C(D),$!==D&&(n=D.defaultView)&&n.top!==n&&(n.addEventListener?n.addEventListener("unload",Ee,!1):n.attachEvent&&n.attachEvent("onunload",Ee)),w.attributes=i(function(e){return e.className="i",!e.getAttribute("className")}),w.getElementsByTagName=i(function(e){return e.appendChild(D.createComment("")),!e.getElementsByTagName("*").length}),w.getElementsByClassName=ve.test(D.getElementsByClassName),w.getById=i(function(e){return R.appendChild(e).id=B,!D.getElementsByName||!D.getElementsByName(B).length}),w.getById?(E.filter.ID=function(e){var t=e.replace(be,_e);return function(e){return e.getAttribute("id")===t}},E.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&P){var n=t.getElementById(e);return n?[n]:[]}}):(E.filter.ID=function(e){var t=e.replace(be,_e);return function(e){var n="undefined"!=typeof e.getAttributeNode&&e.getAttributeNode("id");return n&&n.value===t}},E.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&P){var n,r,i,o=t.getElementById(e);if(o){if(n=o.getAttributeNode("id"),n&&n.value===e)return[o];for(i=t.getElementsByName(e),r=0;o=i[r++];)if(n=o.getAttributeNode("id"),n&&n.value===e)return[o]}return[]}}),E.find.TAG=w.getElementsByTagName?function(e,t){return"undefined"!=typeof t.getElementsByTagName?t.getElementsByTagName(e):w.qsa?t.querySelectorAll(e):void 0}:function(e,t){var n,r=[],i=0,o=t.getElementsByTagName(e);if("*"===e){for(;n=o[i++];)1===n.nodeType&&r.push(n);return r}return o},E.find.CLASS=w.getElementsByClassName&&function(e,t){if("undefined"!=typeof t.getElementsByClassName&&P)return t.getElementsByClassName(e)},I=[],L=[],(w.qsa=ve.test(D.querySelectorAll))&&(i(function(e){R.appendChild(e).innerHTML="",e.querySelectorAll("[msallowcapture^='']").length&&L.push("[*^$]="+ne+"*(?:''|\"\")"),e.querySelectorAll("[selected]").length||L.push("\\["+ne+"*(?:value|"+te+")"),e.querySelectorAll("[id~="+B+"-]").length||L.push("~="),e.querySelectorAll(":checked").length||L.push(":checked"),e.querySelectorAll("a#"+B+"+*").length||L.push(".#.+[+~]")}),i(function(e){e.innerHTML="";var t=D.createElement("input");t.setAttribute("type","hidden"),e.appendChild(t).setAttribute("name","D"),e.querySelectorAll("[name=d]").length&&L.push("name"+ne+"*[*^$|!~]?="),2!==e.querySelectorAll(":enabled").length&&L.push(":enabled",":disabled"),R.appendChild(e).disabled=!0,2!==e.querySelectorAll(":disabled").length&&L.push(":enabled",":disabled"),e.querySelectorAll("*,:x"),L.push(",.*:")})),(w.matchesSelector=ve.test(q=R.matches||R.webkitMatchesSelector||R.mozMatchesSelector||R.oMatchesSelector||R.msMatchesSelector))&&i(function(e){w.disconnectedMatch=q.call(e,"*"),q.call(e,"[s!='']:x"),I.push("!=",oe)}),L=L.length&&new RegExp(L.join("|")),I=I.length&&new RegExp(I.join("|")),t=ve.test(R.compareDocumentPosition),H=t||ve.test(R.contains)?function(e,t){var n=9===e.nodeType?e.documentElement:e,r=t&&t.parentNode;return e===r||!(!r||1!==r.nodeType||!(n.contains?n.contains(r):e.compareDocumentPosition&&16&e.compareDocumentPosition(r)))}:function(e,t){if(t)for(;t=t.parentNode;)if(t===e)return!0;return!1},G=t?function(e,t){if(e===t)return O=!0,0;var n=!e.compareDocumentPosition-!t.compareDocumentPosition;return n?n:(n=(e.ownerDocument||e)===(t.ownerDocument||t)?e.compareDocumentPosition(t):1,1&n||!w.sortDetached&&t.compareDocumentPosition(e)===n?e===D||e.ownerDocument===$&&H($,e)?-1:t===D||t.ownerDocument===$&&H($,t)?1:j?ee(j,e)-ee(j,t):0:4&n?-1:1)}:function(e,t){if(e===t)return O=!0,0;var n,r=0,i=e.parentNode,o=t.parentNode,u=[e],s=[t];if(!i||!o)return e===D?-1:t===D?1:i?-1:o?1:j?ee(j,e)-ee(j,t):0;if(i===o)return a(e,t);for(n=e;n=n.parentNode;)u.unshift(n);for(n=t;n=n.parentNode;)s.unshift(n);for(;u[r]===s[r];)r++;return r?a(u[r],s[r]):u[r]===$?-1:s[r]===$?1:0},D):D},t.matches=function(e,n){return t(e,null,null,n)},t.matchesSelector=function(e,n){if((e.ownerDocument||e)!==D&&M(e),n=n.replace(ce,"='$1']"),w.matchesSelector&&P&&!K[n+" "]&&(!I||!I.test(n))&&(!L||!L.test(n)))try{var r=q.call(e,n);if(r||w.disconnectedMatch||e.document&&11!==e.document.nodeType)return r}catch(i){}return t(n,D,null,[e]).length>0},t.contains=function(e,t){return(e.ownerDocument||e)!==D&&M(e),H(e,t)},t.attr=function(e,t){(e.ownerDocument||e)!==D&&M(e);var n=E.attrHandle[t.toLowerCase()],r=n&&V.call(E.attrHandle,t.toLowerCase())?n(e,t,!P):void 0;return void 0!==r?r:w.attributes||!P?e.getAttribute(t):(r=e.getAttributeNode(t))&&r.specified?r.value:null},t.escape=function(e){return(e+"").replace(xe,we)},t.error=function(e){throw new Error("Syntax error, unrecognized expression: "+e)},t.uniqueSort=function(e){var t,n=[],r=0,i=0;if(O=!w.detectDuplicates,j=!w.sortStable&&e.slice(0),e.sort(G),O){for(;t=e[i++];)t===e[i]&&(r=n.push(i));for(;r--;)e.splice(n[r],1)}return j=null,e},k=t.getText=function(e){var t,n="",r=0,i=e.nodeType;if(i){if(1===i||9===i||11===i){if("string"==typeof e.textContent)return e.textContent;for(e=e.firstChild;e;e=e.nextSibling)n+=k(e)}else if(3===i||4===i)return e.nodeValue}else for(;t=e[r++];)n+=k(t);return n},E=t.selectors={cacheLength:50,createPseudo:r,match:de,attrHandle:{},find:{},relative:{">":{dir:"parentNode",first:!0}," ":{dir:"parentNode"},"+":{dir:"previousSibling",first:!0},"~":{dir:"previousSibling"}},preFilter:{ATTR:function(e){return e[1]=e[1].replace(be,_e),e[3]=(e[3]||e[4]||e[5]||"").replace(be,_e),"~="===e[2]&&(e[3]=" "+e[3]+" "),e.slice(0,4)},CHILD:function(e){return e[1]=e[1].toLowerCase(),"nth"===e[1].slice(0,3)?(e[3]||t.error(e[0]),e[4]=+(e[4]?e[5]+(e[6]||1):2*("even"===e[3]||"odd"===e[3])),e[5]=+(e[7]+e[8]||"odd"===e[3])):e[3]&&t.error(e[0]),e},PSEUDO:function(e){var t,n=!e[6]&&e[2];return de.CHILD.test(e[0])?null:(e[3]?e[2]=e[4]||e[5]||"":n&&fe.test(n)&&(t=N(n,!0))&&(t=n.indexOf(")",n.length-t)-n.length)&&(e[0]=e[0].slice(0,t),e[2]=n.slice(0,t)),e.slice(0,3))}},filter:{TAG:function(e){var t=e.replace(be,_e).toLowerCase();return"*"===e?function(){return!0}:function(e){return e.nodeName&&e.nodeName.toLowerCase()===t}},CLASS:function(e){var t=W[e+" "];return t||(t=new RegExp("(^|"+ne+")"+e+"("+ne+"|$)"))&&W(e,function(e){return t.test("string"==typeof e.className&&e.className||"undefined"!=typeof e.getAttribute&&e.getAttribute("class")||"")})},ATTR:function(e,n,r){return function(i){var o=t.attr(i,e);return null==o?"!="===n:!n||(o+="","="===n?o===r:"!="===n?o!==r:"^="===n?r&&0===o.indexOf(r):"*="===n?r&&o.indexOf(r)>-1:"$="===n?r&&o.slice(-r.length)===r:"~="===n?(" "+o.replace(ae," ")+" ").indexOf(r)>-1:"|="===n&&(o===r||o.slice(0,r.length+1)===r+"-"))}},CHILD:function(e,t,n,r,i){var o="nth"!==e.slice(0,3),a="last"!==e.slice(-4),u="of-type"===t;return 1===r&&0===i?function(e){return!!e.parentNode}:function(t,n,s){var l,c,f,p,d,h,g=o!==a?"nextSibling":"previousSibling",v=t.parentNode,m=u&&t.nodeName.toLowerCase(),y=!s&&!u,b=!1;if(v){if(o){for(;g;){for(p=t;p=p[g];)if(u?p.nodeName.toLowerCase()===m:1===p.nodeType)return!1;h=g="only"===e&&!h&&"nextSibling"}return!0}if(h=[a?v.firstChild:v.lastChild],a&&y){for(p=v,f=p[B]||(p[B]={}),c=f[p.uniqueID]||(f[p.uniqueID]={}),l=c[e]||[],d=l[0]===z&&l[1],b=d&&l[2],p=d&&v.childNodes[d];p=++d&&p&&p[g]||(b=d=0)||h.pop();)if(1===p.nodeType&&++b&&p===t){c[e]=[z,d,b];break}}else if(y&&(p=t,f=p[B]||(p[B]={}),c=f[p.uniqueID]||(f[p.uniqueID]={}),l=c[e]||[],d=l[0]===z&&l[1],b=d),b===!1)for(;(p=++d&&p&&p[g]||(b=d=0)||h.pop())&&((u?p.nodeName.toLowerCase()!==m:1!==p.nodeType)||!++b||(y&&(f=p[B]||(p[B]={}),c=f[p.uniqueID]||(f[p.uniqueID]={}),c[e]=[z,b]),p!==t)););return b-=i,b===r||b%r===0&&b/r>=0}}},PSEUDO:function(e,n){var i,o=E.pseudos[e]||E.setFilters[e.toLowerCase()]||t.error("unsupported pseudo: "+e);return o[B]?o(n):o.length>1?(i=[e,e,"",n],E.setFilters.hasOwnProperty(e.toLowerCase())?r(function(e,t){for(var r,i=o(e,n),a=i.length;a--;)r=ee(e,i[a]),e[r]=!(t[r]=i[a])}):function(e){return o(e,0,i)}):o}},pseudos:{not:r(function(e){var t=[],n=[],i=T(e.replace(ue,"$1"));return i[B]?r(function(e,t,n,r){for(var o,a=i(e,null,r,[]),u=e.length;u--;)(o=a[u])&&(e[u]=!(t[u]=o))}):function(e,r,o){return t[0]=e,i(t,null,o,n), +t[0]=null,!n.pop()}}),has:r(function(e){return function(n){return t(e,n).length>0}}),contains:r(function(e){return e=e.replace(be,_e),function(t){return(t.textContent||t.innerText||k(t)).indexOf(e)>-1}}),lang:r(function(e){return pe.test(e||"")||t.error("unsupported lang: "+e),e=e.replace(be,_e).toLowerCase(),function(t){var n;do if(n=P?t.lang:t.getAttribute("xml:lang")||t.getAttribute("lang"))return n=n.toLowerCase(),n===e||0===n.indexOf(e+"-");while((t=t.parentNode)&&1===t.nodeType);return!1}}),target:function(t){var n=e.location&&e.location.hash;return n&&n.slice(1)===t.id},root:function(e){return e===R},focus:function(e){return e===D.activeElement&&(!D.hasFocus||D.hasFocus())&&!!(e.type||e.href||~e.tabIndex)},enabled:l(!1),disabled:l(!0),checked:function(e){var t=e.nodeName.toLowerCase();return"input"===t&&!!e.checked||"option"===t&&!!e.selected},selected:function(e){return e.parentNode&&e.parentNode.selectedIndex,e.selected===!0},empty:function(e){for(e=e.firstChild;e;e=e.nextSibling)if(e.nodeType<6)return!1;return!0},parent:function(e){return!E.pseudos.empty(e)},header:function(e){return ge.test(e.nodeName)},input:function(e){return he.test(e.nodeName)},button:function(e){var t=e.nodeName.toLowerCase();return"input"===t&&"button"===e.type||"button"===t},text:function(e){var t;return"input"===e.nodeName.toLowerCase()&&"text"===e.type&&(null==(t=e.getAttribute("type"))||"text"===t.toLowerCase())},first:c(function(){return[0]}),last:c(function(e,t){return[t-1]}),eq:c(function(e,t,n){return[n<0?n+t:n]}),even:c(function(e,t){for(var n=0;n=0;)e.push(r);return e}),gt:c(function(e,t,n){for(var r=n<0?n+t:n;++r2&&"ID"===(a=o[0]).type&&9===t.nodeType&&P&&E.relative[o[1].type]){if(t=(E.find.ID(a.matches[0].replace(be,_e),t)||[])[0],!t)return n;l&&(t=t.parentNode),e=e.slice(o.shift().value.length)}for(i=de.needsContext.test(e)?0:o.length;i--&&(a=o[i],!E.relative[u=a.type]);)if((s=E.find[u])&&(r=s(a.matches[0].replace(be,_e),ye.test(o[0].type)&&f(t.parentNode)||t))){if(o.splice(i,1),e=r.length&&d(o),!e)return Y.apply(n,r),n;break}}return(l||T(e,c))(r,t,!P,n,!t||ye.test(e)&&f(t.parentNode)||t),n},w.sortStable=B.split("").sort(G).join("")===B,w.detectDuplicates=!!O,M(),w.sortDetached=i(function(e){return 1&e.compareDocumentPosition(D.createElement("fieldset"))}),i(function(e){return e.innerHTML="","#"===e.firstChild.getAttribute("href")})||o("type|href|height|width",function(e,t,n){if(!n)return e.getAttribute(t,"type"===t.toLowerCase()?1:2)}),w.attributes&&i(function(e){return e.innerHTML="",e.firstChild.setAttribute("value",""),""===e.firstChild.getAttribute("value")})||o("value",function(e,t,n){if(!n&&"input"===e.nodeName.toLowerCase())return e.defaultValue}),i(function(e){return null==e.getAttribute("disabled")})||o(te,function(e,t,n){var r;if(!n)return e[t]===!0?t.toLowerCase():(r=e.getAttributeNode(t))&&r.specified?r.value:null}),t}(n);Ne.find=Ae,Ne.expr=Ae.selectors,Ne.expr[":"]=Ne.expr.pseudos,Ne.uniqueSort=Ne.unique=Ae.uniqueSort,Ne.text=Ae.getText,Ne.isXMLDoc=Ae.isXML,Ne.contains=Ae.contains,Ne.escapeSelector=Ae.escape;var Se=function(e,t,n){for(var r=[],i=void 0!==n;(e=e[t])&&9!==e.nodeType;)if(1===e.nodeType){if(i&&Ne(e).is(n))break;r.push(e)}return r},je=function(e,t){for(var n=[];e;e=e.nextSibling)1===e.nodeType&&e!==t&&n.push(e);return n},Oe=Ne.expr.match.needsContext,Me=/^<([a-z][^\/\0>:\x20\t\r\n\f]*)[\x20\t\r\n\f]*\/?>(?:<\/\1>|)$/i;Ne.filter=function(e,t,n){var r=t[0];return n&&(e=":not("+e+")"),1===t.length&&1===r.nodeType?Ne.find.matchesSelector(r,e)?[r]:[]:Ne.find.matches(e,Ne.grep(t,function(e){return 1===e.nodeType}))},Ne.fn.extend({find:function(e){var t,n,r=this.length,i=this;if("string"!=typeof e)return this.pushStack(Ne(e).filter(function(){for(t=0;t1?Ne.uniqueSort(n):n},filter:function(e){return this.pushStack(c(this,e||[],!1))},not:function(e){return this.pushStack(c(this,e||[],!0))},is:function(e){return!!c(this,"string"==typeof e&&Oe.test(e)?Ne(e):e||[],!1).length}});var De,Re=/^(?:\s*(<[\w\W]+>)[^>]*|#([\w-]+))$/,Pe=Ne.fn.init=function(e,t,n){var r,i;if(!e)return this;if(n=n||De,"string"==typeof e){if(r="<"===e[0]&&">"===e[e.length-1]&&e.length>=3?[null,e,null]:Re.exec(e),!r||!r[1]&&t)return!t||t.jquery?(t||n).find(e):this.constructor(t).find(e);if(r[1]){if(t=t instanceof Ne?t[0]:t,Ne.merge(this,Ne.parseHTML(r[1],t&&t.nodeType?t.ownerDocument||t:ce,!0)),Me.test(r[1])&&Ne.isPlainObject(t))for(r in t)we(this[r])?this[r](t[r]):this.attr(r,t[r]);return this}return i=ce.getElementById(r[2]),i&&(this[0]=i,this.length=1),this}return e.nodeType?(this[0]=e,this.length=1,this):we(e)?void 0!==n.ready?n.ready(e):e(Ne):Ne.makeArray(e,this)};Pe.prototype=Ne.fn,De=Ne(ce);var Le=/^(?:parents|prev(?:Until|All))/,Ie={children:!0,contents:!0,next:!0,prev:!0};Ne.fn.extend({has:function(e){var t=Ne(e,this),n=t.length;return this.filter(function(){for(var e=0;e-1:1===n.nodeType&&Ne.find.matchesSelector(n,e))){o.push(n);break}return this.pushStack(o.length>1?Ne.uniqueSort(o):o)},index:function(e){return e?"string"==typeof e?ge.call(Ne(e),this[0]):ge.call(this,e.jquery?e[0]:e):this[0]&&this[0].parentNode?this.first().prevAll().length:-1},add:function(e,t){return this.pushStack(Ne.uniqueSort(Ne.merge(this.get(),Ne(e,t))))},addBack:function(e){return this.add(null==e?this.prevObject:this.prevObject.filter(e))}}),Ne.each({parent:function(e){var t=e.parentNode;return t&&11!==t.nodeType?t:null},parents:function(e){return Se(e,"parentNode")},parentsUntil:function(e,t,n){return Se(e,"parentNode",n)},next:function(e){return f(e,"nextSibling")},prev:function(e){return f(e,"previousSibling")},nextAll:function(e){return Se(e,"nextSibling")},prevAll:function(e){return Se(e,"previousSibling")},nextUntil:function(e,t,n){return Se(e,"nextSibling",n)},prevUntil:function(e,t,n){return Se(e,"previousSibling",n)},siblings:function(e){return je((e.parentNode||{}).firstChild,e)},children:function(e){return je(e.firstChild)},contents:function(e){return l(e,"iframe")?e.contentDocument:(l(e,"template")&&(e=e.content||e),Ne.merge([],e.childNodes))}},function(e,t){Ne.fn[e]=function(n,r){var i=Ne.map(this,t,n);return"Until"!==e.slice(-5)&&(r=n),r&&"string"==typeof r&&(i=Ne.filter(r,i)),this.length>1&&(Ie[e]||Ne.uniqueSort(i),Le.test(e)&&i.reverse()),this.pushStack(i)}});var qe=/[^\x20\t\r\n\f]+/g;Ne.Callbacks=function(e){e="string"==typeof e?p(e):Ne.extend({},e);var t,n,r,i,o=[],a=[],s=-1,l=function(){for(i=i||e.once,r=t=!0;a.length;s=-1)for(n=a.shift();++s-1;)o.splice(n,1),n<=s&&s--}),this},has:function(e){return e?Ne.inArray(e,o)>-1:o.length>0},empty:function(){return o&&(o=[]),this},disable:function(){return i=a=[],o=n="",this},disabled:function(){return!o},lock:function(){return i=a=[],n||t||(o=n=""),this},locked:function(){return!!i},fireWith:function(e,n){return i||(n=n||[],n=[e,n.slice?n.slice():n],a.push(n),t||l()),this},fire:function(){return c.fireWith(this,arguments),this},fired:function(){return!!r}};return c},Ne.extend({Deferred:function(e){var t=[["notify","progress",Ne.Callbacks("memory"),Ne.Callbacks("memory"),2],["resolve","done",Ne.Callbacks("once memory"),Ne.Callbacks("once memory"),0,"resolved"],["reject","fail",Ne.Callbacks("once memory"),Ne.Callbacks("once memory"),1,"rejected"]],r="pending",i={state:function(){return r},always:function(){return o.done(arguments).fail(arguments),this},"catch":function(e){return i.then(null,e)},pipe:function(){var e=arguments;return Ne.Deferred(function(n){Ne.each(t,function(t,r){var i=we(e[r[4]])&&e[r[4]];o[r[1]](function(){var e=i&&i.apply(this,arguments);e&&we(e.promise)?e.promise().progress(n.notify).done(n.resolve).fail(n.reject):n[r[0]+"With"](this,i?[e]:arguments)})}),e=null}).promise()},then:function(e,r,i){function o(e,t,r,i){return function(){var u=this,s=arguments,l=function(){var n,l;if(!(e=a&&(r!==h&&(u=void 0,s=[n]),t.rejectWith(u,s))}};e?c():(Ne.Deferred.getStackHook&&(c.stackTrace=Ne.Deferred.getStackHook()),n.setTimeout(c))}}var a=0;return Ne.Deferred(function(n){t[0][3].add(o(0,n,we(i)?i:d,n.notifyWith)),t[1][3].add(o(0,n,we(e)?e:d)),t[2][3].add(o(0,n,we(r)?r:h))}).promise()},promise:function(e){return null!=e?Ne.extend(e,i):i}},o={};return Ne.each(t,function(e,n){var a=n[2],u=n[5];i[n[1]]=a.add,u&&a.add(function(){r=u},t[3-e][2].disable,t[3-e][3].disable,t[0][2].lock,t[0][3].lock),a.add(n[3].fire),o[n[0]]=function(){return o[n[0]+"With"](this===o?void 0:this,arguments),this},o[n[0]+"With"]=a.fireWith}),i.promise(o),e&&e.call(o,o),o},when:function(e){var t=arguments.length,n=t,r=Array(n),i=pe.call(arguments),o=Ne.Deferred(),a=function(e){return function(n){r[e]=this,i[e]=arguments.length>1?pe.call(arguments):n,--t||o.resolveWith(r,i)}};if(t<=1&&(g(e,o.done(a(n)).resolve,o.reject,!t),"pending"===o.state()||we(i[n]&&i[n].then)))return o.then();for(;n--;)g(i[n],a(n),o.reject);return o.promise()}});var He=/^(Eval|Internal|Range|Reference|Syntax|Type|URI)Error$/;Ne.Deferred.exceptionHook=function(e,t){n.console&&n.console.warn&&e&&He.test(e.name)&&n.console.warn("jQuery.Deferred exception: "+e.message,e.stack,t)},Ne.readyException=function(e){n.setTimeout(function(){throw e})};var Be=Ne.Deferred();Ne.fn.ready=function(e){return Be.then(e)["catch"](function(e){Ne.readyException(e)}),this},Ne.extend({isReady:!1,readyWait:1,ready:function(e){(e===!0?--Ne.readyWait:Ne.isReady)||(Ne.isReady=!0,e!==!0&&--Ne.readyWait>0||Be.resolveWith(ce,[Ne]))}}),Ne.ready.then=Be.then,"complete"===ce.readyState||"loading"!==ce.readyState&&!ce.documentElement.doScroll?n.setTimeout(Ne.ready):(ce.addEventListener("DOMContentLoaded",v),n.addEventListener("load",v));var $e=function(e,t,n,r,i,o,a){var s=0,l=e.length,c=null==n;if("object"===u(n)){i=!0;for(s in n)$e(e,t,s,n[s],!0,o,a)}else if(void 0!==r&&(i=!0,we(r)||(a=!0),c&&(a?(t.call(e,r),t=null):(c=t,t=function(e,t,n){return c.call(Ne(e),n)})),t))for(;s1,null,!0)},removeData:function(e){return this.each(function(){Ke.remove(this,e)})}}),Ne.extend({queue:function(e,t,n){var r;if(e)return t=(t||"fx")+"queue",r=Ue.get(e,t),n&&(!r||Array.isArray(n)?r=Ue.access(e,t,Ne.makeArray(n)):r.push(n)),r||[]},dequeue:function(e,t){t=t||"fx";var n=Ne.queue(e,t),r=n.length,i=n.shift(),o=Ne._queueHooks(e,t),a=function(){Ne.dequeue(e,t)};"inprogress"===i&&(i=n.shift(),r--),i&&("fx"===t&&n.unshift("inprogress"),delete o.stop,i.call(e,a,o)),!r&&o&&o.empty.fire()},_queueHooks:function(e,t){var n=t+"queueHooks";return Ue.get(e,n)||Ue.access(e,n,{empty:Ne.Callbacks("once memory").add(function(){Ue.remove(e,[t+"queue",n])})})}}),Ne.fn.extend({queue:function(e,t){var n=2;return"string"!=typeof e&&(t=e,e="fx",n--),arguments.length\x20\t\r\n\f]+)/i,rt=/^$|^module$|\/(?:java|ecma)script/i,it={option:[1,""],thead:[1,"","
"],col:[2,"","
"],tr:[2,"","
"],td:[3,"","
"],_default:[0,"",""]};it.optgroup=it.option,it.tbody=it.tfoot=it.colgroup=it.caption=it.thead,it.th=it.td;var ot=/<|&#?\w+;/;!function(){var e=ce.createDocumentFragment(),t=e.appendChild(ce.createElement("div")),n=ce.createElement("input");n.setAttribute("type","radio"),n.setAttribute("checked","checked"),n.setAttribute("name","t"),t.appendChild(n),xe.checkClone=t.cloneNode(!0).cloneNode(!0).lastChild.checked,t.innerHTML="",xe.noCloneChecked=!!t.cloneNode(!0).lastChild.defaultValue}();var at=ce.documentElement,ut=/^key/,st=/^(?:mouse|pointer|contextmenu|drag|drop)|click/,lt=/^([^.]*)(?:\.(.+)|)/;Ne.event={global:{},add:function(e,t,n,r,i){var o,a,u,s,l,c,f,p,d,h,g,v=Ue.get(e);if(v)for(n.handler&&(o=n,n=o.handler,i=o.selector),i&&Ne.find.matchesSelector(at,i),n.guid||(n.guid=Ne.guid++),(s=v.events)||(s=v.events={}),(a=v.handle)||(a=v.handle=function(t){return"undefined"!=typeof Ne&&Ne.event.triggered!==t.type?Ne.event.dispatch.apply(e,arguments):void 0}),t=(t||"").match(qe)||[""],l=t.length;l--;)u=lt.exec(t[l])||[],d=g=u[1],h=(u[2]||"").split(".").sort(),d&&(f=Ne.event.special[d]||{},d=(i?f.delegateType:f.bindType)||d,f=Ne.event.special[d]||{},c=Ne.extend({type:d,origType:g,data:r,handler:n,guid:n.guid,selector:i,needsContext:i&&Ne.expr.match.needsContext.test(i),namespace:h.join(".")},o),(p=s[d])||(p=s[d]=[],p.delegateCount=0,f.setup&&f.setup.call(e,r,h,a)!==!1||e.addEventListener&&e.addEventListener(d,a)),f.add&&(f.add.call(e,c),c.handler.guid||(c.handler.guid=n.guid)),i?p.splice(p.delegateCount++,0,c):p.push(c),Ne.event.global[d]=!0)},remove:function(e,t,n,r,i){var o,a,u,s,l,c,f,p,d,h,g,v=Ue.hasData(e)&&Ue.get(e);if(v&&(s=v.events)){for(t=(t||"").match(qe)||[""],l=t.length;l--;)if(u=lt.exec(t[l])||[],d=g=u[1],h=(u[2]||"").split(".").sort(),d){for(f=Ne.event.special[d]||{},d=(r?f.delegateType:f.bindType)||d,p=s[d]||[],u=u[2]&&new RegExp("(^|\\.)"+h.join("\\.(?:.*\\.|)")+"(\\.|$)"),a=o=p.length;o--;)c=p[o],!i&&g!==c.origType||n&&n.guid!==c.guid||u&&!u.test(c.namespace)||r&&r!==c.selector&&("**"!==r||!c.selector)||(p.splice(o,1),c.selector&&p.delegateCount--,f.remove&&f.remove.call(e,c));a&&!p.length&&(f.teardown&&f.teardown.call(e,h,v.handle)!==!1||Ne.removeEvent(e,d,v.handle),delete s[d])}else for(d in s)Ne.event.remove(e,d+t[l],n,r,!0);Ne.isEmptyObject(s)&&Ue.remove(e,"handle events")}},dispatch:function(e){var t,n,r,i,o,a,u=Ne.event.fix(e),s=new Array(arguments.length),l=(Ue.get(this,"events")||{})[u.type]||[],c=Ne.event.special[u.type]||{};for(s[0]=u,t=1;t=1))for(;l!==this;l=l.parentNode||this)if(1===l.nodeType&&("click"!==e.type||l.disabled!==!0)){for(o=[],a={},n=0;n-1:Ne.find(i,this,null,[l]).length),a[i]&&o.push(r);o.length&&u.push({elem:l,handlers:o})}return l=this,s\x20\t\r\n\f]*)[^>]*)\/>/gi,ft=/\s*$/g;Ne.extend({htmlPrefilter:function(e){return e.replace(ct,"<$1>")},clone:function(e,t,n){var r,i,o,a,u=e.cloneNode(!0),s=Ne.contains(e.ownerDocument,e);if(!(xe.noCloneChecked||1!==e.nodeType&&11!==e.nodeType||Ne.isXMLDoc(e)))for(a=C(u),o=C(e),r=0,i=o.length;r0&&N(a,!s&&C(e,"script")),u},cleanData:function(e){for(var t,n,r,i=Ne.event.special,o=0;void 0!==(n=e[o]);o++)if(We(n)){if(t=n[Ue.expando]){if(t.events)for(r in t.events)i[r]?Ne.event.remove(n,r):Ne.removeEvent(n,r,t.handle);n[Ue.expando]=void 0}n[Ke.expando]&&(n[Ke.expando]=void 0)}}}),Ne.fn.extend({detach:function(e){return q(this,e,!0)},remove:function(e){return q(this,e)},text:function(e){return $e(this,function(e){return void 0===e?Ne.text(this):this.empty().each(function(){1!==this.nodeType&&11!==this.nodeType&&9!==this.nodeType||(this.textContent=e)})},null,e,arguments.length)},append:function(){return I(this,arguments,function(e){if(1===this.nodeType||11===this.nodeType||9===this.nodeType){var t=M(this,e);t.appendChild(e)}})},prepend:function(){return I(this,arguments,function(e){if(1===this.nodeType||11===this.nodeType||9===this.nodeType){var t=M(this,e);t.insertBefore(e,t.firstChild)}})},before:function(){return I(this,arguments,function(e){this.parentNode&&this.parentNode.insertBefore(e,this)})},after:function(){return I(this,arguments,function(e){this.parentNode&&this.parentNode.insertBefore(e,this.nextSibling)})},empty:function(){for(var e,t=0;null!=(e=this[t]);t++)1===e.nodeType&&(Ne.cleanData(C(e,!1)),e.textContent="");return this},clone:function(e,t){return e=null!=e&&e,t=null==t?e:t,this.map(function(){return Ne.clone(this,e,t)})},html:function(e){return $e(this,function(e){var t=this[0]||{},n=0,r=this.length;if(void 0===e&&1===t.nodeType)return t.innerHTML;if("string"==typeof e&&!ft.test(e)&&!it[(nt.exec(e)||["",""])[1].toLowerCase()]){e=Ne.htmlPrefilter(e);try{for(;n1)}}),Ne.Tween=K,K.prototype={constructor:K,init:function(e,t,n,r,i,o){this.elem=e,this.prop=n,this.easing=i||Ne.easing._default,this.options=t,this.start=this.now=this.cur(),this.end=r,this.unit=o||(Ne.cssNumber[n]?"":"px")},cur:function(){var e=K.propHooks[this.prop];return e&&e.get?e.get(this):K.propHooks._default.get(this)},run:function(e){var t,n=K.propHooks[this.prop];return this.options.duration?this.pos=t=Ne.easing[this.easing](e,this.options.duration*e,0,1,this.options.duration):this.pos=t=e,this.now=(this.end-this.start)*t+this.start,this.options.step&&this.options.step.call(this.elem,this.now,this),n&&n.set?n.set(this):K.propHooks._default.set(this),this}},K.prototype.init.prototype=K.prototype,K.propHooks={_default:{get:function(e){var t;return 1!==e.elem.nodeType||null!=e.elem[e.prop]&&null==e.elem.style[e.prop]?e.elem[e.prop]:(t=Ne.css(e.elem,e.prop,""),t&&"auto"!==t?t:0)},set:function(e){Ne.fx.step[e.prop]?Ne.fx.step[e.prop](e):1!==e.elem.nodeType||null==e.elem.style[Ne.cssProps[e.prop]]&&!Ne.cssHooks[e.prop]?e.elem[e.prop]=e.now:Ne.style(e.elem,e.prop,e.now+e.unit)}}},K.propHooks.scrollTop=K.propHooks.scrollLeft={set:function(e){e.elem.nodeType&&e.elem.parentNode&&(e.elem[e.prop]=e.now)}},Ne.easing={linear:function(e){return e},swing:function(e){return.5-Math.cos(e*Math.PI)/2},_default:"swing"},Ne.fx=K.prototype.init,Ne.fx.step={};var Et,kt,Ct=/^(?:toggle|show|hide)$/,Nt=/queueHooks$/;Ne.Animation=Ne.extend(J,{tweeners:{"*":[function(e,t){var n=this.createTween(e,t);return w(n.elem,e,Ze.exec(t),n),n}]},tweener:function(e,t){we(e)?(t=e,e=["*"]):e=e.match(qe);for(var n,r=0,i=e.length;r1)},removeAttr:function(e){return this.each(function(){Ne.removeAttr(this,e)})}}),Ne.extend({attr:function(e,t,n){var r,i,o=e.nodeType;if(3!==o&&8!==o&&2!==o)return"undefined"==typeof e.getAttribute?Ne.prop(e,t,n):(1===o&&Ne.isXMLDoc(e)||(i=Ne.attrHooks[t.toLowerCase()]||(Ne.expr.match.bool.test(t)?Tt:void 0)),void 0!==n?null===n?void Ne.removeAttr(e,t):i&&"set"in i&&void 0!==(r=i.set(e,n,t))?r:(e.setAttribute(t,n+""),n):i&&"get"in i&&null!==(r=i.get(e,t))?r:(r=Ne.find.attr(e,t),null==r?void 0:r))},attrHooks:{type:{set:function(e,t){if(!xe.radioValue&&"radio"===t&&l(e,"input")){var n=e.value;return e.setAttribute("type",t),n&&(e.value=n),t}}}},removeAttr:function(e,t){var n,r=0,i=t&&t.match(qe);if(i&&1===e.nodeType)for(;n=i[r++];)e.removeAttribute(n)}}),Tt={set:function(e,t,n){return t===!1?Ne.removeAttr(e,n):e.setAttribute(n,n),n}},Ne.each(Ne.expr.match.bool.source.match(/\w+/g),function(e,t){var n=At[t]||Ne.find.attr;At[t]=function(e,t,r){var i,o,a=t.toLowerCase();return r||(o=At[a],At[a]=i,i=null!=n(e,t,r)?a:null,At[a]=o),i}});var St=/^(?:input|select|textarea|button)$/i,jt=/^(?:a|area)$/i;Ne.fn.extend({prop:function(e,t){return $e(this,Ne.prop,e,t,arguments.length>1)},removeProp:function(e){return this.each(function(){delete this[Ne.propFix[e]||e]})}}),Ne.extend({prop:function(e,t,n){var r,i,o=e.nodeType;if(3!==o&&8!==o&&2!==o)return 1===o&&Ne.isXMLDoc(e)||(t=Ne.propFix[t]||t,i=Ne.propHooks[t]),void 0!==n?i&&"set"in i&&void 0!==(r=i.set(e,n,t))?r:e[t]=n:i&&"get"in i&&null!==(r=i.get(e,t))?r:e[t]},propHooks:{tabIndex:{get:function(e){var t=Ne.find.attr(e,"tabindex");return t?parseInt(t,10):St.test(e.nodeName)||jt.test(e.nodeName)&&e.href?0:-1}}},propFix:{"for":"htmlFor","class":"className"}}),xe.optSelected||(Ne.propHooks.selected={get:function(e){var t=e.parentNode;return t&&t.parentNode&&t.parentNode.selectedIndex,null},set:function(e){var t=e.parentNode;t&&(t.selectedIndex,t.parentNode&&t.parentNode.selectedIndex)}}),Ne.each(["tabIndex","readOnly","maxLength","cellSpacing","cellPadding","rowSpan","colSpan","useMap","frameBorder","contentEditable"],function(){Ne.propFix[this.toLowerCase()]=this}),Ne.fn.extend({addClass:function(e){var t,n,r,i,o,a,u,s=0;if(we(e))return this.each(function(t){Ne(this).addClass(e.call(this,t,te(this)))});if(t=ne(e),t.length)for(;n=this[s++];)if(i=te(n),r=1===n.nodeType&&" "+ee(i)+" "){for(a=0;o=t[a++];)r.indexOf(" "+o+" ")<0&&(r+=o+" ");u=ee(r),i!==u&&n.setAttribute("class",u)}return this},removeClass:function(e){var t,n,r,i,o,a,u,s=0;if(we(e))return this.each(function(t){Ne(this).removeClass(e.call(this,t,te(this)))});if(!arguments.length)return this.attr("class","");if(t=ne(e),t.length)for(;n=this[s++];)if(i=te(n),r=1===n.nodeType&&" "+ee(i)+" "){for(a=0;o=t[a++];)for(;r.indexOf(" "+o+" ")>-1;)r=r.replace(" "+o+" "," ");u=ee(r),i!==u&&n.setAttribute("class",u)}return this},toggleClass:function(e,t){var n=typeof e,r="string"===n||Array.isArray(e);return"boolean"==typeof t&&r?t?this.addClass(e):this.removeClass(e):we(e)?this.each(function(n){Ne(this).toggleClass(e.call(this,n,te(this),t),t)}):this.each(function(){var t,i,o,a;if(r)for(i=0,o=Ne(this),a=ne(e);t=a[i++];)o.hasClass(t)?o.removeClass(t):o.addClass(t);else void 0!==e&&"boolean"!==n||(t=te(this),t&&Ue.set(this,"__className__",t),this.setAttribute&&this.setAttribute("class",t||e===!1?"":Ue.get(this,"__className__")||""))})},hasClass:function(e){var t,n,r=0;for(t=" "+e+" ";n=this[r++];)if(1===n.nodeType&&(" "+ee(te(n))+" ").indexOf(t)>-1)return!0;return!1}});var Ot=/\r/g;Ne.fn.extend({val:function(e){var t,n,r,i=this[0];{if(arguments.length)return r=we(e),this.each(function(n){var i;1===this.nodeType&&(i=r?e.call(this,n,Ne(this).val()):e,null==i?i="":"number"==typeof i?i+="":Array.isArray(i)&&(i=Ne.map(i,function(e){return null==e?"":e+""})),t=Ne.valHooks[this.type]||Ne.valHooks[this.nodeName.toLowerCase()],t&&"set"in t&&void 0!==t.set(this,i,"value")||(this.value=i))});if(i)return t=Ne.valHooks[i.type]||Ne.valHooks[i.nodeName.toLowerCase()],t&&"get"in t&&void 0!==(n=t.get(i,"value"))?n:(n=i.value,"string"==typeof n?n.replace(Ot,""):null==n?"":n)}}}),Ne.extend({valHooks:{option:{get:function(e){var t=Ne.find.attr(e,"value");return null!=t?t:ee(Ne.text(e))}},select:{get:function(e){var t,n,r,i=e.options,o=e.selectedIndex,a="select-one"===e.type,u=a?null:[],s=a?o+1:i.length;for(r=o<0?s:a?o:0;r-1)&&(n=!0);return n||(e.selectedIndex=-1),o}}}}),Ne.each(["radio","checkbox"],function(){Ne.valHooks[this]={set:function(e,t){if(Array.isArray(t))return e.checked=Ne.inArray(Ne(e).val(),t)>-1}},xe.checkOn||(Ne.valHooks[this].get=function(e){return null===e.getAttribute("value")?"on":e.value})}),xe.focusin="onfocusin"in n;var Mt=/^(?:focusinfocus|focusoutblur)$/,Dt=function(e){e.stopPropagation()};Ne.extend(Ne.event,{trigger:function(e,t,r,i){var o,a,u,s,l,c,f,p,d=[r||ce],h=ye.call(e,"type")?e.type:e,g=ye.call(e,"namespace")?e.namespace.split("."):[];if(a=p=u=r=r||ce,3!==r.nodeType&&8!==r.nodeType&&!Mt.test(h+Ne.event.triggered)&&(h.indexOf(".")>-1&&(g=h.split("."),h=g.shift(),g.sort()),l=h.indexOf(":")<0&&"on"+h,e=e[Ne.expando]?e:new Ne.Event(h,"object"==typeof e&&e),e.isTrigger=i?2:3,e.namespace=g.join("."),e.rnamespace=e.namespace?new RegExp("(^|\\.)"+g.join("\\.(?:.*\\.|)")+"(\\.|$)"):null,e.result=void 0,e.target||(e.target=r),t=null==t?[e]:Ne.makeArray(t,[e]),f=Ne.event.special[h]||{},i||!f.trigger||f.trigger.apply(r,t)!==!1)){if(!i&&!f.noBubble&&!Ee(r)){for(s=f.delegateType||h,Mt.test(s+h)||(a=a.parentNode);a;a=a.parentNode)d.push(a),u=a;u===(r.ownerDocument||ce)&&d.push(u.defaultView||u.parentWindow||n)}for(o=0;(a=d[o++])&&!e.isPropagationStopped();)p=a,e.type=o>1?s:f.bindType||h,c=(Ue.get(a,"events")||{})[e.type]&&Ue.get(a,"handle"),c&&c.apply(a,t),c=l&&a[l],c&&c.apply&&We(a)&&(e.result=c.apply(a,t),e.result===!1&&e.preventDefault());return e.type=h,i||e.isDefaultPrevented()||f._default&&f._default.apply(d.pop(),t)!==!1||!We(r)||l&&we(r[h])&&!Ee(r)&&(u=r[l],u&&(r[l]=null),Ne.event.triggered=h,e.isPropagationStopped()&&p.addEventListener(h,Dt),r[h](),e.isPropagationStopped()&&p.removeEventListener(h,Dt),Ne.event.triggered=void 0,u&&(r[l]=u)),e.result}},simulate:function(e,t,n){var r=Ne.extend(new Ne.Event,n,{type:e,isSimulated:!0});Ne.event.trigger(r,null,t)}}),Ne.fn.extend({trigger:function(e,t){return this.each(function(){Ne.event.trigger(e,t,this)})},triggerHandler:function(e,t){var n=this[0];if(n)return Ne.event.trigger(e,t,n,!0)}}),xe.focusin||Ne.each({focus:"focusin",blur:"focusout"},function(e,t){var n=function(e){Ne.event.simulate(t,e.target,Ne.event.fix(e))};Ne.event.special[t]={setup:function(){var r=this.ownerDocument||this,i=Ue.access(r,t);i||r.addEventListener(e,n,!0),Ue.access(r,t,(i||0)+1)},teardown:function(){var r=this.ownerDocument||this,i=Ue.access(r,t)-1;i?Ue.access(r,t,i):(r.removeEventListener(e,n,!0),Ue.remove(r,t))}}});var Rt=n.location,Pt=Date.now(),Lt=/\?/;Ne.parseXML=function(e){var t;if(!e||"string"!=typeof e)return null;try{t=(new n.DOMParser).parseFromString(e,"text/xml")}catch(r){t=void 0}return t&&!t.getElementsByTagName("parsererror").length||Ne.error("Invalid XML: "+e),t};var It=/\[\]$/,qt=/\r?\n/g,Ht=/^(?:submit|button|image|reset|file)$/i,Bt=/^(?:input|select|textarea|keygen)/i;Ne.param=function(e,t){var n,r=[],i=function(e,t){var n=we(t)?t():t;r[r.length]=encodeURIComponent(e)+"="+encodeURIComponent(null==n?"":n)};if(Array.isArray(e)||e.jquery&&!Ne.isPlainObject(e))Ne.each(e,function(){i(this.name,this.value)});else for(n in e)re(n,e[n],t,i);return r.join("&")},Ne.fn.extend({serialize:function(){return Ne.param(this.serializeArray())},serializeArray:function(){return this.map(function(){var e=Ne.prop(this,"elements");return e?Ne.makeArray(e):this}).filter(function(){var e=this.type;return this.name&&!Ne(this).is(":disabled")&&Bt.test(this.nodeName)&&!Ht.test(e)&&(this.checked||!tt.test(e))}).map(function(e,t){var n=Ne(this).val();return null==n?null:Array.isArray(n)?Ne.map(n,function(e){return{name:t.name,value:e.replace(qt,"\r\n")}}):{name:t.name,value:n.replace(qt,"\r\n")}}).get()}});var $t=/%20/g,zt=/#.*$/,Ft=/([?&])_=[^&]*/,Wt=/^(.*?):[ \t]*([^\r\n]*)$/gm,Ut=/^(?:about|app|app-storage|.+-extension|file|res|widget):$/,Kt=/^(?:GET|HEAD)$/,Gt=/^\/\//,Vt={},Xt={},Zt="*/".concat("*"),Qt=ce.createElement("a");Qt.href=Rt.href,Ne.extend({active:0,lastModified:{},etag:{},ajaxSettings:{url:Rt.href,type:"GET",isLocal:Ut.test(Rt.protocol),global:!0,processData:!0,async:!0,contentType:"application/x-www-form-urlencoded; charset=UTF-8",accepts:{"*":Zt,text:"text/plain",html:"text/html",xml:"application/xml, text/xml",json:"application/json, text/javascript"},contents:{xml:/\bxml\b/,html:/\bhtml/,json:/\bjson\b/},responseFields:{xml:"responseXML",text:"responseText",json:"responseJSON"},converters:{"* text":String,"text html":!0,"text json":JSON.parse,"text xml":Ne.parseXML},flatOptions:{url:!0,context:!0}},ajaxSetup:function(e,t){return t?ae(ae(e,Ne.ajaxSettings),t):ae(Ne.ajaxSettings,e)},ajaxPrefilter:ie(Vt),ajaxTransport:ie(Xt),ajax:function(e,t){function r(e,t,r,u){var l,p,d,_,x,w=t;c||(c=!0,s&&n.clearTimeout(s),i=void 0,a=u||"",E.readyState=e>0?4:0,l=e>=200&&e<300||304===e,r&&(_=ue(h,E,r)),_=se(h,_,E,l),l?(h.ifModified&&(x=E.getResponseHeader("Last-Modified"),x&&(Ne.lastModified[o]=x),x=E.getResponseHeader("etag"),x&&(Ne.etag[o]=x)),204===e||"HEAD"===h.type?w="nocontent":304===e?w="notmodified":(w=_.state,p=_.data,d=_.error,l=!d)):(d=w,!e&&w||(w="error",e<0&&(e=0))),E.status=e,E.statusText=(t||w)+"",l?m.resolveWith(g,[p,w,E]):m.rejectWith(g,[E,w,d]),E.statusCode(b),b=void 0,f&&v.trigger(l?"ajaxSuccess":"ajaxError",[E,h,l?p:d]),y.fireWith(g,[E,w]),f&&(v.trigger("ajaxComplete",[E,h]),--Ne.active||Ne.event.trigger("ajaxStop")))}"object"==typeof e&&(t=e,e=void 0),t=t||{};var i,o,a,u,s,l,c,f,p,d,h=Ne.ajaxSetup({},t),g=h.context||h,v=h.context&&(g.nodeType||g.jquery)?Ne(g):Ne.event,m=Ne.Deferred(),y=Ne.Callbacks("once memory"),b=h.statusCode||{},_={},x={},w="canceled",E={readyState:0,getResponseHeader:function(e){var t;if(c){if(!u)for(u={};t=Wt.exec(a);)u[t[1].toLowerCase()]=t[2];t=u[e.toLowerCase()]}return null==t?null:t},getAllResponseHeaders:function(){return c?a:null},setRequestHeader:function(e,t){return null==c&&(e=x[e.toLowerCase()]=x[e.toLowerCase()]||e,_[e]=t),this},overrideMimeType:function(e){return null==c&&(h.mimeType=e),this},statusCode:function(e){var t;if(e)if(c)E.always(e[E.status]);else for(t in e)b[t]=[b[t],e[t]];return this},abort:function(e){var t=e||w;return i&&i.abort(t),r(0,t),this}};if(m.promise(E),h.url=((e||h.url||Rt.href)+"").replace(Gt,Rt.protocol+"//"),h.type=t.method||t.type||h.method||h.type,h.dataTypes=(h.dataType||"*").toLowerCase().match(qe)||[""],null==h.crossDomain){l=ce.createElement("a");try{l.href=h.url,l.href=l.href,h.crossDomain=Qt.protocol+"//"+Qt.host!=l.protocol+"//"+l.host}catch(k){h.crossDomain=!0}}if(h.data&&h.processData&&"string"!=typeof h.data&&(h.data=Ne.param(h.data,h.traditional)),oe(Vt,h,t,E),c)return E;f=Ne.event&&h.global,f&&0===Ne.active++&&Ne.event.trigger("ajaxStart"),h.type=h.type.toUpperCase(),h.hasContent=!Kt.test(h.type),o=h.url.replace(zt,""),h.hasContent?h.data&&h.processData&&0===(h.contentType||"").indexOf("application/x-www-form-urlencoded")&&(h.data=h.data.replace($t,"+")):(d=h.url.slice(o.length),h.data&&(h.processData||"string"==typeof h.data)&&(o+=(Lt.test(o)?"&":"?")+h.data,delete h.data),h.cache===!1&&(o=o.replace(Ft,"$1"),d=(Lt.test(o)?"&":"?")+"_="+Pt++ +d),h.url=o+d),h.ifModified&&(Ne.lastModified[o]&&E.setRequestHeader("If-Modified-Since",Ne.lastModified[o]),Ne.etag[o]&&E.setRequestHeader("If-None-Match",Ne.etag[o])),(h.data&&h.hasContent&&h.contentType!==!1||t.contentType)&&E.setRequestHeader("Content-Type",h.contentType),E.setRequestHeader("Accept",h.dataTypes[0]&&h.accepts[h.dataTypes[0]]?h.accepts[h.dataTypes[0]]+("*"!==h.dataTypes[0]?", "+Zt+"; q=0.01":""):h.accepts["*"]);for(p in h.headers)E.setRequestHeader(p,h.headers[p]);if(h.beforeSend&&(h.beforeSend.call(g,E,h)===!1||c))return E.abort();if(w="abort",y.add(h.complete),E.done(h.success),E.fail(h.error),i=oe(Xt,h,t,E)){if(E.readyState=1,f&&v.trigger("ajaxSend",[E,h]),c)return E;h.async&&h.timeout>0&&(s=n.setTimeout(function(){E.abort("timeout")},h.timeout));try{c=!1,i.send(_,r)}catch(k){if(c)throw k;r(-1,k)}}else r(-1,"No Transport");return E},getJSON:function(e,t,n){return Ne.get(e,t,n,"json")},getScript:function(e,t){return Ne.get(e,void 0,t,"script")}}),Ne.each(["get","post"],function(e,t){Ne[t]=function(e,n,r,i){return we(n)&&(i=i||r,r=n,n=void 0),Ne.ajax(Ne.extend({url:e,type:t,dataType:i,data:n,success:r},Ne.isPlainObject(e)&&e))}}),Ne._evalUrl=function(e){return Ne.ajax({url:e,type:"GET",dataType:"script",cache:!0,async:!1,global:!1,"throws":!0})},Ne.fn.extend({wrapAll:function(e){var t;return this[0]&&(we(e)&&(e=e.call(this[0])),t=Ne(e,this[0].ownerDocument).eq(0).clone(!0),this[0].parentNode&&t.insertBefore(this[0]),t.map(function(){for(var e=this;e.firstElementChild;)e=e.firstElementChild;return e}).append(this)),this},wrapInner:function(e){return we(e)?this.each(function(t){Ne(this).wrapInner(e.call(this,t))}):this.each(function(){var t=Ne(this),n=t.contents();n.length?n.wrapAll(e):t.append(e)})},wrap:function(e){var t=we(e);return this.each(function(n){Ne(this).wrapAll(t?e.call(this,n):e)})},unwrap:function(e){return this.parent(e).not("body").each(function(){Ne(this).replaceWith(this.childNodes)}),this}}),Ne.expr.pseudos.hidden=function(e){return!Ne.expr.pseudos.visible(e)},Ne.expr.pseudos.visible=function(e){return!!(e.offsetWidth||e.offsetHeight||e.getClientRects().length)},Ne.ajaxSettings.xhr=function(){try{return new n.XMLHttpRequest}catch(e){}};var Yt={0:200,1223:204},Jt=Ne.ajaxSettings.xhr();xe.cors=!!Jt&&"withCredentials"in Jt,xe.ajax=Jt=!!Jt,Ne.ajaxTransport(function(e){var t,r;if(xe.cors||Jt&&!e.crossDomain)return{send:function(i,o){var a,u=e.xhr();if(u.open(e.type,e.url,e.async,e.username,e.password),e.xhrFields)for(a in e.xhrFields)u[a]=e.xhrFields[a];e.mimeType&&u.overrideMimeType&&u.overrideMimeType(e.mimeType),e.crossDomain||i["X-Requested-With"]||(i["X-Requested-With"]="XMLHttpRequest");for(a in i)u.setRequestHeader(a,i[a]);t=function(e){return function(){t&&(t=r=u.onload=u.onerror=u.onabort=u.ontimeout=u.onreadystatechange=null,"abort"===e?u.abort():"error"===e?"number"!=typeof u.status?o(0,"error"):o(u.status,u.statusText):o(Yt[u.status]||u.status,u.statusText,"text"!==(u.responseType||"text")||"string"!=typeof u.responseText?{binary:u.response}:{text:u.responseText},u.getAllResponseHeaders()))}},u.onload=t(),r=u.onerror=u.ontimeout=t("error"),void 0!==u.onabort?u.onabort=r:u.onreadystatechange=function(){4===u.readyState&&n.setTimeout(function(){t&&r()})},t=t("abort");try{u.send(e.hasContent&&e.data||null)}catch(s){if(t)throw s}},abort:function(){t&&t()}}}),Ne.ajaxPrefilter(function(e){e.crossDomain&&(e.contents.script=!1)}),Ne.ajaxSetup({accepts:{script:"text/javascript, application/javascript, application/ecmascript, application/x-ecmascript"},contents:{script:/\b(?:java|ecma)script\b/},converters:{"text script":function(e){return Ne.globalEval(e),e}}}),Ne.ajaxPrefilter("script",function(e){void 0===e.cache&&(e.cache=!1),e.crossDomain&&(e.type="GET")}),Ne.ajaxTransport("script",function(e){if(e.crossDomain){var t,n;return{send:function(r,i){t=Ne("",returnEnd:!0,subLanguage:["actionscript","javascript","handlebars","xml"]}},{className:"meta",variants:[{begin:/<\?xml/,end:/\?>/,relevance:10},{begin:/<\?\w+/,end:/\?>/}]},{className:"tag",begin:"",contains:[{className:"name",begin:/[^\/><\s]+/,relevance:0},n]}]}}),e.registerLanguage("markdown",function(e){return{aliases:["md","mkdown","mkd"],contains:[{className:"section",variants:[{begin:"^#{1,6}",end:"$"},{begin:"^.+?\\n[=-]{2,}$"}]},{begin:"<",end:">",subLanguage:"xml",relevance:0},{className:"bullet",begin:"^([*+-]|(\\d+\\.))\\s+"},{className:"strong",begin:"[*_]{2}.+?[*_]{2}"},{className:"emphasis",variants:[{begin:"\\*.+?\\*"},{begin:"_.+?_",relevance:0}]},{className:"quote",begin:"^>\\s+",end:"$"},{className:"code",variants:[{begin:"^```w*s*$",end:"^```s*$"},{begin:"`.+?`"},{begin:"^( {4}|\t)",end:"$",relevance:0}]},{begin:"^[-\\*]{3,}",end:"$"},{begin:"\\[.+?\\][\\(\\[].*?[\\)\\]]",returnBegin:!0,contains:[{className:"string",begin:"\\[",end:"\\]",excludeBegin:!0,returnEnd:!0,relevance:0},{className:"link",begin:"\\]\\(",end:"\\)",excludeBegin:!0,excludeEnd:!0},{className:"symbol",begin:"\\]\\[",end:"\\]",excludeBegin:!0,excludeEnd:!0}],relevance:10},{begin:/^\[[^\n]+\]:/,returnBegin:!0,contains:[{className:"symbol",begin:/\[/,end:/\]/,excludeBegin:!0,excludeEnd:!0},{className:"link",begin:/:\s*/,end:/$/,excludeBegin:!0}]}]}}),e.registerLanguage("sql",function(e){var t=e.COMMENT("--","$");return{case_insensitive:!0,illegal:/[<>{}*#]/,contains:[{beginKeywords:"begin end start commit rollback savepoint lock alter create drop rename call delete do handler insert load replace select truncate update set show pragma grant merge describe use explain help declare prepare execute deallocate release unlock purge reset change stop analyze cache flush optimize repair kill install uninstall checksum restore check backup revoke comment",end:/;/,endsWithParent:!0,lexemes:/[\w\.]+/,keywords:{keyword:"abort abs absolute acc acce accep accept access accessed accessible account acos action activate add addtime admin administer advanced advise aes_decrypt aes_encrypt after agent aggregate ali alia alias allocate allow alter always analyze ancillary and any anydata anydataset anyschema anytype apply archive archived archivelog are as asc ascii asin assembly assertion associate asynchronous at atan atn2 attr attri attrib attribu attribut attribute attributes audit authenticated authentication authid authors auto autoallocate autodblink autoextend automatic availability avg backup badfile basicfile before begin beginning benchmark between bfile bfile_base big bigfile bin binary_double binary_float binlog bit_and bit_count bit_length bit_or bit_xor bitmap blob_base block blocksize body both bound buffer_cache buffer_pool build bulk by byte byteordermark bytes cache caching call calling cancel capacity cascade cascaded case cast catalog category ceil ceiling chain change changed char_base char_length character_length characters characterset charindex charset charsetform charsetid check checksum checksum_agg child choose chr chunk class cleanup clear client clob clob_base clone close cluster_id cluster_probability cluster_set clustering coalesce coercibility col collate collation collect colu colum column column_value columns columns_updated comment commit compact compatibility compiled complete composite_limit compound compress compute concat concat_ws concurrent confirm conn connec connect connect_by_iscycle connect_by_isleaf connect_by_root connect_time connection consider consistent constant constraint constraints constructor container content contents context contributors controlfile conv convert convert_tz corr corr_k corr_s corresponding corruption cos cost count count_big counted covar_pop covar_samp cpu_per_call cpu_per_session crc32 create creation critical cross cube cume_dist curdate current current_date current_time current_timestamp current_user cursor curtime customdatum cycle data database databases datafile datafiles datalength date_add date_cache date_format date_sub dateadd datediff datefromparts datename datepart datetime2fromparts day day_to_second dayname dayofmonth dayofweek dayofyear days db_role_change dbtimezone ddl deallocate declare decode decompose decrement decrypt deduplicate def defa defau defaul default defaults deferred defi defin define degrees delayed delegate delete delete_all delimited demand dense_rank depth dequeue des_decrypt des_encrypt des_key_file desc descr descri describ describe descriptor deterministic diagnostics difference dimension direct_load directory disable disable_all disallow disassociate discardfile disconnect diskgroup distinct distinctrow distribute distributed div do document domain dotnet double downgrade drop dumpfile duplicate duration each edition editionable editions element ellipsis else elsif elt empty enable enable_all enclosed encode encoding encrypt end end-exec endian enforced engine engines enqueue enterprise entityescaping eomonth error errors escaped evalname evaluate event eventdata events except exception exceptions exchange exclude excluding execu execut execute exempt exists exit exp expire explain export export_set extended extent external external_1 external_2 externally extract failed failed_login_attempts failover failure far fast feature_set feature_value fetch field fields file file_name_convert filesystem_like_logging final finish first first_value fixed flash_cache flashback floor flush following follows for forall force form forma format found found_rows freelist freelists freepools fresh from from_base64 from_days ftp full function general generated get get_format get_lock getdate getutcdate global global_name globally go goto grant grants greatest group group_concat group_id grouping grouping_id groups gtid_subtract guarantee guard handler hash hashkeys having hea head headi headin heading heap help hex hierarchy high high_priority hosts hour http id ident_current ident_incr ident_seed identified identity idle_time if ifnull ignore iif ilike ilm immediate import in include including increment index indexes indexing indextype indicator indices inet6_aton inet6_ntoa inet_aton inet_ntoa infile initial initialized initially initrans inmemory inner innodb input insert install instance instantiable instr interface interleaved intersect into invalidate invisible is is_free_lock is_ipv4 is_ipv4_compat is_not is_not_null is_used_lock isdate isnull isolation iterate java join json json_exists keep keep_duplicates key keys kill language large last last_day last_insert_id last_value lax lcase lead leading least leaves left len lenght length less level levels library like like2 like4 likec limit lines link list listagg little ln load load_file lob lobs local localtime localtimestamp locate locator lock locked log log10 log2 logfile logfiles logging logical logical_reads_per_call logoff logon logs long loop low low_priority lower lpad lrtrim ltrim main make_set makedate maketime managed management manual map mapping mask master master_pos_wait match matched materialized max maxextents maximize maxinstances maxlen maxlogfiles maxloghistory maxlogmembers maxsize maxtrans md5 measures median medium member memcompress memory merge microsecond mid migration min minextents minimum mining minus minute minvalue missing mod mode model modification modify module monitoring month months mount move movement multiset mutex name name_const names nan national native natural nav nchar nclob nested never new newline next nextval no no_write_to_binlog noarchivelog noaudit nobadfile nocheck nocompress nocopy nocycle nodelay nodiscardfile noentityescaping noguarantee nokeep nologfile nomapping nomaxvalue nominimize nominvalue nomonitoring none noneditionable nonschema noorder nopr nopro noprom nopromp noprompt norely noresetlogs noreverse normal norowdependencies noschemacheck noswitch not nothing notice notrim novalidate now nowait nth_value nullif nulls num numb numbe nvarchar nvarchar2 object ocicoll ocidate ocidatetime ociduration ociinterval ociloblocator ocinumber ociref ocirefcursor ocirowid ocistring ocitype oct octet_length of off offline offset oid oidindex old on online only opaque open operations operator optimal optimize option optionally or oracle oracle_date oradata ord ordaudio orddicom orddoc order ordimage ordinality ordvideo organization orlany orlvary out outer outfile outline output over overflow overriding package pad parallel parallel_enable parameters parent parse partial partition partitions pascal passing password password_grace_time password_lock_time password_reuse_max password_reuse_time password_verify_function patch path patindex pctincrease pctthreshold pctused pctversion percent percent_rank percentile_cont percentile_disc performance period period_add period_diff permanent physical pi pipe pipelined pivot pluggable plugin policy position post_transaction pow power pragma prebuilt precedes preceding precision prediction prediction_cost prediction_details prediction_probability prediction_set prepare present preserve prior priority private private_sga privileges procedural procedure procedure_analyze processlist profiles project prompt protection public publishingservername purge quarter query quick quiesce quota quotename radians raise rand range rank raw read reads readsize rebuild record records recover recovery recursive recycle redo reduced ref reference referenced references referencing refresh regexp_like register regr_avgx regr_avgy regr_count regr_intercept regr_r2 regr_slope regr_sxx regr_sxy reject rekey relational relative relaylog release release_lock relies_on relocate rely rem remainder rename repair repeat replace replicate replication required reset resetlogs resize resource respect restore restricted result result_cache resumable resume retention return returning returns reuse reverse revoke right rlike role roles rollback rolling rollup round row row_count rowdependencies rowid rownum rows rtrim rules safe salt sample save savepoint sb1 sb2 sb4 scan schema schemacheck scn scope scroll sdo_georaster sdo_topo_geometry search sec_to_time second section securefile security seed segment select self sequence sequential serializable server servererror session session_user sessions_per_user set sets settings sha sha1 sha2 share shared shared_pool short show shrink shutdown si_averagecolor si_colorhistogram si_featurelist si_positionalcolor si_stillimage si_texture siblings sid sign sin size size_t sizes skip slave sleep smalldatetimefromparts smallfile snapshot some soname sort soundex source space sparse spfile split sql sql_big_result sql_buffer_result sql_cache sql_calc_found_rows sql_small_result sql_variant_property sqlcode sqldata sqlerror sqlname sqlstate sqrt square standalone standby start starting startup statement static statistics stats_binomial_test stats_crosstab stats_ks_test stats_mode stats_mw_test stats_one_way_anova stats_t_test_ stats_t_test_indep stats_t_test_one stats_t_test_paired stats_wsr_test status std stddev stddev_pop stddev_samp stdev stop storage store stored str str_to_date straight_join strcmp strict string struct stuff style subdate subpartition subpartitions substitutable substr substring subtime subtring_index subtype success sum suspend switch switchoffset switchover sync synchronous synonym sys sys_xmlagg sysasm sysaux sysdate sysdatetimeoffset sysdba sysoper system system_user sysutcdatetime table tables tablespace tan tdo template temporary terminated tertiary_weights test than then thread through tier ties time time_format time_zone timediff timefromparts timeout timestamp timestampadd timestampdiff timezone_abbr timezone_minute timezone_region to to_base64 to_date to_days to_seconds todatetimeoffset trace tracking transaction transactional translate translation treat trigger trigger_nestlevel triggers trim truncate try_cast try_convert try_parse type ub1 ub2 ub4 ucase unarchived unbounded uncompress under undo unhex unicode uniform uninstall union unique unix_timestamp unknown unlimited unlock unpivot unrecoverable unsafe unsigned until untrusted unusable unused update updated upgrade upped upper upsert url urowid usable usage use use_stored_outlines user user_data user_resources users using utc_date utc_timestamp uuid uuid_short validate validate_password_strength validation valist value values var var_samp varcharc vari varia variab variabl variable variables variance varp varraw varrawc varray verify version versions view virtual visible void wait wallet warning warnings week weekday weekofyear wellformed when whene whenev wheneve whenever where while whitespace with within without work wrapped xdb xml xmlagg xmlattributes xmlcast xmlcolattval xmlelement xmlexists xmlforest xmlindex xmlnamespaces xmlpi xmlquery xmlroot xmlschema xmlserialize xmltable xmltype xor year year_to_month years yearweek",literal:"true false null",built_in:"array bigint binary bit blob boolean char character date dec decimal float int int8 integer interval number numeric real record serial serial8 smallint text varchar varying void"},contains:[{className:"string",begin:"'",end:"'",contains:[e.BACKSLASH_ESCAPE,{begin:"''"}]},{className:"string",begin:'"',end:'"',contains:[e.BACKSLASH_ESCAPE,{begin:'""'}]},{className:"string",begin:"`",end:"`",contains:[e.BACKSLASH_ESCAPE]},e.C_NUMBER_MODE,e.C_BLOCK_COMMENT_MODE,t]},e.C_BLOCK_COMMENT_MODE,t]}}),e})},function(e,t,n){"use strict";function r(e){w.forEach(function(t){t===e?(0,v["default"])(t).parent().addClass("selected"):(0,v["default"])(t).parent().removeClass("selected")})}function i(){var e=(0,v["default"])("#full-list"),t=(0,v["default"])("#full-list .clicked");t.length>0&&e.scrollTop(t.offset().top-e.offset().top-40)}function o(e,t){var n=b.getModuleType();t=t||n;var i=e[t]||[],o=(0,v["default"])("#full-list");o.replaceWith((0,x["default"])({nodes:i,group:""})),r(["#",t,"-list"].join("")),(0,v["default"])("#full-list li a").on("click",function(e){var t=(0,v["default"])(e.target);t.hasClass("expand")?(e.preventDefault(),(0,v["default"])(e.target).closest("li").toggleClass("open")):((0,v["default"])("#full-list .clicked li.active").removeClass("active"),(0,v["default"])(e.target).closest("li").addClass("active"))})}function a(e){return function(t){t.preventDefault(),o(sidebarNodes,e),i()}}function u(){E.on("click","#extras-list",a("extras")),E.on("click","#modules-list",a("modules")),E.on("click","#exceptions-list",a("exceptions")),E.on("click","#tasks-list",a("tasks")),(0,v["default"])(".sidebar-search input").on("keydown",function(e){27===e.keyCode?(0,v["default"])(this).val(""):(event.metaKey||event.ctrlKey)&&13===e.keyCode&&((0,v["default"])(this).parent().attr("target","_blank").submit().removeAttr(""),e.preventDefault())});var e=window.location.pathname;"search.html"===e.substr(e.lastIndexOf("/")+1)&&(0,m.search)(s("q"))}function s(e){var t=window.location.href,n=e.replace(/[\[\]]/g,"\\$&"),r=new RegExp("[?&]"+n+"(=([^&#]*)|&|#|$)"),i=r.exec(t);return i&&i[2]?decodeURIComponent(i[2].replace(/\+/g," ")):""}function l(){var e=b.getLocationHash()||"content",t=sidebarNodes[b.getModuleType()],n=b.findSidebarCategory(t,e);(0,v["default"])('#full-list .clicked a.expand[href$="#'+n+'"]').closest("li").addClass("open"),(0,v["default"])('#full-list .clicked a[href$="#'+e+'"]').closest("li").addClass("active")}function c(){k.find("a").has("code").addClass("no-underline"),k.find("a").has("img").addClass("no-underline")}function f(){k.attr("tabindex",-1).focus()}function p(){o(sidebarNodes),u(),i(),l(),c(),f()}var d=n(1)["default"],h=n(5)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.initialize=p;var g=n(2),v=d(g),m=n(6),y=n(7),b=h(y),_=n(81),x=d(_),w=["#extras-list","#modules-list","#exceptions-list","#tasks-list","#search-list"],E=(0,v["default"])(".sidebar-listNav"),k=(0,v["default"])(".content")},function(e,t){"use strict";t["default"]=function(e){if(e&&e.__esModule)return e;var t={};if(null!=e)for(var n in e)Object.prototype.hasOwnProperty.call(e,n)&&(t[n]=e[n]);return t["default"]=e,t},t.__esModule=!0},function(e,t,n){"use strict";function r(e){var t=e.index,n=e.index+e[0].length,r=e.input,i=""+e[0]+"";return r.slice(0,t)+i+r.slice(n)}function i(e){return!!e}function o(e,t,n){return(e||[]).map(function(e){var i=(t+"."+e.id).match(n),o=e.id&&e.id.match(n);if(i||o){var a=JSON.parse(JSON.stringify(e));return a.match=o?r(o):e.id,a}}).filter(i)}function a(e,t,n){t.length>0&&e.push({name:n,results:t})}function u(e,t){return e.map(function(e){var n=e.title,i=n&&n.match(t),a=o(e.functions,n,t),u=o(e.macros,n,t),s=o(e.callbacks,n,t),l=o(e.types,n,t),c={id:e.id,match:i?r(i):e.title};if(a.length>0&&(c.functions=a),u.length>0&&(c.macros=u),s.length>0&&(c.callbacks=s),l.length>0&&(c.types=l),i||a.length>0||u.length>0||s.length>0||l.length>0)return c}).filter(i)}function s(e){var t=sidebarNodes;if(""!==e.replace(/\s/,"")){var n=new RegExp(h.escapeText(e),"i"),r=[],i=u(t.modules,n),o=u(t.exceptions,n),s=u(t.tasks,n);a(r,i,"Modules"),a(r,o,"Exceptions"),a(r,s,"Mix Tasks");var l=(0,v["default"])({value:e,levels:r,empty:0===r.length});y.val(e),m.html(l)}}var l=n(1)["default"],c=n(5)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.findIn=u,t.search=s;var f=n(2),p=l(f),d=n(7),h=c(d),g=n(61),v=l(g),m=(0,p["default"])("#search"),y=(0,p["default"])(".sidebar-search input")},function(e,t,n){"use strict";function r(e){return e.replace(/[\-\[\]{}()*+?.,\\\^$|#\s]/g,"\\$&")}function i(){return(0,c["default"])("body").data("type")}function o(e,t){var n=!0,r=!1,i=void 0;try{for(var o,a=u(e);!(n=(o=a.next()).done);n=!0){var s=o.value,l=(0,h["default"])(s,function(e,n){var r=(0,p["default"])(e,function(e){var n=e.anchor;return n===t});return r});if(l)return l}}catch(c){r=!0,i=c}finally{try{!n&&a["return"]&&a["return"]()}finally{if(r)throw i}}}function a(){return window.location.hash.replace(/^#/,"")}var u=n(8)["default"],s=n(1)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.escapeText=r,t.getModuleType=i,t.findSidebarCategory=o,t.getLocationHash=a;var l=n(2),c=s(l),f=n(46),p=s(f),d=n(59),h=s(d)},function(e,t,n){e.exports={"default":n(9),__esModule:!0}},function(e,t,n){n(10),n(38),e.exports=n(41)},function(e,t,n){n(11);var r=n(14);r.NodeList=r.HTMLCollection=r.Array},function(e,t,n){"use strict";var r=n(12),i=n(13),o=n(14),a=n(15);e.exports=n(19)(Array,"Array",function(e,t){this._t=a(e),this._i=0,this._k=t},function(){var e=this._t,t=this._k,n=this._i++;return!e||n>=e.length?(this._t=void 0,i(1)):"keys"==t?i(0,n):"values"==t?i(0,e[n]):i(0,[n,e[n]])},"values"),o.Arguments=o.Array,r("keys"),r("values"),r("entries")},function(e,t){e.exports=function(){}},function(e,t){e.exports=function(e,t){return{value:t,done:!!e}}},function(e,t){e.exports={}},function(e,t,n){var r=n(16),i=n(18);e.exports=function(e){return r(i(e))}},function(e,t,n){var r=n(17);e.exports=Object("z").propertyIsEnumerable(0)?Object:function(e){return"String"==r(e)?e.split(""):Object(e)}},function(e,t){ +var n={}.toString;e.exports=function(e){return n.call(e).slice(8,-1)}},function(e,t){e.exports=function(e){if(void 0==e)throw TypeError("Can't call method on "+e);return e}},function(e,t,n){"use strict";var r=n(20),i=n(21),o=n(26),a=n(27),u=n(32),s=n(14),l=n(33),c=n(34),f=n(28).getProto,p=n(35)("iterator"),d=!([].keys&&"next"in[].keys()),h="@@iterator",g="keys",v="values",m=function(){return this};e.exports=function(e,t,n,y,b,_,x){l(n,t,y);var w,E,k=function(e){if(!d&&e in A)return A[e];switch(e){case g:return function(){return new n(this,e)};case v:return function(){return new n(this,e)}}return function(){return new n(this,e)}},C=t+" Iterator",N=b==v,T=!1,A=e.prototype,S=A[p]||A[h]||b&&A[b],j=S||k(b);if(S){var O=f(j.call(new e));c(O,C,!0),!r&&u(A,h)&&a(O,p,m),N&&S.name!==v&&(T=!0,j=function(){return S.call(this)})}if(r&&!x||!d&&!T&&A[p]||a(A,p,j),s[t]=j,s[C]=m,b)if(w={values:N?j:k(v),keys:_?j:k(g),entries:N?k("entries"):j},x)for(E in w)E in A||o(A,E,w[E]);else i(i.P+i.F*(d||T),t,w);return w}},function(e,t){e.exports=!0},function(e,t,n){var r=n(22),i=n(23),o=n(24),a="prototype",u=function(e,t,n){var s,l,c,f=e&u.F,p=e&u.G,d=e&u.S,h=e&u.P,g=e&u.B,v=e&u.W,m=p?i:i[t]||(i[t]={}),y=p?r:d?r[t]:(r[t]||{})[a];p&&(n=t);for(s in n)l=!f&&y&&s in y,l&&s in m||(c=l?y[s]:n[s],m[s]=p&&"function"!=typeof y[s]?n[s]:g&&l?o(c,r):v&&y[s]==c?function(e){var t=function(t){return this instanceof e?new e(t):e(t)};return t[a]=e[a],t}(c):h&&"function"==typeof c?o(Function.call,c):c,h&&((m[a]||(m[a]={}))[s]=c))};u.F=1,u.G=2,u.S=4,u.P=8,u.B=16,u.W=32,e.exports=u},function(e,t){var n=e.exports="undefined"!=typeof window&&window.Math==Math?window:"undefined"!=typeof self&&self.Math==Math?self:Function("return this")();"number"==typeof __g&&(__g=n)},function(e,t){var n=e.exports={version:"1.2.6"};"number"==typeof __e&&(__e=n)},function(e,t,n){var r=n(25);e.exports=function(e,t,n){if(r(e),void 0===t)return e;switch(n){case 1:return function(n){return e.call(t,n)};case 2:return function(n,r){return e.call(t,n,r)};case 3:return function(n,r,i){return e.call(t,n,r,i)}}return function(){return e.apply(t,arguments)}}},function(e,t){e.exports=function(e){if("function"!=typeof e)throw TypeError(e+" is not a function!");return e}},function(e,t,n){e.exports=n(27)},function(e,t,n){var r=n(28),i=n(29);e.exports=n(30)?function(e,t,n){return r.setDesc(e,t,i(1,n))}:function(e,t,n){return e[t]=n,e}},function(e,t){var n=Object;e.exports={create:n.create,getProto:n.getPrototypeOf,isEnum:{}.propertyIsEnumerable,getDesc:n.getOwnPropertyDescriptor,setDesc:n.defineProperty,setDescs:n.defineProperties,getKeys:n.keys,getNames:n.getOwnPropertyNames,getSymbols:n.getOwnPropertySymbols,each:[].forEach}},function(e,t){e.exports=function(e,t){return{enumerable:!(1&e),configurable:!(2&e),writable:!(4&e),value:t}}},function(e,t,n){e.exports=!n(31)(function(){return 7!=Object.defineProperty({},"a",{get:function(){return 7}}).a})},function(e,t){e.exports=function(e){try{return!!e()}catch(t){return!0}}},function(e,t){var n={}.hasOwnProperty;e.exports=function(e,t){return n.call(e,t)}},function(e,t,n){"use strict";var r=n(28),i=n(29),o=n(34),a={};n(27)(a,n(35)("iterator"),function(){return this}),e.exports=function(e,t,n){e.prototype=r.create(a,{next:i(1,n)}),o(e,t+" Iterator")}},function(e,t,n){var r=n(28).setDesc,i=n(32),o=n(35)("toStringTag");e.exports=function(e,t,n){e&&!i(e=n?e:e.prototype,o)&&r(e,o,{configurable:!0,value:t})}},function(e,t,n){var r=n(36)("wks"),i=n(37),o=n(22).Symbol;e.exports=function(e){return r[e]||(r[e]=o&&o[e]||(o||i)("Symbol."+e))}},function(e,t,n){var r=n(22),i="__core-js_shared__",o=r[i]||(r[i]={});e.exports=function(e){return o[e]||(o[e]={})}},function(e,t){var n=0,r=Math.random();e.exports=function(e){return"Symbol(".concat(void 0===e?"":e,")_",(++n+r).toString(36))}},function(e,t,n){"use strict";var r=n(39)(!0);n(19)(String,"String",function(e){this._t=String(e),this._i=0},function(){var e,t=this._t,n=this._i;return n>=t.length?{value:void 0,done:!0}:(e=r(t,n),this._i+=e.length,{value:e,done:!1})})},function(e,t,n){var r=n(40),i=n(18);e.exports=function(e){return function(t,n){var o,a,u=String(i(t)),s=r(n),l=u.length;return s<0||s>=l?e?"":void 0:(o=u.charCodeAt(s),o<55296||o>56319||s+1===l||(a=u.charCodeAt(s+1))<56320||a>57343?e?u.charAt(s):o:e?u.slice(s,s+2):(o-55296<<10)+(a-56320)+65536)}}},function(e,t){var n=Math.ceil,r=Math.floor;e.exports=function(e){return isNaN(e=+e)?0:(e>0?r:n)(e)}},function(e,t,n){var r=n(42),i=n(44);e.exports=n(23).getIterator=function(e){var t=i(e);if("function"!=typeof t)throw TypeError(e+" is not iterable!");return r(t.call(e))}},function(e,t,n){var r=n(43);e.exports=function(e){if(!r(e))throw TypeError(e+" is not an object!");return e}},function(e,t){e.exports=function(e){return"object"==typeof e?null!==e:"function"==typeof e}},function(e,t,n){var r=n(45),i=n(35)("iterator"),o=n(14);e.exports=n(23).getIteratorMethod=function(e){if(void 0!=e)return e[i]||e["@@iterator"]||o[r(e)]}},function(e,t,n){var r=n(17),i=n(35)("toStringTag"),o="Arguments"==r(function(){return arguments}());e.exports=function(e){var t,n,a;return void 0===e?"Undefined":null===e?"Null":"string"==typeof(n=(t=Object(e))[i])?n:o?r(t):"Object"==(a=r(t))&&"function"==typeof t.callee?"Arguments":a}},function(e,t,n){function r(e,t){return function(n,r,o){if(r=i(r,o,3),s(n)){var l=u(n,r,t);return l>-1?n[l]:void 0}return a(n,r,e)}}var i=n(47),o=n(56),a=n(57),u=n(58),s=n(49),l=r(o);e.exports=l},function(e,t,n){function r(e){return null==e?"":e+""}function i(e,t,n){var r=typeof e;return"function"==r?void 0===t?e:w(e,t,n):null==e?b:"object"==r?u(e):void 0===t?_(e):s(e,t)}function o(e,t,n){if(null!=e){void 0!==n&&n in g(e)&&(t=[n]);for(var r=0,i=t.length;null!=e&&ri?0:i+t),n=void 0===n||n>i?i:+n||0,n<0&&(n+=i),i=t>n?0:n-t>>>0,t>>>=0;for(var o=Array(i);++rl))return!1;for(;++s-1&&e%1==0&&e<=m}function o(e){return a(e)&&h.call(e)==l}function a(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}function u(e){return null!=e&&(o(e)?g.test(p.call(e)):n(e)&&c.test(e))}var s="[object Array]",l="[object Function]",c=/^\[object .+?Constructor\]$/,f=Object.prototype,p=Function.prototype.toString,d=f.hasOwnProperty,h=f.toString,g=RegExp("^"+p.call(d).replace(/[\\^$.*+?()[\]{}|]/g,"\\$&").replace(/hasOwnProperty|(function).*?(?=\\\()| for .+?(?=\\\])/g,"$1.*?")+"$"),v=r(Array,"isArray"),m=9007199254740991,y=v||function(e){return n(e)&&i(e.length)&&h.call(e)==s};e.exports=y},function(e,t){function n(e){return"number"==typeof e&&e>-1&&e%1==0&&e<=o}function r(e){return!!e&&"object"==typeof e}function i(e){return r(e)&&n(e.length)&&!!j[M.call(e)]}var o=9007199254740991,a="[object Arguments]",u="[object Array]",s="[object Boolean]",l="[object Date]",c="[object Error]",f="[object Function]",p="[object Map]",d="[object Number]",h="[object Object]",g="[object RegExp]",v="[object Set]",m="[object String]",y="[object WeakMap]",b="[object ArrayBuffer]",_="[object DataView]",x="[object Float32Array]",w="[object Float64Array]",E="[object Int8Array]",k="[object Int16Array]",C="[object Int32Array]",N="[object Uint8Array]",T="[object Uint8ClampedArray]",A="[object Uint16Array]",S="[object Uint32Array]",j={};j[x]=j[w]=j[E]=j[k]=j[C]=j[N]=j[T]=j[A]=j[S]=!0,j[a]=j[u]=j[b]=j[s]=j[_]=j[l]=j[c]=j[f]=j[p]=j[d]=j[h]=j[g]=j[v]=j[m]=j[y]=!1;var O=Object.prototype,M=O.toString;e.exports=i},function(e,t,n){function r(e){return function(t){return null==t?void 0:t[e]}}function i(e){return null!=e&&a(y(e))}function o(e,t){return e="number"==typeof e||d.test(e)?+e:-1,t=null==t?m:t,e>-1&&e%1==0&&e-1&&e%1==0&&e<=m}function u(e){for(var t=l(e),n=t.length,r=n&&e.length,i=!!r&&a(r)&&(p(e)||f(e)),u=-1,s=[];++u0;++r-1&&e%1==0&&e<=l}function u(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}function s(e){return!!e&&"object"==typeof e}var l=9007199254740991,c="[object Arguments]",f="[object Function]",p="[object GeneratorFunction]",d=Object.prototype,h=d.hasOwnProperty,g=d.toString,v=d.propertyIsEnumerable;e.exports=n},function(e,t){function n(e,t,n){if("function"!=typeof e)return r;if(void 0===t)return e;switch(n){case 1:return function(n){return e.call(t,n)};case 3:return function(n,r,i){return e.call(t,n,r,i)};case 4:return function(n,r,i,o){return e.call(t,n,r,i,o)};case 5:return function(n,r,i,o,a){return e.call(t,n,r,i,o,a)}}return function(){return e.apply(t,arguments)}}function r(e){return e}e.exports=n},function(e,t,n){function r(e){return i(e)?e:Object(e)}function i(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}function o(e){e=r(e);for(var t=-1,n=a(e),i=n.length,o=Array(i);++t-1&&e%1==0&&e<=f}function s(e){return l(e)?e:Object(e)}function l(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}var c=n(51),f=9007199254740991,p=o(r),d=a(),h=i("length");e.exports=p},function(e,t){function n(e,t,n,r){var i;return n(e,function(e,n,o){if(t(e,n,o))return i=r?n:e,!1}),i}e.exports=n},function(e,t){function n(e,t,n){for(var r=e.length,i=n?r:-1;n?i--:++iSorry, we couldn't find anything for "+e.escapeExpression((o=null!=(o=n.value||(null!=t?t.value:t))?o:n.helperMissing,"function"==typeof o?o.call(null!=t?t:e.nullContext||{},{name:"value",hash:{},data:i}):o))+".

\n"},3:function(e,t,n,r,i,o,a){var u;return null!=(u=n.each.call(null!=t?t:e.nullContext||{},null!=t?t.levels:t,{name:"each",hash:{},fn:e.program(4,i,0,o,a),inverse:e.noop,data:i}))?u:""},4:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{};return'

'+e.escapeExpression((s=null!=(s=n.name||(null!=t?t.name:t))?s:n.helperMissing,"function"==typeof s?s.call(l,{name:"name",hash:{},data:i}):s))+"

\n"+(null!=(u=n.each.call(l,null!=t?t.results:t,{name:"each",hash:{},fn:e.program(5,i,0,o,a),inverse:e.noop,data:i}))?u:"")},5:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{},c=n.helperMissing,f="function";return'
\n

\n '+(null!=(s=null!=(s=n.match||(null!=t?t.match:t))?s:c,u=typeof s===f?s.call(l,{name:"match",hash:{},data:i}):s)?u:"")+'\n

\n
    \n'+(null!=(u=n.each.call(l,null!=t?t.functions:t,{name:"each",hash:{},fn:e.program(6,i,0,o,a),inverse:e.noop,data:i}))?u:"")+'
\n
    \n'+(null!=(u=n.each.call(l,null!=t?t.macros:t,{name:"each",hash:{},fn:e.program(6,i,0,o,a),inverse:e.noop,data:i}))?u:"")+'
\n
    \n'+(null!=(u=n.each.call(l,null!=t?t.callbacks:t,{name:"each",hash:{},fn:e.program(8,i,0,o,a),inverse:e.noop,data:i}))?u:"")+'
\n
    \n'+(null!=(u=n.each.call(l,null!=t?t.types:t,{name:"each",hash:{},fn:e.program(10,i,0,o,a),inverse:e.noop,data:i}))?u:"")+"
\n
\n"},6:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{},c=n.helperMissing,f="function";return'
  • '+(null!=(s=null!=(s=n.match||(null!=t?t.match:t))?s:c,u=typeof s===f?s.call(l,{name:"match",hash:{},data:i}):s)?u:"")+"
    • \n"},8:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{},c=n.helperMissing,f="function";return'
  • '+(null!=(s=null!=(s=n.match||(null!=t?t.match:t))?s:c,u=typeof s===f?s.call(l,{name:"match",hash:{},data:i}):s)?u:"")+" (callback)
    • \n"},10:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{},c=n.helperMissing,f="function";return'
  • '+(null!=(s=null!=(s=n.match||(null!=t?t.match:t))?s:c,u=typeof s===f?s.call(l,{name:"match",hash:{},data:i}):s)?u:"")+" (type)
    • \n"},compiler:[7,">= 4.0.0"],main:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{};return"

    Search Results for "+e.escapeExpression((s=null!=(s=n.value||(null!=t?t.value:t))?s:n.helperMissing,"function"==typeof s?s.call(l,{name:"value",hash:{},data:i}):s))+"

    \n\n"+(null!=(u=n["if"].call(l,null!=t?t.empty:t,{name:"if",hash:{},fn:e.program(1,i,0,o,a),inverse:e.program(3,i,0,o,a),data:i}))?u:"")},useData:!0,useDepths:!0})},function(e,t,n){e.exports=n(63)["default"]},function(e,t,n){"use strict";function r(e){return e&&e.__esModule?e:{"default":e}}function i(e){if(e&&e.__esModule)return e;var t={};if(null!=e)for(var n in e)Object.prototype.hasOwnProperty.call(e,n)&&(t[n]=e[n]);return t["default"]=e,t}function o(){var e=new u.HandlebarsEnvironment;return d.extend(e,u),e.SafeString=l["default"],e.Exception=f["default"],e.Utils=d,e.escapeExpression=d.escapeExpression,e.VM=g,e.template=function(t){return g.template(t,e)},e}t.__esModule=!0;var a=n(64),u=i(a),s=n(78),l=r(s),c=n(66),f=r(c),p=n(65),d=i(p),h=n(79),g=i(h),v=n(80),m=r(v),y=o();y.create=o,m["default"](y),y["default"]=y,t["default"]=y,e.exports=t["default"]},function(e,t,n){"use strict";function r(e){return e&&e.__esModule?e:{"default":e}}function i(e,t,n){this.helpers=e||{},this.partials=t||{},this.decorators=n||{},s.registerDefaultHelpers(this),l.registerDefaultDecorators(this)}t.__esModule=!0,t.HandlebarsEnvironment=i;var o=n(65),a=n(66),u=r(a),s=n(67),l=n(75),c=n(77),f=r(c),p="4.0.10";t.VERSION=p;var d=7;t.COMPILER_REVISION=d;var h={1:"<= 1.0.rc.2",2:"== 1.0.0-rc.3",3:"== 1.0.0-rc.4",4:"== 1.x.x",5:"== 2.0.0-alpha.x",6:">= 2.0.0-beta.1",7:">= 4.0.0"};t.REVISION_CHANGES=h;var g="[object Object]";i.prototype={constructor:i,logger:f["default"],log:f["default"].log,registerHelper:function(e,t){if(o.toString.call(e)===g){if(t)throw new u["default"]("Arg not supported with multiple helpers");o.extend(this.helpers,e)}else this.helpers[e]=t},unregisterHelper:function(e){delete this.helpers[e]},registerPartial:function(e,t){if(o.toString.call(e)===g)o.extend(this.partials,e);else{if("undefined"==typeof t)throw new u["default"]('Attempting to register a partial called "'+e+'" as undefined');this.partials[e]=t}},unregisterPartial:function(e){delete this.partials[e]},registerDecorator:function(e,t){if(o.toString.call(e)===g){if(t)throw new u["default"]("Arg not supported with multiple decorators");o.extend(this.decorators,e)}else this.decorators[e]=t},unregisterDecorator:function(e){delete this.decorators[e]}};var v=f["default"].log;t.log=v,t.createFrame=o.createFrame,t.logger=f["default"]},function(e,t){"use strict";function n(e){return c[e]}function r(e){for(var t=1;t":">",'"':""","'":"'","`":"`","=":"="},f=/[&<>"'`=]/g,p=/[&<>"'`=]/,d=Object.prototype.toString;t.toString=d;var h=function(e){return"function"==typeof e};h(/x/)&&(t.isFunction=h=function(e){return"function"==typeof e&&"[object Function]"===d.call(e)}),t.isFunction=h;var g=Array.isArray||function(e){return!(!e||"object"!=typeof e)&&"[object Array]"===d.call(e)};t.isArray=g},function(e,t){"use strict";function n(e,t){var i=t&&t.loc,o=void 0,a=void 0;i&&(o=i.start.line,a=i.start.column,e+=" - "+o+":"+a);for(var u=Error.prototype.constructor.call(this,e),s=0;s0?(n.ids&&(n.ids=[n.name]),e.helpers.each(t,n)):i(this);if(n.data&&n.ids){var a=r.createFrame(n.data);a.contextPath=r.appendContextPath(n.data.contextPath,n.name),n={data:a}}return o(t,n)})},e.exports=t["default"]},function(e,t,n){"use strict";function r(e){return e&&e.__esModule?e:{"default":e}}t.__esModule=!0;var i=n(65),o=n(66),a=r(o);t["default"]=function(e){e.registerHelper("each",function(e,t){function n(t,n,o){l&&(l.key=t,l.index=n,l.first=0===n,l.last=!!o,c&&(l.contextPath=c+t)),s+=r(e[t],{data:l,blockParams:i.blockParams([e[t],t],[c+t,null])})}if(!t)throw new a["default"]("Must pass iterator to #each");var r=t.fn,o=t.inverse,u=0,s="",l=void 0,c=void 0;if(t.data&&t.ids&&(c=i.appendContextPath(t.data.contextPath,t.ids[0])+"."),i.isFunction(e)&&(e=e.call(this)),t.data&&(l=i.createFrame(t.data)),e&&"object"==typeof e)if(i.isArray(e))for(var f=e.length;u=0?t:parseInt(e,10)}return e},log:function(e){if(e=i.lookupLevel(e),"undefined"!=typeof console&&i.lookupLevel(i.level)<=e){var t=i.methodMap[e];console[t]||(t="log");for(var n=arguments.length,r=Array(n>1?n-1:0),o=1;o\n\n\n '+p(f(null!=(l=u[0][0])?l.title:l,t))+'\n\n
      \n
    • \n Top\n
      • \n\n'+(null!=(l=r(n(84)).call(c,null!=(l=u[0][0])?l.headers:l,{name:"isArray",hash:{},fn:e.program(6,a,0,u,s),inverse:e.program(9,a,0,u,s),data:a,blockParams:u}))?l:"")+"
    \n \n"},2:function(e,t,n,r,i,o){var a;return'
  • '+e.escapeExpression(e.lambda(null!=(a=o[1][0])?a.group:a,t))+"
    • \n"},4:function(e,t,n,r,i){return"clicked open"},6:function(e,t,n,r,i,o){var a;return null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[1][0])?a.headers:a,{name:"each",hash:{},fn:e.program(7,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:""},7:function(e,t,n,r,i,o){var a,u=e.lambda,s=e.escapeExpression;return'
  • \n '+s(u(null!=t?t.id:t,t))+"\n
    • \n"},9:function(e,t,i,o,a,u){var s,l=null!=t?t:e.nullContext||{};return(null!=(s=r(n(85)).call(l,u[1][0],{name:"showSummary",hash:{},fn:e.program(10,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")+(null!=(s=i["if"].call(l,null!=(s=u[1][0])?s.types:s,{name:"if",hash:{},fn:e.program(12,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")+(null!=(s=i["if"].call(l,null!=(s=u[1][0])?s.functions:s,{name:"if",hash:{},fn:e.program(15,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")+(null!=(s=i["if"].call(l,null!=(s=u[1][0])?s.macros:s,{name:"if",hash:{},fn:e.program(17,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")+(null!=(s=i["if"].call(l,null!=(s=u[1][0])?s.callbacks:s,{name:"if",hash:{},fn:e.program(20,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")},10:function(e,t,n,r,i,o){var a;return'
  • \n Summary\n
    • \n'},12:function(e,t,n,r,i,o){var a;return'
  • \n Types\n
      \n'+(null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[2][0])?a.types:a,{name:"each",hash:{},fn:e.program(13,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:"")+"
    \n
    • \n"},13:function(e,t,n,r,i,o){var a,u=e.lambda,s=e.escapeExpression;return'
  • \n '+s(u(null!=t?t.id:t,t))+"\n
    • \n"},15:function(e,t,n,r,i,o){var a;return'
  • \n Functions\n
      \n'+(null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[2][0])?a.functions:a,{name:"each",hash:{},fn:e.program(13,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:"")+"
    \n
    • \n"},17:function(e,t,n,r,i,o){var a;return'
  • \n Macros\n
      \n'+(null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[2][0])?a.macros:a,{name:"each",hash:{},fn:e.program(18,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:"")+"
    \n
    • \n"},18:function(e,t,n,r,i,o){var a,u=e.lambda,s=e.escapeExpression;return'
  • \n '+s(u(null!=t?t.id:t,t))+"\n
    • \n"},20:function(e,t,n,r,i,o){var a;return'
  • \n Callbacks\n
      \n'+(null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[2][0])?a.callbacks:a,{name:"each",hash:{},fn:e.program(18,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:"")+"
    \n
    • \n"},compiler:[7,">= 4.0.0"],main:function(e,t,n,r,i,o,a){var u;return'
      \n'+(null!=(u=n.each.call(null!=t?t:e.nullContext||{},null!=t?t.nodes:t,{name:"each",hash:{},fn:e.program(1,i,2,o,a),inverse:e.noop,data:i,blockParams:o}))?u:"")+"
    \n"},useData:!0,useDepths:!0,useBlockParams:!0})},function(e,t){"use strict";Object.defineProperty(t,"__esModule",{value:!0}),t["default"]=function(e,t,n){var r=t||"";if(e.group!==r)return e.group=r,n.fn(this)},e.exports=t["default"]},function(e,t){"use strict";Object.defineProperty(t,"__esModule",{value:!0}),t["default"]=function(e,t){var n=window.location.pathname.split("/");return e+=".html",e===n[n.length-1]?t.fn(this):t.inverse(this)},e.exports=t["default"]},function(e,t){"use strict";Object.defineProperty(t,"__esModule",{value:!0}),t["default"]=function(e,t){return Array.isArray(e)?t.fn(this):t.inverse(this)},e.exports=t["default"]},function(e,t){"use strict";Object.defineProperty(t,"__esModule",{value:!0}),t["default"]=function(e,t){if(e.types||e.functions||e.macros||e.callbacks)return t.fn(this)},e.exports=t["default"]},function(e,t,n){"use strict";function r(){p.addClass(y).removeClass(g).removeClass(v),_=setTimeout(function(){return p.addClass(m).removeClass(y)},h)}function i(){p.addClass(v).removeClass(m).removeClass(y),_=setTimeout(function(){return p.addClass(g).removeClass(v)},h)}function o(){var e=p.attr("class")||"";clearTimeout(_),e.includes(m)||e.includes(y)?i():r()}function a(){p.removeClass(b),p.addClass(window.innerWidth>d?g:m)}function u(){a();var e=window.innerWidth;(0,c["default"])(window).resize((0,f.throttle)(function(){e!==window.innerWidth&&(e=window.innerWidth,a())},100)),(0,c["default"])(".sidebar-toggle").click(function(){o()})}var s=n(1)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.initialize=u;var l=n(2),c=s(l),f=n(87),p=(0,c["default"])("body"),d=768,h=300,g="sidebar-opened",v="sidebar-opening",m="sidebar-closed",y="sidebar-closing",b=[g,v,m,y].join(" "),_=void 0;t.breakpoint=d,t.closeSidebar=r},function(e,t,n){var r;(function(e,i){(function(){function o(e,t){if(e!==t){var n=null===e,r=e===T,i=e===e,o=null===t,a=t===T,u=t===t;if(e>t&&!o||!i||n&&!a&&u||r&&u)return 1;if(e-1;);return n}function f(e,t){for(var n=e.length;n--&&t.indexOf(e.charAt(n))>-1;);return n}function p(e,t){return o(e.criteria,t.criteria)||e.index-t.index}function d(e,t,n){for(var r=-1,i=e.criteria,a=t.criteria,u=i.length,s=n.length;++r=s)return l;var c=n[r];return l*("asc"===c||c===!0?1:-1)}}return e.index-t.index}function h(e){return Ke[e]}function g(e){return Ge[e]}function v(e,t,n){return t?e=Ze[e]:n&&(e=Qe[e]),"\\"+e}function m(e){return"\\"+Qe[e]}function y(e,t,n){for(var r=e.length,i=t+(n?0:-1);n?i--:++i=9&&e<=13||32==e||160==e||5760==e||6158==e||e>=8192&&(e<=8202||8232==e||8233==e||8239==e||8287==e||12288==e||65279==e)}function x(e,t){for(var n=-1,r=e.length,i=-1,o=[];++n=z?gn(t):null,l=t.length;s&&(o=Qe,a=!1,t=s);e:for(;++ii?0:i+n),r=r===T||r>i?i:+r||0,r<0&&(r+=i),i=n>r?0:r>>>0,n>>>=0;ni?0:i+t),n=n===T||n>i?i:+n||0,n<0&&(n+=i),i=t>n?0:n-t>>>0,t>>>=0;for(var o=$o(i);++r=z,s=a?gn():null,l=[];s?(r=Qe,o=!1):(a=!1,s=t?[]:l);e:for(;++n>>1,a=e[o];(n?a<=t:a2?n[i-2]:T,a=i>2?n[2]:T,u=i>1?n[i-1]:T;for("function"==typeof o?(o=an(o,u,5),i-=2):(o="function"==typeof u?u:T,i-=o?1:0),a&&Jn(n[0],n[1],a)&&(o=i<3?T:o,i=1);++r-1?n[o]:T}return At(n,r,e)}}function wn(e){return function(t,n,r){return t&&t.length?(n=$n(n,r,3),a(t,n,e)):-1}}function En(e){return function(t,n,r){return n=$n(n,r,3),At(t,n,e,!0)}}function kn(e){return function(){for(var t,n=arguments.length,i=e?n:-1,o=0,a=$o(n);e?i--:++i=z)return t.plant(r).value();for(var i=0,o=n?a[i].apply(this,e):r;++i=t||!_a(t))return"";var i=t-r;return n=null==n?" ":n+"",mo(n,va(i/n.length)).slice(0,i)}function Rn(e,t,n,r){function i(){for(var t=-1,u=arguments.length,s=-1,l=r.length,c=$o(l+u);++ss))return!1;for(;++u-1&&e%1==0&&e-1&&e%1==0&&e<=Ma}function rr(e){return e===e&&!Ri(e)}function ir(e,t){var n=e[1],r=t[1],i=n|r,o=i-1;)da.call(t,o,1);return t}function Sr(e,t,n){var r=[];if(!e||!e.length)return r;var i=-1,o=[],a=e.length;for(t=$n(t,n,3);++i-1:!!i&&Fn(e,t,n)>-1}function ei(e,t,n){var r=Su(e)?st:qt;return t=$n(t,n,3),r(e,t)}function ti(e,t){return ei(e,Ro(t))}function ni(e,t,n){var r=Su(e)?ut:Tt;return t=$n(t,n,3),r(e,function(e,n,r){return!t(e,n,r)})}function ri(e,t,n){if(n?Jn(e,t,n):null==t){e=cr(e);var r=e.length;return r>0?e[Kt(0,r-1)]:T}var i=-1,o=Gi(e),r=o.length,a=r-1;for(t=Ea(t<0?0:+t||0,r);++i0&&(n=t.apply(this,arguments)),e<=1&&(t=T),n}}function di(e,t,n){function r(){d&&ua(d),l&&ua(l),g=0,l=d=h=T}function i(t,n){n&&ua(n),l=d=h=T,t&&(g=gu(),c=e.apply(p,s),d||l||(s=p=T))}function o(){var e=t-(gu()-f);e<=0||e>t?i(h,l):d=pa(o,e)}function a(){i(m,d)}function u(){if(s=arguments,f=gu(),p=this,h=m&&(d||!y),v===!1)var n=y&&!d;else{l||y||(g=f);var r=v-(f-g),i=r<=0||r>v;i?(l&&(l=ua(l)),g=f,c=e.apply(p,s)):l||(l=pa(a,r))}return i&&d?d=ua(d):d||t===v||(d=pa(o,t)),n&&(i=!0,c=e.apply(p,s)),!i||d||l||(s=p=T),c}var s,l,c,f,p,d,h,g=0,v=!1,m=!0;if("function"!=typeof e)throw new Zo(U);if(t=t<0?0:+t||0,n===!0){var y=!0;m=!1}else Ri(n)&&(y=!!n.leading,v="maxWait"in n&&wa(+n.maxWait||0,t),m="trailing"in n?!!n.trailing:m);return u.cancel=r,u}function hi(e,t){if("function"!=typeof e||t&&"function"!=typeof t)throw new Zo(U);var n=function(){var r=arguments,i=t?t.apply(this,r):r[0],o=n.cache;if(o.has(i))return o.get(i);var a=e.apply(this,r);return n.cache=o.set(i,a),a};return n.cache=new hi.Cache,n}function gi(e){if("function"!=typeof e)throw new Zo(U);return function(){return!e.apply(this,arguments)}}function vi(e){return pi(2,e)}function mi(e,t){if("function"!=typeof e)throw new Zo(U);return t=wa(t===T?e.length-1:+t||0,0),function(){for(var n=arguments,r=-1,i=wa(n.length-t,0),o=$o(i);++rt}function ki(e,t){return e>=t}function Ci(e){return b(e)&&Qn(e)&&ta.call(e,"callee")&&!ca.call(e,"callee")}function Ni(e){return e===!0||e===!1||b(e)&&ra.call(e)==X}function Ti(e){return b(e)&&ra.call(e)==Z}function Ai(e){return!!e&&1===e.nodeType&&b(e)&&!Bi(e)}function Si(e){return null==e||(Qn(e)&&(Su(e)||zi(e)||Ci(e)||b(e)&&Di(e.splice))?!e.length:!Bu(e).length)}function ji(e,t,n,r){n="function"==typeof n?an(n,r,3):T;var i=n?n(e,t):T;return i===T?Pt(e,t,n):!!i}function Oi(e){return b(e)&&"string"==typeof e.message&&ra.call(e)==Q}function Mi(e){return"number"==typeof e&&_a(e)}function Di(e){return Ri(e)&&ra.call(e)==Y}function Ri(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}function Pi(e,t,n,r){return n="function"==typeof n?an(n,r,3):T,It(e,Wn(t),n)}function Li(e){return Hi(e)&&e!=+e}function Ii(e){return null!=e&&(Di(e)?oa.test(ea.call(e)):b(e)&&Le.test(e))}function qi(e){return null===e}function Hi(e){return"number"==typeof e||b(e)&&ra.call(e)==ee}function Bi(e){var t;if(!b(e)||ra.call(e)!=te||Ci(e)||!ta.call(e,"constructor")&&(t=e.constructor,"function"==typeof t&&!(t instanceof t)))return!1;var n;return jt(e,function(e,t){n=t}),n===T||ta.call(e,n)}function $i(e){return Ri(e)&&ra.call(e)==ne}function zi(e){return"string"==typeof e||b(e)&&ra.call(e)==ie}function Fi(e){return b(e)&&nr(e.length)&&!!We[ra.call(e)]}function Wi(e){return e===T}function Ui(e,t){return e0;++r=Ea(t,n)&&e=0&&e.indexOf(t,n)==n}function po(e){return e=l(e),e&&we.test(e)?e.replace(_e,g):e}function ho(e){return e=l(e),e&&je.test(e)?e.replace(Se,v):e||"(?:)"}function go(e,t,n){e=l(e),t=+t;var r=e.length;if(r>=t||!_a(t))return e;var i=(t-r)/2,o=ya(i),a=va(i);return n=Dn("",a,n),n.slice(0,o)+e+n}function vo(e,t,n){return(n?Jn(e,t,n):null==t)?t=0:t&&(t=+t),e=_o(e),Ca(e,t||(Pe.test(e)?16:10))}function mo(e,t){var n="";if(e=l(e),t=+t,t<1||!e||!_a(t))return n;do t%2&&(n+=e),t=ya(t/2),e+=e;while(t);return n}function yo(e,t,n){return e=l(e),n=null==n?0:Ea(n<0?0:+n||0,e.length),e.lastIndexOf(t,n)==n}function bo(e,n,r){var i=t.templateSettings;r&&Jn(e,n,r)&&(n=r=T),e=l(e),n=vt(mt({},r||n),i,gt);var o,a,u=vt(mt({},n.imports),i.imports,gt),s=Bu(u),c=en(u,s),f=0,p=n.interpolate||He,d="__p += '",h=Vo((n.escape||He).source+"|"+p.source+"|"+(p===Ce?De:He).source+"|"+(n.evaluate||He).source+"|$","g"),g="//# sourceURL="+("sourceURL"in n?n.sourceURL:"lodash.templateSources["+ ++Fe+"]")+"\n";e.replace(h,function(t,n,r,i,u,s){return r||(r=i),d+=e.slice(f,s).replace(Be,m),n&&(o=!0,d+="' +\n__e("+n+") +\n'"),u&&(a=!0,d+="';\n"+u+";\n__p += '"),r&&(d+="' +\n((__t = ("+r+")) == null ? '' : __t) +\n'"),f=s+t.length,t}),d+="';\n";var v=n.variable;v||(d="with (obj) {\n"+d+"\n}\n"),d=(a?d.replace(ve,""):d).replace(me,"$1").replace(ye,"$1;"),d="function("+(v||"obj")+") {\n"+(v?"":"obj || (obj = {});\n")+"var __t, __p = ''"+(o?", __e = _.escape":"")+(a?", __j = Array.prototype.join;\nfunction print() { __p += __j.call(arguments, '') }\n":";\n")+d+"return __p\n}";var y=Qu(function(){return Wo(s,g+"return "+d).apply(T,c)});if(y.source=d,Oi(y))throw y;return y}function _o(e,t,n){var r=e;return(e=l(e))?(n?Jn(r,t,n):null==t)?e.slice(E(e),k(e)+1):(t+="",e.slice(c(e,t),f(e,t)+1)):e}function xo(e,t,n){var r=e;return e=l(e),e?(n?Jn(r,t,n):null==t)?e.slice(E(e)):e.slice(c(e,t+"")):e}function wo(e,t,n){var r=e;return e=l(e),e?(n?Jn(r,t,n):null==t)?e.slice(0,k(e)+1):e.slice(0,f(e,t+"")+1):e}function Eo(e,t,n){n&&Jn(e,t,n)&&(t=T);var r=q,i=H;if(null!=t)if(Ri(t)){var o="separator"in t?t.separator:o;r="length"in t?+t.length||0:r,i="omission"in t?l(t.omission):i}else r=+t||0;if(e=l(e),r>=e.length)return e;var a=r-i.length;if(a<1)return i;var u=e.slice(0,a);if(null==o)return u+i;if($i(o)){if(e.slice(a).search(o)){var s,c,f=e.slice(0,a);for(o.global||(o=Vo(o.source,(Re.exec(o)||"")+"g")),o.lastIndex=0;s=o.exec(f);)c=s.index;u=u.slice(0,null==c?a:c)}}else if(e.indexOf(o,a)!=a){var p=u.lastIndexOf(o);p>-1&&(u=u.slice(0,p))}return u+i}function ko(e){return e=l(e),e&&xe.test(e)?e.replace(be,C):e}function Co(e,t,n){return n&&Jn(e,t,n)&&(t=T),e=l(e),e.match(t||$e)||[]}function No(e,t,n){return n&&Jn(e,t,n)&&(t=T),b(e)?So(e):_t(e,t)}function To(e){return function(){return e}}function Ao(e){return e}function So(e){return Ht(xt(e,!0))}function jo(e,t){return Bt(e,xt(t,!0))}function Oo(e,t,n){if(null==n){var r=Ri(t),i=r?Bu(t):T,o=i&&i.length?Dt(t,i):T;(o?o.length:r)||(o=!1,n=t,t=e,e=this)}o||(o=Dt(t,Bu(t)));var a=!0,u=-1,s=Di(e),l=o.length;n===!1?a=!1:Ri(n)&&"chain"in n&&(a=n.chain);for(;++u>>1,Ma=9007199254740991,Da=ga&&new ga,Ra={};t.support={};t.templateSettings={escape:Ee,evaluate:ke,interpolate:Ce,variable:"",imports:{_:t}};var Pa=function(){function e(){}return function(t){if(Ri(t)){e.prototype=t;var n=new e;e.prototype=T}return n||{}}}(),La=pn(Ot),Ia=pn(Mt,!0),qa=dn(),Ha=dn(!0),Ba=Da?function(e,t){return Da.set(e,t),e}:Ao,$a=Da?function(e){return Da.get(e)}:Do,za=Ft("length"),Fa=function(){var e=0,t=0;return function(n,r){var i=gu(),o=$-(i-t);if(t=i,o>0){if(++e>=B)return n}else e=0;return Ba(n,r)}}(),Wa=mi(function(e,t){return b(e)&&Qn(e)?Et(e,St(t,!1,!0)):[]}),Ua=wn(),Ka=wn(!0),Ga=mi(function(e){for(var t=e.length,n=t,r=$o(f),i=Fn(),o=i==u,a=[];n--;){var s=e[n]=Qn(s=e[n])?s:[];r[n]=o&&s.length>=120?gn(n&&s):null}var l=e[0],c=-1,f=l?l.length:0,p=r[0];e:for(;++c2?e[t-2]:T,r=t>1?e[t-1]:T;return t>2&&"function"==typeof n?t-=2:(n=t>1&&"function"==typeof r?(--t,r):T,r=T),e.length=t,qr(e,n,r)}),tu=mi(function(e){return e=St(e),this.thru(function(t){return Je(Su(t)?t:[fr(t)],e)})}),nu=mi(function(e,t){return yt(e,St(t))}),ru=cn(function(e,t,n){ta.call(e,n)?++e[n]:e[n]=1}),iu=xn(La),ou=xn(Ia,!0),au=Cn(tt,La),uu=Cn(nt,Ia),su=cn(function(e,t,n){ta.call(e,n)?e[n].push(t):e[n]=[t]}),lu=cn(function(e,t,n){e[n]=t}),cu=mi(function(e,t,n){var r=-1,i="function"==typeof t,o=er(t),a=Qn(e)?$o(e.length):[];return La(e,function(e){var u=i?t:o&&null!=e?e[t]:T;a[++r]=u?u.apply(e,n):Zn(e,t,n)}),a}),fu=cn(function(e,t,n){e[n?0:1].push(t)},function(){return[[],[]]}),pu=On(ct,La),du=On(ft,Ia),hu=mi(function(e,t){if(null==e)return[];var n=t[2];return n&&Jn(t[0],t[1],n)&&(t.length=1),Qt(e,St(t),[])}),gu=ka||function(){return(new zo).getTime()},vu=mi(function(e,t,n){var r=S;if(n.length){var i=x(n,vu.placeholder);r|=R}return In(e,r,t,n,i)}),mu=mi(function(e,t){t=t.length?St(t):Zi(e);for(var n=-1,r=t.length;++n0||t<0)?new i(n):(e<0?n=n.takeRight(-e):e&&(n=n.drop(e)),t!==T&&(t=+t||0,n=t<0?n.dropRight(-t):n.take(t-e)),n)},i.prototype.takeRightWhile=function(e,t){return this.reverse().takeWhile(e,t).reverse()},i.prototype.toArray=function(){return this.take(Aa)},Ot(i.prototype,function(e,n){var o=/^(?:filter|map|reject)|While$/.test(n),a=/^(?:first|last)$/.test(n),u=t[a?"take"+("last"==n?"Right":""):n];u&&(t.prototype[n]=function(){var t=a?[1]:arguments,n=this.__chain__,s=this.__wrapped__,l=!!this.__actions__.length,c=s instanceof i,f=t[0],p=c||Su(s);p&&o&&"function"==typeof f&&1!=f.length&&(c=p=!1);var d=function(e){return a&&n?u(e,1)[0]:u.apply(T,lt([e],t))},h={func:Fr,args:[d],thisArg:T},g=c&&!l;if(a&&!n)return g?(s=s.clone(),s.__actions__.push(h),e.call(s)):u.call(T,this.value())[0];if(!a&&p){s=g?s:new i(this);var v=e.apply(s,t);return v.__actions__.push(h),new r(v,n)}return this.thru(d)})}),tt(["join","pop","push","replace","shift","sort","splice","split","unshift"],function(e){var n=(/^(?:replace|split)$/.test(e)?Jo:Qo)[e],r=/^(?:push|sort|unshift)$/.test(e)?"tap":"thru",i=/^(?:join|pop|replace|shift)$/.test(e);t.prototype[e]=function(){var e=arguments;return i&&!this.__chain__?n.apply(this.value(),e):this[r](function(t){return n.apply(t,e)})}}),Ot(i.prototype,function(e,n){var r=t[n];if(r){var i=r.name,o=Ra[i]||(Ra[i]=[]);o.push({name:n,func:r})}}),Ra[Mn(T,j).name]=[{name:"wrapper",func:T}],i.prototype.clone=_,i.prototype.reverse=J,i.prototype.value=re,t.prototype.chain=Wr,t.prototype.commit=Ur,t.prototype.concat=tu,t.prototype.plant=Kr,t.prototype.reverse=Gr,t.prototype.toString=Vr,t.prototype.run=t.prototype.toJSON=t.prototype.valueOf=t.prototype.value=Xr,t.prototype.collect=t.prototype.map,t.prototype.head=t.prototype.first,t.prototype.select=t.prototype.filter,t.prototype.tail=t.prototype.rest,t}var T,A="3.10.1",S=1,j=2,O=4,M=8,D=16,R=32,P=64,L=128,I=256,q=30,H="...",B=150,$=16,z=200,F=1,W=2,U="Expected a function",K="__lodash_placeholder__",G="[object Arguments]",V="[object Array]",X="[object Boolean]",Z="[object Date]",Q="[object Error]",Y="[object Function]",J="[object Map]",ee="[object Number]",te="[object Object]",ne="[object RegExp]",re="[object Set]",ie="[object String]",oe="[object WeakMap]",ae="[object ArrayBuffer]",ue="[object Float32Array]",se="[object Float64Array]",le="[object Int8Array]",ce="[object Int16Array]",fe="[object Int32Array]",pe="[object Uint8Array]",de="[object Uint8ClampedArray]",he="[object Uint16Array]",ge="[object Uint32Array]",ve=/\b__p \+= '';/g,me=/\b(__p \+=) '' \+/g,ye=/(__e\(.*?\)|\b__t\)) \+\n'';/g,be=/&(?:amp|lt|gt|quot|#39|#96);/g,_e=/[&<>"'`]/g,xe=RegExp(be.source),we=RegExp(_e.source),Ee=/<%-([\s\S]+?)%>/g,ke=/<%([\s\S]+?)%>/g,Ce=/<%=([\s\S]+?)%>/g,Ne=/\.|\[(?:[^[\]]*|(["'])(?:(?!\1)[^\n\\]|\\.)*?\1)\]/,Te=/^\w*$/,Ae=/[^.[\]]+|\[(?:(-?\d+(?:\.\d+)?)|(["'])((?:(?!\2)[^\n\\]|\\.)*?)\2)\]/g,Se=/^[:!,]|[\\^$.*+?()[\]{}|\/]|(^[0-9a-fA-Fnrtuvx])|([\n\r\u2028\u2029])/g,je=RegExp(Se.source),Oe=/[\u0300-\u036f\ufe20-\ufe23]/g,Me=/\\(\\)?/g,De=/\$\{([^\\}]*(?:\\.[^\\}]*)*)\}/g,Re=/\w*$/,Pe=/^0[xX]/,Le=/^\[object .+?Constructor\]$/,Ie=/^\d+$/,qe=/[\xc0-\xd6\xd8-\xde\xdf-\xf6\xf8-\xff]/g,He=/($^)/,Be=/['\n\r\u2028\u2029\\]/g,$e=function(){var e="[A-Z\\xc0-\\xd6\\xd8-\\xde]",t="[a-z\\xdf-\\xf6\\xf8-\\xff]+";return RegExp(e+"+(?="+e+t+")|"+e+"?"+t+"|"+e+"+|[0-9]+","g")}(),ze=["Array","ArrayBuffer","Date","Error","Float32Array","Float64Array","Function","Int8Array","Int16Array","Int32Array","Math","Number","Object","RegExp","Set","String","_","clearTimeout","isFinite","parseFloat","parseInt","setTimeout","TypeError","Uint8Array","Uint8ClampedArray","Uint16Array","Uint32Array","WeakMap"],Fe=-1,We={};We[ue]=We[se]=We[le]=We[ce]=We[fe]=We[pe]=We[de]=We[he]=We[ge]=!0,We[G]=We[V]=We[ae]=We[X]=We[Z]=We[Q]=We[Y]=We[J]=We[ee]=We[te]=We[ne]=We[re]=We[ie]=We[oe]=!1;var Ue={};Ue[G]=Ue[V]=Ue[ae]=Ue[X]=Ue[Z]=Ue[ue]=Ue[se]=Ue[le]=Ue[ce]=Ue[fe]=Ue[ee]=Ue[te]=Ue[ne]=Ue[ie]=Ue[pe]=Ue[de]=Ue[he]=Ue[ge]=!0,Ue[Q]=Ue[Y]=Ue[J]=Ue[re]=Ue[oe]=!1;var Ke={"À":"A","Á":"A","Â":"A","Ã":"A","Ä":"A","Å":"A","à":"a","á":"a","â":"a","ã":"a","ä":"a","å":"a","Ç":"C","ç":"c","Ð":"D","ð":"d","È":"E","É":"E","Ê":"E","Ë":"E","è":"e","é":"e","ê":"e","ë":"e","Ì":"I","Í":"I","Î":"I","Ï":"I","ì":"i","í":"i","î":"i","ï":"i","Ñ":"N","ñ":"n","Ò":"O","Ó":"O","Ô":"O","Õ":"O","Ö":"O","Ø":"O","ò":"o","ó":"o","ô":"o","õ":"o","ö":"o","ø":"o","Ù":"U","Ú":"U","Û":"U","Ü":"U","ù":"u","ú":"u","û":"u","ü":"u","Ý":"Y","ý":"y","ÿ":"y","Æ":"Ae","æ":"ae","Þ":"Th","þ":"th","ß":"ss"},Ge={"&":"&","<":"<",">":">",'"':""","'":"'","`":"`"},Ve={"&":"&","<":"<",">":">",""":'"',"'":"'","`":"`"},Xe={"function":!0,object:!0},Ze={0:"x30",1:"x31",2:"x32",3:"x33",4:"x34",5:"x35",6:"x36",7:"x37",8:"x38",9:"x39",A:"x41",B:"x42",C:"x43",D:"x44",E:"x45",F:"x46",a:"x61",b:"x62",c:"x63",d:"x64",e:"x65",f:"x66",n:"x6e",r:"x72",t:"x74",u:"x75",v:"x76",x:"x78"},Qe={"\\":"\\","'":"'","\n":"n","\r":"r","\u2028":"u2028","\u2029":"u2029"},Ye=Xe[typeof t]&&t&&!t.nodeType&&t,Je=Xe[typeof e]&&e&&!e.nodeType&&e,et=Ye&&Je&&"object"==typeof i&&i&&i.Object&&i,tt=Xe[typeof self]&&self&&self.Object&&self,nt=Xe[typeof window]&&window&&window.Object&&window,rt=(Je&&Je.exports===Ye&&Ye,et||nt!==(this&&this.window)&&nt||tt||this),it=N();rt._=it,r=function(){return it}.call(t,n,t,e),!(r!==T&&(e.exports=r))}).call(this)}).call(t,n(88)(e),function(){return this}())},function(e,t){e.exports=function(e){return e.webpackPolyfill||(e.deprecate=function(){},e.paths=[],e.children=[],e.webpackPolyfill=1),e}},function(e,t,n){"use strict";function r(){f.addClass(p);try{localStorage.setItem(p,!0)}catch(e){}}function i(){f.removeClass(p);try{localStorage.removeItem(p)}catch(e){}}function o(){try{localStorage.getItem(p)&&r()}catch(e){}}function a(){f.hasClass(p)?i():r()}function u(){o(),d.click(function(){a()})}var s=n(1)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.initialize=u;var l=n(2),c=s(l),f=(0,c["default"])("body"),p="night-mode",d=(0,c["default"])(".night-mode-toggle")}]); +//# sourceMappingURL=app-9bd040e5e5.js.map diff --git a/doc/dist/sidebar_items-3a30d3745e.js b/doc/dist/sidebar_items-3a30d3745e.js new file mode 100644 index 0000000..2957579 --- /dev/null +++ b/doc/dist/sidebar_items-3a30d3745e.js @@ -0,0 +1 @@ +sidebarNodes={"extras":[{"id":"api-reference","title":"API Reference","group":"","headers":[{"id":"Modules","anchor":"modules"}]}],"exceptions":[],"modules":[{"id":"Tensorflex","title":"Tensorflex","group":"","functions":[{"id":"append_to_matrix/2","anchor":"append_to_matrix/2"},{"id":"create_matrix/3","anchor":"create_matrix/3"},{"id":"float32_tensor/1","anchor":"float32_tensor/1"},{"id":"float32_tensor/2","anchor":"float32_tensor/2"},{"id":"float32_tensor_alloc/1","anchor":"float32_tensor_alloc/1"},{"id":"float64_tensor/1","anchor":"float64_tensor/1"},{"id":"float64_tensor/2","anchor":"float64_tensor/2"},{"id":"float64_tensor_alloc/1","anchor":"float64_tensor_alloc/1"},{"id":"get_graph_ops/1","anchor":"get_graph_ops/1"},{"id":"int32_tensor/1","anchor":"int32_tensor/1"},{"id":"int32_tensor/2","anchor":"int32_tensor/2"},{"id":"int32_tensor_alloc/1","anchor":"int32_tensor_alloc/1"},{"id":"load_csv_as_matrix/2","anchor":"load_csv_as_matrix/2"},{"id":"load_image_as_tensor/1","anchor":"load_image_as_tensor/1"},{"id":"matrix_pos/3","anchor":"matrix_pos/3"},{"id":"matrix_to_lists/1","anchor":"matrix_to_lists/1"},{"id":"read_graph/1","anchor":"read_graph/1"},{"id":"run_session/5","anchor":"run_session/5"},{"id":"size_of_matrix/1","anchor":"size_of_matrix/1"},{"id":"string_tensor/1","anchor":"string_tensor/1"},{"id":"tensor_datatype/1","anchor":"tensor_datatype/1"}]}],"tasks":[]} \ No newline at end of file diff --git a/doc/fonts/icomoon.eot b/doc/fonts/icomoon.eot new file mode 100644 index 0000000000000000000000000000000000000000..2786a1e5b0e9ee7a593e544f2bc840cf821af61a GIT binary patch literal 3096 zcma)8O>7&-6@Ig`JIh@zx%~f=#Qz`HvSgX0$ka`2ra&DylHtIK5~K)FF|s1LF-2LH zfCwlGqp%&IhsHf5Xwpjqni=_g{O(1V51EhdrqCmzJ5ty{VeJ^N)9f%zb zyZ+nz*YCq<9P;&*?$veB5aeHg`c^J)zi{KHe?AZSZ6f`-)#dKRzDy;%fU^X}jVEJ8bxAkH2!~7@go_4P=_|IVAcsRgmA7_J) zbA#L;v!7-Ep7hyT@Kg90jqP;w8b%DB>=8XSQ~R5a`9Ytr#}tYWcG*AK9r`xiz%c8{ zg#5*0y0Dqh@hcT;Rs32FN7bmeTc=s8-I{GWaJ5vFmXf%5Lw8+x{aiKN#AB{!*FmCD zZ@2+G_tn{^=eXT!uou(GR3vJcrWxrdJd#KxCo-{6$Owdm?hnUPp-{-`Sh^6pWq}We z?7&n!$#t%DBB@mRwg1~XnoMSfv5cq}i`_BI$WQ|;orbrqakU1q&)q=p@au9kxheDp3owTfp?n?OiHhW6NFlRiynDn?{FSlS3EA zo=5+uyyzr{4I<}kK%>&NQ9-W%6e-B@HdUxiRHapZJ7rlwP4bCa5%Ur$82=$)yb4>i_G z>3|)t;Qld*20IF|9eg2@$h$$`p=a=M*`%BFGpuqEdlH8nJF%)4G0nI>FzbTXTaU0N zZVNHNVu`3rU?xGHb%xh!9%ohT>sXFy9z9T0n$=0y_qSNRRe>Lbku2hrG<3pLz#VQV zTu(*4vd`i&b99Vm=jLW*E!FEKSxvv+9I0<|A6I=`*C%vbHhPv#3rQz{IdFFGuHT$A zEX$ZQ{p>X~p1AkJtgdsPivI*wJuFyW)j2Q`hA*8N8e5o~ZO+d(XXke6rE+bWFNUwu z486+(A)>$4Aazh^)EIr-i9+Fwro=?`-FF2zYm;UwVQv;Uk!ALso*iaaf( zq#Q4j9Tv;=Fs;D!c?7OOs-hrCs&}`w81sPVxvp#ovcUI>WBEQ&4In{i zTwVdc;_)2(BOb4U|AEJAcpu@F=JXZ7FL=C3O?vCnQoq;luRZvpeR_H0>ZSf#{>1dm z;m_DZdmRd1SYFw>+}%LXk@xYTmS_5FoB5UHwdIZO=JLh-^V|8eOZ{`mZJN$FPv}jI kVg+v2P&e?*-UPh}?0MQod)9f9pYxvN)0l7hQv90oU%@nu1ONa4 literal 0 HcmV?d00001 diff --git a/doc/fonts/icomoon.svg b/doc/fonts/icomoon.svg new file mode 100644 index 0000000..0998daf --- /dev/null +++ b/doc/fonts/icomoon.svg @@ -0,0 +1,18 @@ + + + +Generated by IcoMoon + + + + + + + + + + + + + + \ No newline at end of file diff --git a/doc/fonts/icomoon.ttf b/doc/fonts/icomoon.ttf new file mode 100644 index 0000000000000000000000000000000000000000..538fce7b3e34530ac4af0c2d8b5c3447b3020a87 GIT binary patch literal 2932 zcma)8TWnm#8UE+YIdk@$J?pdgo9(?{(l~aU-SsX_Az3OlQR+lNDFziG)r+&(DOs=M z*b4Sj>dJ(E3#7j`U{pOsr@wQmd z*`4{X^Z)~Br`hA+ob=!~l%lIrZ=^AV{ok!kbUV*>20N5v=n%@i7`B{)NN=)FK;}a3Xibnx zE8xem8=M4qxgBk%`B~z#+~3)4?RI|f@4>*au*avnd^Y$ZJIH>R`B&!4gwNK-{s9}K zzMYBQz=*+;Jt9|#T0eHw_vU;*Podaghy9b?q@U6?EVGt~%U>+23!4cYzha?U!LM0& zMD<#$d7L#{&AEodt`rMWQv#P?*Ig6dJXZ`ev8Zd=H4v}Y>aGLNd}XfT8E!S}?1@w& z8IBmHX@=Vh569z)$#gUrGy)-^`$Ms0Fc`GjmM(;DS-?XHpS_ zBogVNFC*glVmA#lJd}VJ^cevo7>opzu-}5q_RuE?U0BnCsqkwWFIuJ@j9N+vS|R`~ zI~1~mRzRQyriIGda!E!coJgip$wZ{h)S%Btw08yvX`U!Dat6A46ve_}c^IR6)xsIn zW7vH88{J$fxO+I;)lxB?wr#%|+k&EDh^Ye8>69G|+NpG@Sj|j~N23;6G&(+! z$&OD9sfqC{yJ_3$RH;%PlreMhI>W>@6Wg|X|wAEXzZ;E*74&jvI~T^S|h z`A?IA7;jLST14dvPP9;+VXb2fV{we>B9Xy{C62OstqE?cQK{-0vp_NZg=BuhuA5FIlCmzw&W7E<6OmX3Wk^WR=S}K)W zIiv=JCiLFvLq9rjJQh_`$?^Fq&CjpKBV6=O)XoO$tHo5nj+Jr$m_&nZh1fP;hy>!U z)6eK(>e2>Xr_`(9GO_r3)p5{AJ}!lYt8#u1Gj~kWYKuU1yGYro^h7fY#e44 za(v3tWZ51N)Uuo6eZ%nZ^$ zBr;#SV94B(YKS50jD~4Fmw}d)W-v#HuZD=r#(6QkO0)D1J40!_5vV_0CfMg;)nUVA z*~Me;@M~P(Nls#aPNTvBj{oE}k^<#;!csQx?0AsplHxRepS{cEJQ8d4R7J;A^Zt4#R=Vo?>F z^GgWMR#X`>zu$=w8pFshVYbTIk=^H1g)xnRVwW)6L2nJUM%Tyjtjl*)_Fvx9`IUX| zH{{22iuTFTk3QIgsDvac0pC5SWw&9D^Zh`ItQJy|julVn0Bfi!Dd_@q%HbxW|khpO|wpyp#TBw96|2Q z?mKH==DF_j=EcrBoOTiDCS9Zst?#ROqQAP4TkfuQ*E<{C^SQ^ia;KL1XAs*Aop5|_ m!22@HtU3{OaQ}7bJa~`M7RFP~0?xP#K#cAeqqFh*5B~v!Lw&9Q literal 0 HcmV?d00001 diff --git a/doc/fonts/icomoon.woff b/doc/fonts/icomoon.woff new file mode 100644 index 0000000000000000000000000000000000000000..a27a656044f41d8a567a5ffa9ed74c75ceae77f8 GIT binary patch literal 3008 zcma)8U2Ggz6+U{O-{;fWW>9 zxgv=_(4uRDt>=Ng0r`}q$h-8%Yr_?3_b%kDBs*OOgjf04wYnd+qm9?-l;U?i=@uqkoS^(8JFG zP6s#}{XRd+|1$UY+#gaCt}*%Nz^m%;p^6cMCs#%FRT@#}Q!3TNd_Nx3C^6b&e_?m& zF}j6eHd0CXOTE=fz-_;}OwT?GAnHQXfP zzGvS-ve{_*0Rs26g;wCW({8foGO2VlX4$qK?P@%lOr~bCanG~D5n+ZRiL~c=PS-Jo zFdYYc#B;-Qi4-@v){UmqnOFXA>sTt49mld_K`eIHvZG@SFmJ*NTb>sSYtfK{klk@i z2&Qo61k(}MwqAB@*NZz^1XdypD>o8xJtr(s!gE4rT{$Ex7EPrynRF`FWx6+k<8-(N z$7zWuc4Q4qe<{kPlX5YR&Q%9%&`e<9qB-bop>k|bgfq^ z8Mr?#`@;I6p0GccagOe=Z93-futeNthXeD+wa_}p>S`A>8+j%-!OIpF_{A5o$cTFT zwy-`dvuyzoDnBi0DvgvWShTnYXqFb(F1t(-D$9E*R>btm?OiTnW6NFlQKa(`n`W22 zAcroj{haWPnI zT=JUhn7FriBv=)|j8|6x) zQ7JbVDL=}^{vF3YZab=C6^??QIc)!k$o<1_#^8=LLk@XWTekC34pvf|LmwgjHfDSn z>&5UYoub#-MftvVaQ<+aV4laP4ihGiT{7`Hzr}UVCpq8PP>l}Y9!CWXijC_W4b(7` zZw{yM=mRayz5&1UbRNFtaSl8OKF6mW+sG79)oYy?E-8e=o<3qOkK;iiaq}lCAOhpB zLduL@`wF(U9?><8J;wk?b*B9cx#*h9`E?}cYPw9BKdN$s#W?fp=&f^hdjBO|W6WTX z*md+)?7gwo(eZIS>+&6y^o39MzT|%9eG1_@MfuH{U;W~+#YsrxB*v?6S#EX z(5v!XqNr|)JFF(@%^(LxCSch^TAFk@;5qnq!0X_*1Kz+HPXxRGz7+5_*8E%Y`$V^Z zOhV)H8u-dJ619In$Ttbqdj1qWjhkkZZomKv_&JTecj}?N9tu9yU){de+eFYl65Xb2)T7OZS}qRPw+gHM z_5NmWtAC~N>`vkQ%J3p`o2Nw;dkfK5;bvWB)WiMPrz^mor5)7g)c`K~13-@QynJl= E4 + + + + tensorflex v0.1.0 – Documentation + + + + + + diff --git a/doc/search.html b/doc/search.html new file mode 100644 index 0000000..f3bb0da --- /dev/null +++ b/doc/search.html @@ -0,0 +1,96 @@ + + + + + + + + Search – tensorflex v0.1.0 + + + + + + + + + + + +
    + + +
    + + + +
    +

    + tensorflex +

    +

    + v0.1.0 +

    +
    + +     + +
    + + +
    + + +
    + +
    + +
    +
    +
    + + + +
    +
    +
    +
    + + + + + + + + From aae0e6589430667775ad1aa52d11271192501f20 Mon Sep 17 00:00:00 2001 From: Anshuman Chhabra Date: Sun, 29 Jul 2018 21:27:31 +0530 Subject: [PATCH 3/5] Made requested changes --- doc/.build | 12 - doc/404.html | 101 -- doc/Tensorflex.html | 1399 -------------------------- doc/api-reference.html | 123 --- doc/dist/app-480ffdc169.css | 1 - doc/dist/app-9bd040e5e5.js | 8 - doc/dist/sidebar_items-3a30d3745e.js | 1 - doc/fonts/icomoon.eot | Bin 3096 -> 0 bytes doc/fonts/icomoon.svg | 18 - doc/fonts/icomoon.ttf | Bin 2932 -> 0 bytes doc/fonts/icomoon.woff | Bin 3008 -> 0 bytes doc/index.html | 11 - doc/search.html | 96 -- lib/tensorflex.ex | 317 ++++-- 14 files changed, 237 insertions(+), 1850 deletions(-) delete mode 100644 doc/.build delete mode 100644 doc/404.html delete mode 100644 doc/Tensorflex.html delete mode 100644 doc/api-reference.html delete mode 100644 doc/dist/app-480ffdc169.css delete mode 100644 doc/dist/app-9bd040e5e5.js delete mode 100644 doc/dist/sidebar_items-3a30d3745e.js delete mode 100644 doc/fonts/icomoon.eot delete mode 100644 doc/fonts/icomoon.svg delete mode 100644 doc/fonts/icomoon.ttf delete mode 100644 doc/fonts/icomoon.woff delete mode 100644 doc/index.html delete mode 100644 doc/search.html diff --git a/doc/.build b/doc/.build deleted file mode 100644 index 149b262..0000000 --- a/doc/.build +++ /dev/null @@ -1,12 +0,0 @@ -dist/app-480ffdc169.css -dist/app-9bd040e5e5.js -fonts/icomoon.eot -fonts/icomoon.svg -fonts/icomoon.ttf -fonts/icomoon.woff -dist/sidebar_items-3a30d3745e.js -api-reference.html -search.html -404.html -Tensorflex.html -index.html diff --git a/doc/404.html b/doc/404.html deleted file mode 100644 index 409102f..0000000 --- a/doc/404.html +++ /dev/null @@ -1,101 +0,0 @@ - - - - - - - - 404 – tensorflex v0.1.0 - - - - - - - - - - - -
    - - -
    - - - -
    -

    - tensorflex -

    -

    - v0.1.0 -

    -
    - -     - -
    - - -
    - - -
    - -
    - -
    -
    -
    - - -

    Page not found

    - -

    Sorry, but the page you were trying to get to, does not exist. You -may want to try searching this site using the sidebar or using our -API Reference page to find what -you were looking for.

    - - -
    -
    -
    -
    - - - - - - - - diff --git a/doc/Tensorflex.html b/doc/Tensorflex.html deleted file mode 100644 index ebbd702..0000000 --- a/doc/Tensorflex.html +++ /dev/null @@ -1,1399 +0,0 @@ - - - - - - - - Tensorflex – tensorflex v0.1.0 - - - - - - - - - - - -
    - - -
    - - - -
    -

    - tensorflex -

    -

    - v0.1.0 -

    -
    - -     - -
    - - -
    - - -
    - -
    - -
    -
    -
    - - -

    - tensorflex v0.1.0 - Tensorflex - -

    - - -
    -

    A simple and fast library for running Tensorflow graph models in Elixir. Tensorflex is written around the Tensorflow C API, and allows Elixir developers to leverage Machine Learning and Deep Learning solutions in their projects.

    -

    NOTE:

    -
      -

    • Make sure that the C API version and Python API version (assuming you are using the Python API for first training your models) are the latest. As of July 2018, the latest version is r1.9.

      -
      • -

    • Since Tensorflex provides Inference capability for pre-trained graph models, it is assumed you have adequate knowledge of the pre-trained models you are using (such as the input data type/dimensions, input and output operation names, etc.). Some basic understanding of the Tensorflow Python API can come in very handy.

      -
      • -

    • Tensorflex consists of multiple NIFs, so exercise caution while using it— providing incorrect operation names for running sessions, incorrect dimensions of tensors than the actual pre-trained graph requires, providing different tensor datatypes than the ones required by the graph can all lead to failure. While these are not easy errors to make, do ensure that you test your solution well before deployment.

      -
      • -
    - -
    - - - -
    -

    - - - Link to this section - - Summary -

    - - - -
    -

    - Functions -

    -
    -
    - append_to_matrix(matrix, datalist) -
    - -

    Appends a single row to the back of a Tensorflex matrix

    -
    - -
    -
    -
    - create_matrix(nrows, ncols, datalist) -
    - -

    Creates a 2-D Tensorflex matrix from custom input specifications

    -
    - -
    -
    -
    - float32_tensor(floatval) -
    - -

    Creates a TF_FLOAT constant value one-dimensional tensor from the floating point value specified

    -
    - -
    -
    -
    - float32_tensor(matrix1, matrix2) -
    - -

    Creates a TF_FLOAT tensor from Tensorflex matrices containing the values and dimensions specified

    -
    - -
    -
    -
    - float32_tensor_alloc(matrix) -
    - -

    Allocates a TF_FLOAT tensor of specified dimensions

    -
    - -
    -
    -
    - float64_tensor(floatval) -
    - -

    Creates a TF_DOUBLE constant value one-dimensional tensor from the floating point value specified

    -
    - -
    -
    -
    - float64_tensor(matrix1, matrix2) -
    - -

    Creates a TF_DOUBLE tensor from Tensorflex matrices containing the values and dimensions specified

    -
    - -
    -
    -
    - float64_tensor_alloc(matrix) -
    - -

    Allocates a TF_DOUBLE tensor of specified dimensions

    -
    - -
    -
    -
    - get_graph_ops(graph) -
    - -

    Used for listing all the operations in a Tensorflow .pb graph

    -
    - -
    -
    -
    - int32_tensor(intval) -
    - -

    Creates a TF_INT32 constant value one-dimensional tensor from the integer value specified

    -
    - -
    -
    -
    - int32_tensor(matrix1, matrix2) -
    - -

    Creates a TF_INT32 tensor from Tensorflex matrices containing the values and dimensions specified

    -
    - -
    -
    -
    - int32_tensor_alloc(matrix) -
    - -

    Allocates a TF_INT32 tensor of specified dimensions

    -
    - -
    -
    -
    - load_csv_as_matrix(filepath, opts \\ []) -
    - -

    Loads high-dimensional data from a CSV file as a Tensorflex 2-D matrix in a super-fast manner

    -
    - -
    -
    -
    - load_image_as_tensor(imagepath) -
    - -

    Loads JPEG images into Tensorflex directly as a TF_UINT8 tensor of dimensions image height x image width x number of color channels

    -
    - -
    -
    -
    - matrix_pos(matrix, row, col) -
    - -

    Used for accessing an element of a Tensorflex matrix

    -
    - -
    -
    -
    - matrix_to_lists(matrix) -
    - -

    Converts a Tensorflex matrix (back) to a list of lists format

    -
    - -
    -
    -
    - read_graph(filepath) -
    - -

    Used for loading a Tensorflow .pb graph model in Tensorflex

    -
    - -
    -
    -
    - run_session(graph, tensor1, tensor2, input_opname, output_opname) -
    - -

    Runs a Tensorflow session to generate predictions for a given graph, input data, and required input/output operations

    -
    - -
    -
    -
    - size_of_matrix(matrix) -
    - -

    Used for obtaining the size of a Tensorflex matrix

    -
    - -
    -
    -
    - string_tensor(stringval) -
    - -

    Creates a TF_STRING constant value string tensor from the string value specified

    -
    - -
    -
    -
    - tensor_datatype(tensor) -
    - -

    Used to get the datatype of a created tensor

    -
    - -
    - -
    - - - - -
    - - - - - -
    -

    - - - Link to this section - - Functions -

    -
    - - -
    - - - Link to this function - - append_to_matrix(matrix, datalist) - - - -
    -
    -

    Appends a single row to the back of a Tensorflex matrix.

    -

    Takes a Tensorflex %Matrix matrix as input and a single row of data (with the same number of columns as the original matrix) as a list of lists (datalist) to append to the original matrix.

    -

    Returns the extended and modified %Matrix struct matrix.

    -

    - - Examples -

    - -
    iex(1)> m = Tensorflex.create_matrix(2,3,[[23,23,23],[32,32,32]])
-%Tensorflex.Matrix{
-data: #Reference<0.153563642.2042232833.193025>,
-ncols: 3,
-nrows: 2
-}
-
-iex(2)> m = Tensorflex.append_to_matrix(m,[[2,2,2]])
-%Tensorflex.Matrix{
-data: #Reference<0.153563642.2042232833.193025>,
-ncols: 3,
-nrows: 3
-}
-
-iex(3)> m = Tensorflex.append_to_matrix(m,[[3,3,3]])
-%Tensorflex.Matrix{
-data: #Reference<0.153563642.2042232833.193025>,
-ncols: 3,
-nrows: 4
-}
-
-iex(4)> m |> Tensorflex.matrix_to_lists
-[[23.0, 23.0, 23.0], [32.0, 32.0, 32.0], [2.0, 2.0, 2.0], [3.0, 3.0, 3.0]]
-
    -

    Incorrect usage will raise:

    -
    iex(5)> m = Tensorflex.append_to_matrix(m,[[2,2,2],[3,3,3]])
-** (ArgumentError) data columns must be same as matrix and number of rows must be 1
-(tensorflex) lib/tensorflex.ex:345: Tensorflex.append_to_matrix/2
-
-iex(5)> m = Tensorflex.append_to_matrix(m,[[2,2,2,2]])      
-** (ArgumentError) data columns must be same as matrix and number of rows must be 1
-(tensorflex) lib/tensorflex.ex:345: Tensorflex.append_to_matrix/2
    - -
    -
    -
    - - -
    - - - Link to this function - - create_matrix(nrows, ncols, datalist) - - - -
    -
    -

    Creates a 2-D Tensorflex matrix from custom input specifications.

    -

    Takes three input arguments: number of rows in matrix (nrows), number of columns in matrix (ncols), and a list of lists of the data that will form the matrix (datalist).

    -

    Returns a %Matrix Tensorflex struct type.

    -

    - - Examples: -

    - -

    Creating a new matrix

    -
    iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]])    %Tensorflex.Matrix{
-data: #Reference<0.759278808.823525378.128525>,
-ncols: 3,
-nrows: 2
-}
    -

    All %Matrix Tensorflex matrices can be passed in to the other matrix inspection and manipulation functions— matrix_pos/3,size_of_matrix/1, matrix_to_lists/1, and append_to_matrix/2:

    -
    iex(1)> mat = Tensorflex.create_matrix(4,4,[[123,431,23,1],[1,2,3,4],[5,6,7,8],[768,564,44,5]])
-%Tensorflex.Matrix{
-data: #Reference<0.878138179.2435973124.131489>,
-ncols: 4,
-nrows: 4
-}
-
-iex(2)> mat = Tensorflex.append_to_matrix(mat, [[1,1,1,1]])
-%Tensorflex.Matrix{
-data: #Reference<0.878138179.2435973124.131489>,
-ncols: 4,
-nrows: 5
-}
-
-iex(3)> Tensorflex.matrix_to_lists mat
-[
-[123.0, 431.0, 23.0, 1.0],
-[1.0, 2.0, 3.0, 4.0],
-[5.0, 6.0, 7.0, 8.0],
-[768.0, 564.0, 44.0, 5.0],
-[1.0, 1.0, 1.0, 1.0]
-]
-
-iex(4)> Tensorflex.matrix_pos(mat,5,3)
-1.0
-
-iex(5)> Tensorflex.size_of_matrix mat
-{5, 4}
    -

    Incorrect usage will raise:

    -
    iex(1)> Tensorflex.create_matrix(1,2,[[1,2,3]])
-** (ArgumentError) argument error
-(tensorflex) Tensorflex.NIFs.create_matrix(1, 2, [[1, 2, 3]])
-(tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3
-
-iex(1)> Tensorflex.create_matrix(2,1,[[1,2,3]])
-** (ArgumentError) argument error
-(tensorflex) Tensorflex.NIFs.create_matrix(2, 1, [[1, 2, 3]])
-(tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3
-
-iex(1)> Tensorflex.create_matrix(2,3,[[1.1,23,3.4], []])
-** (ArgumentError) argument error
-  (tensorflex) Tensorflex.NIFs.create_matrix(2, 3, [[1.1, 23, 3.4], []])
-  (tensorflex) lib/tensorflex.ex:247: Tensorflex.create_matrix/3
-  
-iex(1)> Tensorflex.create_matrix(1,2,[[]])              
-** (ArgumentError) data provided cannot be an empty list
-(tensorflex) lib/tensorflex.ex:243: Tensorflex.create_matrix/3
-
-iex(1)> Tensorflex.create_matrix(-1,2,[[3,4]])
-** (FunctionClauseError) no function clause matching in Tensorflex.create_matrix/3
    - -
    -
    -
    - - -
    - - - Link to this function - - float32_tensor(floatval) - - - -
    -
    -

    Creates a TF_FLOAT constant value one-dimensional tensor from the floating point value specified.

    -

    Takes in a float value as input.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

    -

    - - Examples -

    - -
    iex(1)> {:ok, tensor} = Tensorflex.float32_tensor 123.123
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_float,
-tensor: #Reference<0.2011963375.1804468228.236110>
-}}
-
    -

    Incorrect usage will raise:

    -
    iex(2)> {:ok, tensor} = Tensorflex.float32_tensor "123.123"
-** (FunctionClauseError) no function clause matching in Tensorflex.float32_tensor/1 
-
-iex(2)> {:ok, tensor} = Tensorflex.float32_tensor 123      
-** (FunctionClauseError) no function clause matching in Tensorflex.float32_tensor/1
    - -
    -
    -
    - - -
    - - - Link to this function - - float32_tensor(matrix1, matrix2) - - - -
    -
    -

    Creates a TF_FLOAT tensor from Tensorflex matrices containing the values and dimensions specified.

    -

    Takes two arguments: a %Matrix matrix (matrix1) containing the values the tensor should have and another %Matrix matrix (matrix2) containing the dimensions of the required tensor.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

    -

    - - Examples: -

    - -
    iex(1)> vals = Tensorflex.create_matrix(2,3,[[12.0,45.2,2.11],[36.7,8.09,9.81]])
-%Tensorflex.Matrix{
-data: #Reference<0.1251941183.3671982081.254268>,
-ncols: 3,
-nrows: 2
-}
-
-iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]])
-%Tensorflex.Matrix{
-data: #Reference<0.1251941183.3671982081.254723>,
-ncols: 2,
-nrows: 1
-}
-  
-iex(3)> {:ok, tensor} = Tensorflex.float32_tensor vals,dims
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_float,
-tensor: #Reference<0.1251941183.3671982081.255228>
-}}
-
    - -
    -
    -
    - - -
    - - - Link to this function - - float32_tensor_alloc(matrix) - - - -
    -
    -

    Allocates a TF_FLOAT tensor of specified dimensions.

    -

    This function is generally used to allocate output tensors that do not hold any value data yet, but will after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the run_session/5 function to hold the output values generated as predictions.

    -

    Takes a Tensorflex %Matrix struct matrix as input.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding the potential tensor data and type.

    -

    - - Examples -

    - -

    As an example, we can allocate a float32 output tensor that will be a vector of 250 values (1x250 matrix). Therefore, after the session is run, the output will be a float vector containing 250 values:

    -
    iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float32_tensor_alloc
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_float,
-tensor: #Reference<0.961157994.2087059457.19014>
-}}
-
    - -
    -
    -
    - - -
    - - - Link to this function - - float64_tensor(floatval) - - - -
    -
    -

    Creates a TF_DOUBLE constant value one-dimensional tensor from the floating point value specified.

    -

    Takes in a float value as input.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

    -

    - - Examples -

    - -
    iex(1)> {:ok, tensor} = Tensorflex.float64_tensor 123.123
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_double,
-tensor: #Reference<0.2778616536.4219338753.155412>
-}}
-
    -

    Incorrect usage will raise:

    -
    iex(2)> {:ok, tensor} = Tensorflex.float64_tensor "123.123"
-** (FunctionClauseError) no function clause matching in Tensorflex.float64_tensor/1
-
-iex(2)> {:ok, tensor} = Tensorflex.float64_tensor 123      
-** (FunctionClauseError) no function clause matching in Tensorflex.float64_tensor/1
    - -
    -
    -
    - - -
    - - - Link to this function - - float64_tensor(matrix1, matrix2) - - - -
    -
    -

    Creates a TF_DOUBLE tensor from Tensorflex matrices containing the values and dimensions specified.

    -

    Takes two arguments: a %Matrix matrix (matrix1) containing the values the tensor should have and another %Matrix matrix (matrix2) containing the dimensions of the required tensor.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

    -

    - - Examples: -

    - -
    iex(1)> vals = Tensorflex.create_matrix(2,3,[[12.0,45.2,2.11],[36.7,8.09,9.81]])
-%Tensorflex.Matrix{
-data: #Reference<0.1251941183.3671982081.254268>,
-ncols: 3,
-nrows: 2
-}
-
-iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]])
-%Tensorflex.Matrix{
-data: #Reference<0.1251941183.3671982081.254723>,
-ncols: 2,
-nrows: 1
-}
-
-iex(3)> {:ok, tensor} = Tensorflex.float64_tensor vals,dims
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_double,
-tensor: #Reference<0.1251941183.3671982081.255216>
-}}
-
    - -
    -
    -
    - - -
    - - - Link to this function - - float64_tensor_alloc(matrix) - - - -
    -
    -

    Allocates a TF_DOUBLE tensor of specified dimensions.

    -

    This function is generally used to allocate output tensors that do not hold any value data yet, but will after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the run_session/5 function to hold the output values generated as predictions.

    -

    Takes a Tensorflex %Matrix struct matrix as input.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding the potential tensor data and type.

    -

    - - Examples -

    - -

    As an example, we can allocate a float64 output tensor that will be a vector of 250 values (1x250 matrix). Therefore, after the session is run, the output will be a double vector containing 250 values:

    -
    iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float64_tensor_alloc
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_double,
-tensor: #Reference<0.961157994.2087059457.19025>
-}}
-
    - -
    -
    -
    - - -
    - - - Link to this function - - get_graph_ops(graph) - - - -
    -
    -

    Used for listing all the operations in a Tensorflow .pb graph.

    -

    Reads in a Tensorflex %Graph struct obtained from read_graph/1.

    -

    Returns a list of all the operation names (as strings) that populate the graph model.

    -

    - - Examples -

    - -
      -
    • Google Inception CNN Model (source) -
      • -
    -
    iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb"
-2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
-{:ok,
-%Tensorflex.Graph{
-def: #Reference<0.3018278404.759824385.5268>,
-name: "classify_image_graph_def.pb"
-}}
-
-iex(2)> Tensorflex.get_graph_ops graph
-["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims",
-"ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul",
-"conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta",
-"conv/batchnorm/gamma", "conv/batchnorm/moving_mean",
-"conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics",
-"conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D",
-"conv_1/batchnorm/beta", "conv_1/batchnorm/gamma",
-"conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance",
-"conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency",
-"conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta",
-"conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean",
-"conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics",
-"conv_2/control_dependency", "conv_2", "pool/CheckNumerics",
-"pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D",
-"conv_3/batchnorm/beta", "conv_3/batchnorm/gamma",
-"conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...]
    -
      -
    • Iris Dataset MLP Model (source) -
      • -
    -
    iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_iris.pb"
-{:ok,
-%Tensorflex.Graph{
-def: #Reference<0.4109712726.1847984130.24506>,
-name: "graphdef_iris.pb"
-}}
-
-iex(2)> Tensorflex.get_graph_ops graph
-["input", "weights1", "weights1/read", "biases1", "biases1/read", "weights2", "weights2/read", "biases2", "biases2/read", "MatMul", "Add", "Relu", "MatMul_1", "Add_1", "output"]
-
    -
      -
    • Toy Computational Graph Model (source) -
      • -
    -
    iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_toy.pb"
-{:ok,
-%Tensorflex.Graph{
-def: #Reference<0.1274892327.1580335105.235135>,
-name: "graphdef_toy.pb"
-}}
-
-iex(2)> Tensorflex.get_graph_ops graph
-["input", "weights", "weights/read", "biases", "biases/read", "MatMul", "add", "output"]
    -
      -
    • RNN LSTM Sentiment Analysis Model (source) -
      • -
    -
    iex(1)> {:ok, graph} = Tensorflex.read_graph "frozen_model_lstm.pb"
-{:ok,
-%Tensorflex.Graph{
-def: #Reference<0.713975820.1050542081.11558>,
-name: "frozen_model_lstm.pb"
-}}
-
-iex(2)> Tensorflex.get_graph_ops graph
-["Placeholder_1", "embedding_lookup/params_0", "embedding_lookup",
-"transpose/perm", "transpose", "rnn/Shape", "rnn/strided_slice/stack",
-"rnn/strided_slice/stack_1", "rnn/strided_slice/stack_2", "rnn/strided_slice",
-"rnn/stack/1", "rnn/stack", "rnn/zeros/Const", "rnn/zeros", "rnn/stack_1/1",
-"rnn/stack_1", "rnn/zeros_1/Const", "rnn/zeros_1", "rnn/Shape_1",
-"rnn/strided_slice_2/stack", "rnn/strided_slice_2/stack_1",
-"rnn/strided_slice_2/stack_2", "rnn/strided_slice_2", "rnn/time",
-"rnn/TensorArray", "rnn/TensorArray_1", "rnn/TensorArrayUnstack/Shape",
-"rnn/TensorArrayUnstack/strided_slice/stack",
-"rnn/TensorArrayUnstack/strided_slice/stack_1",
-"rnn/TensorArrayUnstack/strided_slice/stack_2",
-"rnn/TensorArrayUnstack/strided_slice", "rnn/TensorArrayUnstack/range/start",
-"rnn/TensorArrayUnstack/range/delta", "rnn/TensorArrayUnstack/range",
-"rnn/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3",
-"rnn/while/Enter", "rnn/while/Enter_1", "rnn/while/Enter_2",
-"rnn/while/Enter_3", "rnn/while/Merge", "rnn/while/Merge_1",
-"rnn/while/Merge_2", "rnn/while/Merge_3", "rnn/while/Less/Enter",
-"rnn/while/Less", "rnn/while/LoopCond", "rnn/while/Switch",
-"rnn/while/Switch_1", "rnn/while/Switch_2", "rnn/while/Switch_3", ...]
    - -
    -
    -
    - - -
    - - - Link to this function - - int32_tensor(intval) - - - -
    -
    -

    Creates a TF_INT32 constant value one-dimensional tensor from the integer value specified.

    -

    Takes in an integer value as input.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

    -

    - - Examples -

    - -
    iex(1)> {:ok, tensor} = Tensorflex.int32_tensor 123
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_int32,
-tensor: #Reference<0.1927663658.3415343105.162588>
-}}
    -

    Incorrect usage will raise:

    -
    iex(2)> {:ok, tensor} = Tensorflex.int32_tensor 123.123
-** (FunctionClauseError) no function clause matching in Tensorflex.int32_tensor/1 
-
-iex(2)> {:ok, tensor} = Tensorflex.int32_tensor "123.123"
-** (FunctionClauseError) no function clause matching in Tensorflex.int32_tensor/1  
-
    - -
    -
    -
    - - -
    - - - Link to this function - - int32_tensor(matrix1, matrix2) - - - -
    -
    -

    Creates a TF_INT32 tensor from Tensorflex matrices containing the values and dimensions specified.

    -

    Takes two arguments: a %Matrix matrix (matrix1) containing the values the tensor should have and another %Matrix matrix (matrix2) containing the dimensions of the required tensor.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

    -

    NOTE: In case floating point values are passed in the values matrix (matrix1) as arguments for this function, the tensor will still be created and all the float values will be typecast to integers.

    -

    - - Examples: -

    - -
    iex(1)> vals = Tensorflex.create_matrix(2,3,[[123,45,333],[2,2,899]]) 
-%Tensorflex.Matrix{
-data: #Reference<0.1256144000.2868510721.170449>,
-ncols: 3,
-nrows: 2
-}
-
-iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]])
-%Tensorflex.Matrix{
-data: #Reference<0.1256144000.2868510721.170894>,
-ncols: 2,
-nrows: 1
-}
-
-iex(3)> {:ok, tensor} = Tensorflex.int32_tensor vals,dims
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_int32,
-tensor: #Reference<0.1256144000.2868510721.171357>
-}}
-
    - -
    -
    -
    - - -
    - - - Link to this function - - int32_tensor_alloc(matrix) - - - -
    -
    -

    Allocates a TF_INT32 tensor of specified dimensions.

    -

    This function is generally used to allocate output tensors that do not hold any value data yet, but will after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the run_session/5 function to hold the output values generated as predictions.

    -

    Takes a Tensorflex %Matrix struct matrix as input.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding the potential tensor data and type.

    -

    - - Examples -

    - -

    As an example, we can allocate an int32 output tensor that will be a vector of 250 values (1x250 matrix). Therefore, after the session is run, the output will be an integer vector containing 250 values:

    -
    iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.int32_tensor_alloc
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_int32,
-tensor: #Reference<0.961157994.2087059457.18950>
-}}
-
    - -
    -
    -
    - - - - -
    - - - Link to this function - - load_csv_as_matrix(filepath, opts \\ []) - - - -
    -
    -

    Loads high-dimensional data from a CSV file as a Tensorflex 2-D matrix in a super-fast manner.

    -

    The load_csv_as_matrix/2 function is very fast— when compared with the Python based pandas library for data science and analysis’ function read_csv on the test.csv file from MNIST Kaggle data (source), the following execution times were obtained:

    - -

    This function takes in 2 arguments: a path to a valid CSV file (filepath) and other optional arguments opts. These include whether or not a header needs to be discarded in the CSV, and what the delimiter type is. These are specified by passing in an atom :true or :false to the header: key, and setting a string value for the delimiter: key. By default, the header is considered to be present (:true) and the delimiter is set to ,.

    -

    Returns a %Matrix Tensorflex struct type.

    -

    - - Examples: -

    - -

    We first exemplify the working with the test.csv file which belongs to the MNIST Kaggle CSV data (source), which contains 28000 rows and 784 columns (without the header). It is comma delimited and also contains a header. From the test.csv file, we also create a custom file withou the header present which we refer to as test_without_header.csv in the examples below:

    -
    iex(1)> mat = Tensorflex.load_csv_as_matrix("test.csv")
-%Tensorflex.Matrix{
-data: #Reference<0.4024686574.590479361.258459>,
-ncols: 784,
-nrows: 28000
-}
-
-iex(2)> Tensorflex.matrix_pos mat, 5,97
-80.0
-
-iex(3)> Tensorflex.matrix_pos mat, 5,96
-13.0
    -

    On a visual inspection of the very large test.csv file, one can see that the values in these particular positions are correct. Now we show usage for the same file but without header, test_without_header.csv:

    -
    iex(1)> no_header = Tensorflex.load_csv_as_matrix("test/test_without_header.csv", header: :false)    
-%Tensorflex.Matrix{
-data: #Reference<0.4024686574.590479364.257078>,
-ncols: 784,
-nrows: 28000
-}
-
-iex(2)> Tensorflex.matrix_pos no_header,5,97
-80.0
-
-iex(3)> Tensorflex.matrix_pos no_header,5,96
-13.0
    -

    Next we see the delimiter functionalities. First, assuming we have two simple CSV files, sample1.csv and sample2.csv

    -

    sample1.csv:

    -
    1,2,3,4,5
-6,7,8,9,10
-11,12,13,14,15
    -

    sample2.csv:

    -
    col1-col2-col3-col4
-1-2-3-4
-5-6-7-8
-9-10-11-12
    -

    The examples are as follows:

    -
    iex(1)> m1 = Tensorflex.load_csv_as_matrix("sample1.csv", header: :false) 
-%Tensorflex.Matrix{
-data: #Reference<0.3878093040.3013214209.247502>,
-ncols: 5,
-nrows: 3
-}
-
-iex(2)> Tensorflex.matrix_to_lists m1
-[
-[1.0, 2.0, 3.0, 4.0, 5.0],
-[6.0, 7.0, 8.0, 9.0, 10.0],
-[11.0, 12.0, 13.0, 14.0, 15.0]
-]
-
-iex(3)> m2 = Tensorflex.load_csv_as_matrix("sample2.csv", header: :true, delimiter: "-")
-%Tensorflex.Matrix{
-data: #Reference<0.4024686574.590479361.258952>,
-ncols: 4,
-nrows: 3
-}
-
-iex(4)> Tensorflex.matrix_to_lists m2
-[[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0], [9.0, 10.0, 11.0, 12.0]]
    -

    Incorrect usage will raise:

    -
    iex(1)> not_working = Tensorflex.load_csv_as_matrix("test.csv", header: :no_header, delimiter: ",")
-** (ArgumentError) header indicator atom must be either :true or :false 
-(tensorflex) lib/tensorflex.ex:122: Tensorflex.load_csv_as_matrix/2
    - -
    -
    -
    - - -
    - - - Link to this function - - load_image_as_tensor(imagepath) - - - -
    -
    -

    Loads JPEG images into Tensorflex directly as a TF_UINT8 tensor of dimensions image height x image width x number of color channels.

    -

    This function is very useful if you wish to do image classification using Convolutional Neural Networks, or other Deep Learning Models. One of the most widely adopted and robust image classification models is the Inception model by Google. It makes classifications on images from over a 1000 classes with highly accurate results. The load_image_as_tensor/1 function is an essential component for the prediction pipeline of the Inception model (and for other similar image classification models) to work in Tensorflex.

    -

    Reads in the path to a JPEG image file (.jpg or .jpeg).

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding the tensor data and type. Here the created Tensor is a uint8 tensor (TF_UINT8).

    -

    NOTE: For now, only 3 channel RGB JPEG color images can be passed as arguments. Support for grayscale images and other image formats such as PNG will be added in the future.

    -

    - - Examples -

    - -

    To exemplify the working of the load_image_as_tensor/1 function we will cover the entire prediction pipeline for the Inception model. However, this makes use of many other Tensorflex functions such as run_session/5 and the other tensor functions so it would be advisable to go through them first. Also, the Inception model can be downloaded here. We will be making use of the cropped_panda.jpg image file that comes along with the model to test out the model in Tensorflex.

    -

    First the graph is loaded:

    -
    iex(1)> {:ok, graph} = Tensorflex.read_graph("classify_image_graph_def.pb")
-2018-07-25 14:20:29.079139: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
-{:ok,
-%Tensorflex.Graph{
-def: #Reference<0.542869014.389152771.105680>,
-name: "classify_image_graph_def.pb"
-}}
    -

    Then we load the image as a uint8 tensor:

    -
    iex(2)> {:ok, input_tensor} = Tensorflex.load_image_as_tensor("cropped_panda.jpg")
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_uint8,
-tensor: #Reference<0.1203951739.122552322.52747>
-}}
    -

    Then we create the output tensor which will hold out output vector values. For the Inception model, the output is received as a 1008x1 float32 tensor, as there are 1008 classes in the model:

    -
    iex(3)> {:ok, output_tensor} = Tensorflex.create_matrix(1,2,[[1008,1]]) |> Tensorflex.float32_tensor_alloc
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_float,
-tensor: #Reference<0.1203951739.122552322.52794>
-}}
    -

    Next, we obtain the results by running the session:

    -
    iex(4)> results = Tensorflex.run_session(graph, input_tensor, output_tensor, "DecodeJpeg", "softmax")
-2018-07-25 14:33:40.992813: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA
-[
-[1.059142014128156e-4, 2.8240500250831246e-4, 8.30648496048525e-5,
-1.2982363114133477e-4, 7.32232874725014e-5, 8.014426566660404e-5,
-6.63459359202534e-5, 0.003170756157487631, 7.931600703159347e-5,
-3.707312498590909e-5, 3.0997329304227605e-5, 1.4232713147066534e-4,
-1.0381334868725389e-4, 1.1057958181481808e-4, 1.4321311027742922e-4,
-1.203602587338537e-4, 1.3130248407833278e-4, 5.850398520124145e-5,
-2.641105093061924e-4, 3.1629020668333396e-5, 3.906813799403608e-5,
-2.8646905775531195e-5, 2.2863158665131778e-4, 1.2222197256051004e-4,
-5.956588938715868e-5, 5.421260357252322e-5, 5.996063555357978e-5,
-4.867801326327026e-4, 1.1005574924638495e-4, 2.3433618480339646e-4,
-1.3062104699201882e-4, 1.317620772169903e-4, 9.388553007738665e-5,
-7.076268957462162e-5, 4.281177825760096e-5, 1.6863139171618968e-4,
-9.093972039408982e-5, 2.611844101920724e-4, 2.7584232157096267e-4,
-5.157176201464608e-5, 2.144951868103817e-4, 1.3628098531626165e-4,
-8.007588621694595e-5, 1.7929042223840952e-4, 2.2831936075817794e-4,
-6.216531619429588e-5, 3.736453436431475e-5, 6.782123091397807e-5,
-1.1538144462974742e-4, ...]
-]
-
    -

    Finally, we need to find which class has the maximum probability and identify it’s label. Since results is a List of Lists, it’s better to read in the flattened list. Then we need to find the index of the element in the new list which as the maximum value. Therefore:

    -
    iex(5)> max_prob = List.flatten(results) |> Enum.max
-0.8849328756332397
-
-iex(6)> Enum.find_index(results |> List.flatten, fn(x) -> x == max_prob end)
-169
    -

    We can thus see that the class with the maximum probability predicted (0.8849328756332397) for the image is 169. We will now find what the 169 label corresponds to. For this we can look back into the unzipped Inception folder, where there is a file called imagenet_2012_challenge_label_map_proto.pbtxt. On opening this file, we can find the string class identifier for the 169 class index. This is n02510455 and is present on Line 1556 in the file. Finally, we need to match this string identifier to a set of identification labels by referring to the file imagenet_synset_to_human_label_map.txt file. Here we can see that corresponding to the string class n02510455 the human labels are giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca (Line 3691 in the file). Thus, we have correctly identified the animal in the image as a panda using Tensorflex.

    - -
    -
    -
    - - -
    - - - Link to this function - - matrix_pos(matrix, row, col) - - - -
    -
    -

    Used for accessing an element of a Tensorflex matrix.

    -

    Takes in three input arguments: a Tensorflex %Matrix struct matrix, and the row (row) and column (col) values of the required element in the matrix. Both row and col here are NOT zero indexed.

    -

    Returns the value as float.

    -

    - - Examples -

    - -
    iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]])
-%Tensorflex.Matrix{
-data: #Reference<0.759278808.823525378.128525>,
-ncols: 3,
-nrows: 2
-}
-
-iex(2)> Tensorflex.matrix_pos(mat,2,1)
-5.5
-
-iex(3)> Tensorflex.matrix_pos(mat,1,3)
-44.5
-
    - -
    -
    -
    - - -
    - - - Link to this function - - matrix_to_lists(matrix) - - - -
    -
    -

    Converts a Tensorflex matrix (back) to a list of lists format.

    -

    Takes a Tensorflex %Matrix struct matrix as input.

    -

    Returns a list of lists representing the data stored in the matrix.

    -

    NOTE: If the matrix contains very high dimensional data, typically obtained from a function like load_csv_as_matrix/2, then it is not recommended to convert the matrix back to a list of lists format due to a possibility of memory errors.

    -

    - - Examples -

    - -
    iex(1)> Tensorflex.create_matrix(2,3,[[23,23,23],[32,32,32]]) |> Tensorflex.matrix_to_lists
-[[23.0, 23.0, 23.0], [32.0, 32.0, 32.0]]
    - -
    -
    -
    - - -
    - - - Link to this function - - read_graph(filepath) - - - -
    -
    -

    Used for loading a Tensorflow .pb graph model in Tensorflex.

    -

    Reads in a pre-trained Tensorflow protobuf (.pb) Graph model binary file.

    -

    Returns a tuple {:ok, %Graph}.

    -

    %Graph is an internal Tensorflex struct which holds the name of the graph file and the binary definition data that is read in via the .pb file.

    -

    - - Examples: -

    - -

    Reading in a graph

    -

    As an example, we can try reading in the Inception convolutional neural network based image classification graph model by Google. The graph file is named classify_image_graph_def.pb:

    -
    iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb"
-2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
-{:ok,
-%Tensorflex.Graph{
-def: #Reference<0.3018278404.759824385.5268>,
-name: "classify_image_graph_def.pb"
-}}
    -

    Generally to check that the loaded graph model is correct and contains computational operations, the get_graph_ops/1 function is useful:

    -
    iex(2)> Tensorflex.get_graph_ops graph
-["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims",
-"ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul",
-"conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta",
-"conv/batchnorm/gamma", "conv/batchnorm/moving_mean",
-"conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics",
-"conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D",
-"conv_1/batchnorm/beta", "conv_1/batchnorm/gamma",
-"conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance",
-"conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency",
-"conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta",
-"conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean",
-"conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics",
-"conv_2/control_dependency", "conv_2", "pool/CheckNumerics",
-"pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D",
-"conv_3/batchnorm/beta", "conv_3/batchnorm/gamma",
-"conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...]
-
    -

    Incorrect usage will raise:

    -
    iex(3)> {:ok, graph} = Tensorflex.read_graph "Makefile"
-** (ArgumentError) file is not a protobuf .pb file
-(tensorflex) lib/tensorflex.ex:27: Tensorflex.read_graph/1
-
-iex(3)> {:ok, graph} = Tensorflex.read_graph "Makefile.pb"
-** (ArgumentError) graph definition file does not exist
-(tensorflex) lib/tensorflex.ex:23: Tensorflex.read_graph/1
-
    - -
    -
    -
    - - -
    - - - Link to this function - - run_session(graph, tensor1, tensor2, input_opname, output_opname) - - - -
    -
    -

    Runs a Tensorflow session to generate predictions for a given graph, input data, and required input/output operations.

    -

    This function is the final step of the Inference (prediction) pipeline and generates output for a given set of input data, a pre-trained graph model, and the specified input and output operations of the graph.

    -

    Takes in five arguments: a pre-trained Tensorflow graph .pb model read in from the read_graph/1 function (graph), an input tensor with the dimensions and data required for the input operation of the graph to run (tensor1), an output tensor allocated with the right dimensions (tensor2), the name of the input operation of the graph that needs where the input data is fed (input_opname), and the output operation name in the graph where the outputs are obtained (output_opname). The input tensor is generally created from the matrices manually or using the load_csv_as_matrix/2 function, and then passed through to one of the tensor creation functions. For image classification the load_image_as_tensor/1 can also be used to create the input tensor from an image. The output tensor is created using the tensor allocation functions (generally containing alloc at the end of the function name).

    -

    Returns a List of Lists (similar to the matrix_to_lists/1 function) containing the generated predictions as per the output tensor dimensions.

    -

    - - Examples -

    - -
      -

    • A blog post here covers generating predictions and running sessions using an MLP model on the Iris Dataset

      -
      • -

    • Generating predictions from the Inception model by Google is covered in the load_image_as_tensor/1 function examples.

      -
      • -

    • Working with an RNN-LSTM example for sentiment analysis is covered here.

      -
      • -
    - -
    -
    -
    - - -
    - - - Link to this function - - size_of_matrix(matrix) - - - -
    -
    -

    Used for obtaining the size of a Tensorflex matrix.

    -

    Takes a Tensorflex %Matrix struct matrix as input.

    -

    Returns a tuple {nrows, ncols} where nrows represents the number of rows of the matrix and ncols represents the number of columns of the matrix.

    -

    - - Examples -

    - -
    iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]])
-%Tensorflex.Matrix{
-data: #Reference<0.759278808.823525378.128525>,
-ncols: 3,
-nrows: 2
-}
-
-iex(2)> Tensorflex.size_of_matrix mat
-{2, 3}
    - -
    -
    -
    - - -
    - - - Link to this function - - string_tensor(stringval) - - - -
    -
    -

    Creates a TF_STRING constant value string tensor from the string value specified.

    -

    Takes in a string value as input.

    -

    Returns a tuple {:ok, %Tensor} where %Tensor represents an internal Tensorflex struct type that is used for holding tensor data and type.

    -

    - - Examples -

    - -
    iex(1)> {:ok, tensor} = Tensorflex.string_tensor "123.123"
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_string,
-tensor: #Reference<0.2069282048.194904065.41126>
-}}
-
    -

    Incorrect usage will raise:

    -
    iex(2)> {:ok, tensor} = Tensorflex.string_tensor 123.123  
-** (FunctionClauseError) no function clause matching in Tensorflex.string_tensor/1 
-
-iex(2)> {:ok, tensor} = Tensorflex.string_tensor 123    
-** (FunctionClauseError) no function clause matching in Tensorflex.string_tensor/1
    - -
    -
    -
    - - -
    - - - Link to this function - - tensor_datatype(tensor) - - - -
    -
    -

    Used to get the datatype of a created tensor.

    -

    Takes in a %Tensor struct tensor as input.

    -

    Returns a tuple {:ok, datatype} where datatype is an atom representing the list of Tensorflow TF_DataType tensor datatypes. Click here to view a list of all possible datatypes.

    -

    - - Examples -

    - -
    iex(1)> {:ok, tensor} = Tensorflex.string_tensor "example"
-{:ok,
-%Tensorflex.Tensor{
-datatype: :tf_string,
-tensor: #Reference<0.4132928949.2894987267.194583>
-}}
-
-iex(2)> Tensorflex.tensor_datatype tensor
-{:ok, :tf_string}
    - -
    -
    - -
    - - - - -
    -
    -
    -
    - - - - - - - - diff --git a/doc/api-reference.html b/doc/api-reference.html deleted file mode 100644 index ce885f9..0000000 --- a/doc/api-reference.html +++ /dev/null @@ -1,123 +0,0 @@ - - - - - - - - API Reference – tensorflex v0.1.0 - - - - - - - - - - - -
    - - -
    - - - -
    -

    - tensorflex -

    -

    - v0.1.0 -

    -
    - -     - -
    - - -
    - - -
    - -
    - -
    -
    -
    - - -

    - tensorflex v0.1.0 - API Reference -

    - - -
    -

    - - Modules -

    - -
    -
    -
    Tensorflex
    - -

    A simple and fast library for running Tensorflow graph models in Elixir. Tensorflex is written around the Tensorflow C API, and allows Elixir developers to leverage Machine Learning and Deep Learning solutions in their projects

    -
    - -
    - -
    -
    - - - - - - - -
    -
    -
    -
    - - - - - - - - diff --git a/doc/dist/app-480ffdc169.css b/doc/dist/app-480ffdc169.css deleted file mode 100644 index c30bd32..0000000 --- a/doc/dist/app-480ffdc169.css +++ /dev/null @@ -1 +0,0 @@ -@import url(https://fonts.googleapis.com/css?family=Lato:300,700|Merriweather:300italic,300|Inconsolata:400,700);.hljs,article,aside,details,figcaption,figure,footer,header,hgroup,main,menu,nav,section,summary{display:block}.hljs-strong,b,optgroup,strong{font-weight:700}.hljs-emphasis,dfn{font-style:italic}img,legend{border:0}#search ul,.sidebar ul{list-style:none}.night-mode-toggle:focus,.sidebar .sidebar-search .search-input:focus,.sidebar .sidebar-search .search-input:hover,.sidebar-button:active,.sidebar-button:focus,.sidebar-button:hover,a:active,a:hover{outline:0}.content-inner .footer a,.content-inner a,body.night-mode .content-inner a{-webkit-text-decoration-skip:ink;text-decoration-skip:ink}.hljs-comment,.hljs-quote{color:#8e908c}.hljs-deletion,.hljs-name,.hljs-regexp,.hljs-selector-class,.hljs-selector-id,.hljs-tag,.hljs-template-variable,.hljs-variable{color:#c82829}.hljs-built_in,.hljs-builtin-name,.hljs-link,.hljs-literal,.hljs-meta,.hljs-number,.hljs-params,.hljs-type{color:#f5871f}.hljs-attribute{color:#eab700}.hljs-addition,.hljs-bullet,.hljs-string,.hljs-symbol{color:#718c00}.hljs-section,.hljs-title{color:#4271ae}.hljs-keyword,.hljs-selector-tag{color:#8959a8}.hljs{overflow-x:auto;background:#fff;color:#4d4d4c;padding:.5em}legend,td,th{padding:0}/*! normalize.css v3.0.3 | MIT License | github.com/necolas/normalize.css */html{font-family:sans-serif;-ms-text-size-adjust:100%;-webkit-text-size-adjust:100%}audio,canvas,progress,video{display:inline-block;vertical-align:baseline}audio:not([controls]){display:none;height:0}[hidden],template{display:none}a{background-color:transparent}abbr[title]{border-bottom:1px dotted}h1{font-size:2em;margin:.67em 0}mark{background:#ff0;color:#000}.content-outer,body{background-color:#fff}small{font-size:80%}sub,sup{font-size:75%;line-height:0;position:relative;vertical-align:baseline}sup{top:-.5em}sub{bottom:-.25em}svg:not(:root){overflow:hidden}figure{margin:1em 40px}hr{box-sizing:content-box;height:0}pre,textarea{overflow:auto}code,kbd,pre,samp{font-family:monospace,monospace;font-size:1em}button,input,optgroup,select,textarea{color:inherit;font:inherit;margin:0}button{overflow:visible}button,select{text-transform:none}button,html input[type=button],input[type=reset],input[type=submit]{-webkit-appearance:button;cursor:pointer}button[disabled],html input[disabled]{cursor:default}button::-moz-focus-inner,input::-moz-focus-inner{border:0;padding:0}input{line-height:normal}input[type=checkbox],input[type=radio]{box-sizing:border-box;padding:0}input[type=number]::-webkit-inner-spin-button,input[type=number]::-webkit-outer-spin-button{height:auto}.content,.main,.sidebar,body,html{height:100%}input[type=search]{-webkit-appearance:textfield;box-sizing:content-box}input[type=search]::-webkit-search-cancel-button,input[type=search]::-webkit-search-decoration{-webkit-appearance:none}fieldset{border:1px solid silver;margin:0 2px;padding:.35em .625em .75em}table{border-collapse:collapse;border-spacing:0}@font-face{font-family:icomoon;src:url(../fonts/icomoon.eot?h5z89e);src:url(../fonts/icomoon.eot?#iefixh5z89e) format('embedded-opentype'),url(../fonts/icomoon.ttf?h5z89e) format('truetype'),url(../fonts/icomoon.woff?h5z89e) format('woff'),url(../fonts/icomoon.svg?h5z89e#icomoon) format('svg');font-weight:400;font-style:normal}.icon-elem,[class*=" icon-"],[class^=icon-]{font-family:icomoon;speak:none;font-style:normal;font-weight:400;font-variant:normal;text-transform:none;line-height:1;-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale}.sidebar,body{font-family:Lato,sans-serif}.icon-link:before{content:"\e005"}.icon-search:before{content:"\e036"}.icon-cross:before{content:"\e117"}@media screen and (max-width:768px){.icon-menu{font-size:1em}}@media screen and (min-width:769px){.icon-menu{font-size:1.25em}}@media screen and (min-width:1281px){.icon-menu{font-size:1.5em}}.icon-menu:before{content:"\e120"}.icon-angle-right:before{content:"\f105"}.icon-code:before{content:"\f121"}body,html{box-sizing:border-box;width:100%}body{margin:0;font-size:16px;line-height:1.6875em}*,:after,:before{box-sizing:inherit}.main{display:-ms-flexbox;display:-ms-flex;display:flex;-ms-flex-pack:end;justify-content:flex-end}.sidebar{display:-ms-flexbox;display:-ms-flex;display:flex;min-height:0;-moz-flex-direction:column;-ms-flex-direction:column;flex-direction:column;width:300px;position:fixed;top:0;left:0;z-index:4}.content{width:100%;padding-left:300px;overflow-y:auto;-webkit-overflow-scrolling:touch;position:relative;z-index:3}@media screen and (max-width:768px){body .content{z-index:0;padding-left:0}body .sidebar{z-index:3;transform:translateX(-102%);will-change:transform}}body.sidebar-closed .sidebar,body.sidebar-closing .sidebar,body.sidebar-opening .sidebar{z-index:0}body.sidebar-opened .sidebar-button,body.sidebar-opening .sidebar-button{transform:translateX(250px)}@media screen and (max-width:768px){body.sidebar-opened .sidebar,body.sidebar-opening .sidebar{transform:translateX(0)}}body.sidebar-closed .content,body.sidebar-closing .content{padding-left:0}body.sidebar-closed .sidebar-button,body.sidebar-closing .sidebar-button{transform:none}body.sidebar-closed .sidebar-button{color:#000}body.sidebar-opening .sidebar-button{transition:transform .3s ease-in-out}body.sidebar-opening .content{padding-left:300px;transition:padding-left .3s ease-in-out}@media screen and (max-width:768px){body.sidebar-opening .content{padding-left:0}body.sidebar-opening .sidebar{transition:transform .3s ease-in-out;z-index:3}}body.sidebar-closing .sidebar-button{transition:transform .3s ease-in-out}body.sidebar-closing .content{transition:padding-left .3s ease-in-out}@media screen and (max-width:768px){body.sidebar-closing .sidebar{z-index:3;transition:transform .3s ease-in-out;transform:translateX(-102%)}}.sidebar a,.sidebar-button{transition:color .3s ease-in-out}body.sidebar-closed .sidebar{visibility:hidden}.content-inner{max-width:949px;margin:0 auto;padding:3px 60px}.content-outer{min-height:100%}@media screen and (max-width:768px){.content-inner{padding:27px 20px 27px 40px}}.sidebar-button{position:fixed;z-index:99;left:18px;color:#e1e1e1;background-color:transparent;border:none;padding:0;font-size:16px;will-change:transform;transform:translateX(250px)}.sidebar-button:hover{color:#e1e1e1}.sidebar-button .sidebar-toggle{top:8px}@media screen and (max-width:768px){.sidebar-button{transform:translateX(0);left:5px;top:5px}.sidebar-opened .sidebar-button{left:18px}}.sidebar{font-size:15px;line-height:18px;background:#373f52;color:#d5dae6;overflow:hidden}.sidebar .gradient{background:linear-gradient(#373f52,rgba(55,63,82,0));height:20px;margin-top:-20px;pointer-events:none;position:relative;top:20px;z-index:100}.sidebar ul li{margin:0;padding:0 10px}.sidebar a{color:#d5dae6;text-decoration:none}.sidebar a:hover{color:#fff}.sidebar .sidebar-projectLink{margin:18px 30px 0}.sidebar .sidebar-projectDetails{display:inline-block;text-align:right;vertical-align:top;margin-top:6px}.sidebar .sidebar-projectImage{display:inline-block;max-width:64px;max-height:64px;margin-left:15px;vertical-align:bottom}.sidebar .sidebar-projectName{font-weight:700;font-size:24px;line-height:30px;color:#fff;margin:0;padding:0;max-width:230px;word-wrap:break-word}.sidebar .sidebar-projectVersion{margin:0;padding:0;font-weight:300;font-size:16px;line-height:20px;color:#fff}.sidebar .sidebar-listNav{padding:10px 30px 20px;margin:0}.sidebar .sidebar-listNav li,.sidebar .sidebar-listNav li a{text-transform:uppercase;font-weight:300;font-size:14px}.sidebar .sidebar-listNav li{padding-left:17px;border-left:3px solid transparent;transition:all .3s linear;line-height:27px}.sidebar .sidebar-listNav li.selected,.sidebar .sidebar-listNav li.selected a,.sidebar .sidebar-listNav li:hover,.sidebar .sidebar-listNav li:hover a{border-color:#9768d1;color:#fff}.sidebar .sidebar-search{margin:18px 30px;display:-ms-flexbox;display:-ms-flex;display:flex}.sidebar .sidebar-search .search-button:hover,.sidebar .sidebar-search.selected .search-button{color:#9768d1}.sidebar .sidebar-search .search-button{font-size:14px;color:#d5dae6;background-color:transparent;border:none;padding:2px 0 0;margin:0}.sidebar .sidebar-search .search-input{background-color:transparent;border:none;border-radius:0;border-bottom:1px solid #959595;margin-left:5px;height:20px}.sidebar .sidebar-search .icon-search{font-weight:700}.sidebar #full-list{margin:0 0 0 30px;padding:10px 20px 40px;overflow-y:auto;-webkit-overflow-scrolling:touch;-moz-flex:1 1 .01%;-ms-flex:1 1 .01%;flex:1 1 .01%;-ms-flex-positive:1;-ms-flex-negative:1;-ms-flex-preferred-size:.01%}.sidebar #full-list ul{display:none;margin:9px 15px;padding:0}.sidebar #full-list ul li{font-weight:300;line-height:18px;padding:2px 10px}.sidebar #full-list ul li a.expand:before{content:"+";font-family:monospaced;font-size:15px;float:left;width:13px;margin-left:-13px}.sidebar #full-list ul li.open a.expand:before{content:"−"}.sidebar #full-list ul li ul{display:none;margin:9px 6px}.sidebar #full-list li.open>ul,.sidebar #full-list ul li.open>ul{display:block}.sidebar #full-list ul li ul li{border-left:1px solid #959595;padding:0 10px}.sidebar #full-list ul li ul li.active:before{font-family:icomoon;speak:none;font-style:normal;font-weight:400;font-variant:normal;text-transform:none;line-height:1;-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale;content:"\f105";margin-left:-10px;font-size:16px;margin-right:5px}.sidebar #full-list li{padding:0;line-height:27px}.sidebar #full-list li.active{border-left:none}.sidebar #full-list li.active>a,.sidebar #full-list li.clicked>a{color:#fff}.sidebar #full-list li.group{text-transform:uppercase;font-weight:700;font-size:.8em;margin:2em 0 0;line-height:1.8em;color:#ddd}@media screen and (max-height:500px){.sidebar{overflow-y:auto}.sidebar #full-list{overflow:visible}}.content-inner{font-family:Merriweather,'Book Antiqua',Georgia,'Century Schoolbook',serif;font-size:1em;line-height:1.6875em}.content-inner h1,.content-inner h2,.content-inner h3,.content-inner h4,.content-inner h5,.content-inner h6{font-family:Lato,sans-serif;font-weight:700;line-height:1.5em;word-wrap:break-word}.content-inner h1{font-size:2em;margin:1em 0 .5em}.content-inner h1.section-heading{margin:1.5em 0 .5em}.content-inner h1 small{font-weight:300}.content-inner h1 a.view-source{font-size:1.2rem}.content-inner h2{font-size:1.6em;margin:1em 0 .5em;font-weight:700}.content-inner h3{font-size:1.375em;margin:1em 0 .5em;font-weight:700}.content-inner a{color:#000;text-decoration:underline}.content-inner a:visited{color:#000}.content-inner ul li{line-height:1.5em}.content-inner ul li>p{margin:0}.content-inner a.view-source{float:right;color:#959595;text-decoration:none;border:none;transition:color .3s ease-in-out;margin-top:1px}.content-inner a.view-source:hover{color:#373f52}.content-inner .note{color:#959595;margin:0 5px;font-size:14px;font-weight:400}.content-inner blockquote{font-style:italic;margin:.5em 0;padding:.25em 1.5em;border-left:3px solid #e1e1e1;display:inline-block}.content-inner blockquote :first-child{padding-top:0;margin-top:0}.content-inner blockquote :last-child{padding-bottom:0;margin-bottom:0}.content-inner table{margin:2em 0}.content-inner th{text-align:left;font-family:Lato,sans-serif;text-transform:uppercase;font-weight:700;padding-bottom:.5em}.content-inner tr{border-bottom:1px solid #d5dae6;vertical-align:bottom;height:2.5em}.content-inner .summary .summary-row .summary-signature a,.content-inner .summary h2 a{border:none;text-decoration:none}.content-inner td,.content-inner th{padding-left:1em;line-height:2em;vertical-align:top}.content-inner .section-heading:hover a.hover-link{opacity:1;text-decoration:none}.content-inner .section-heading a.hover-link{transition:opacity .3s ease-in-out;display:inline-block;opacity:0;padding:.3em .6em .6em;line-height:1em;margin-left:-2.7em;text-decoration:none;border:none;font-size:16px;vertical-align:middle}.content-inner .detail h2.section-heading{margin-left:.3em}.content-inner .visible-xs{display:none!important}@media screen and (max-width:768px){.content-inner .visible-xs{display:block!important}}.content-inner img{max-width:100%}.content-inner .summary h2{font-weight:700}.content-inner .summary .summary-row .summary-signature{font-family:Inconsolata,Menlo,Courier,monospace;font-weight:700}.content-inner .summary .summary-row .summary-synopsis{font-family:Merriweather,'Book Antiqua',Georgia,'Century Schoolbook',serif;font-style:italic;padding:0 1.2em;margin:0 0 .5em}.content-inner .summary .summary-row .summary-synopsis p{margin:0;padding:0}@keyframes blink-background{from{background-color:#f7f7f7}to{background-color:#ff9}}.content-inner .detail:target .detail-header{animation-duration:.55s;animation-name:blink-background;animation-iteration-count:1;animation-timing-function:ease-in-out}.content-inner .detail-header{margin:2em 0 1em;padding:.5em 1em;background:#f7f7f7;border-left:3px solid #9768d1;font-size:1em;font-family:Inconsolata,Menlo,Courier,monospace;position:relative}.content-inner .detail-header .note{float:right}.content-inner .detail-header .signature{font-size:1rem;font-weight:700}.content-inner .detail-header:hover a.detail-link{opacity:1;text-decoration:none}.content-inner .detail-header a.detail-link{transition:opacity .3s ease-in-out;position:absolute;top:0;left:0;display:block;opacity:0;padding:.6em;line-height:1.5em;margin-left:-2.5em;text-decoration:none;border:none}#search h1,.content-inner .footer .line{display:inline-block}.content-inner .specs pre,.content-inner code{font-family:Inconsolata,Menlo,Courier,monospace;font-style:normal;line-height:24px}.content-inner .specs{opacity:.7;padding-bottom:.05em}.content-inner .specs pre{font-size:.9em;white-space:pre-wrap;margin:0;padding:0}.content-inner .docstring{margin:1.2em 0 2.1em 1.2em}.content-inner .docstring h2,.content-inner .docstring h3,.content-inner .docstring h4,.content-inner .docstring h5{font-weight:700}.content-inner .docstring h2{font-size:1.1em}.content-inner .docstring h3{font-size:1em}.content-inner .docstring h4{font-size:.95em}.content-inner .docstring h5{font-size:.9em}.content-inner a.no-underline,.content-inner pre a{color:#9768d1;text-shadow:none;text-decoration:none;background-image:none}.content-inner a.no-underline:active,.content-inner a.no-underline:focus,.content-inner a.no-underline:hover,.content-inner a.no-underline:visited,.content-inner pre a:active,.content-inner pre a:focus,.content-inner pre a:hover,.content-inner pre a:visited{color:#9768d1;text-decoration:none}.content-inner code{font-weight:400;background-color:#f7f9fc;vertical-align:baseline;border-radius:2px;padding:.1em .2em;border:1px solid #d2ddee}.content-inner pre{margin:1.5em 0}.content-inner pre.spec{margin:0}.content-inner pre.spec code{padding:0}.content-inner pre code.hljs{white-space:inherit;padding:.5em 1em;background-color:#f7f9fc}.content-inner .footer{margin:4em auto 1em;text-align:center;font-style:italic;font-size:14px;color:#959595}.content-inner .footer a{color:#959595;text-decoration:underline}.content-inner .footer a:visited{color:#959595}#search{min-height:200px}#search .result-id{font-size:1.2em}#search .result-id a{text-decoration:none;transition:color .3s ease-in-out}#search .result-id a:active,#search .result-id a:focus,#search .result-id a:visited{color:#000}#search .result-id a:hover{color:#9768d1}#search .result-elem em,#search .result-id em{font-style:normal;color:#9768d1}#search ul{margin:0;padding:0}.night-mode-toggle{top:1.5em}.night-mode-toggle .icon-theme::before{content:"\e900"}@media screen and (max-width:768px){.night-mode-toggle .icon-theme::before{font-size:1em}.night-mode-toggle{transform:translateX(0);left:5px;top:1.5em}.sidebar-opened .night-mode-toggle{left:18px}}@media screen and (min-width:769px){.night-mode-toggle .icon-theme::before{font-size:1.25em}}@media screen and (min-width:1281px){.night-mode-toggle .icon-theme::before{font-size:1.5em}}body.night-mode{background:#212127}body.night-mode .hljs-comment,body.night-mode .hljs-quote{color:#969896}body.night-mode .hljs-deletion,body.night-mode .hljs-name,body.night-mode .hljs-regexp,body.night-mode .hljs-selector-class,body.night-mode .hljs-selector-id,body.night-mode .hljs-tag,body.night-mode .hljs-template-variable,body.night-mode .hljs-variable{color:#c66}body.night-mode .hljs-built_in,body.night-mode .hljs-builtin-name,body.night-mode .hljs-link,body.night-mode .hljs-literal,body.night-mode .hljs-meta,body.night-mode .hljs-number,body.night-mode .hljs-params,body.night-mode .hljs-type{color:#de935f}body.night-mode .hljs-attribute{color:#f0c674}body.night-mode .hljs-addition,body.night-mode .hljs-bullet,body.night-mode .hljs-string,body.night-mode .hljs-symbol{color:#b5bd68}body.night-mode .hljs-section,body.night-mode .hljs-title{color:#81a2be}body.night-mode .hljs-keyword,body.night-mode .hljs-selector-tag{color:#b294bb}body.night-mode .hljs{display:block;overflow-x:auto;background:#1d1f21;color:#c5c8c6;padding:.5em}body.night-mode .hljs-emphasis{font-style:italic}body.night-mode .hljs-strong{font-weight:700}body.night-mode .content-outer{background:#212127}body.night-mode .night-mode-toggle .icon-theme::before{content:"\e901"}body.night-mode #search .result-id a:active,body.night-mode #search .result-id a:focus,body.night-mode #search .result-id a:visited{color:#D2D2D2}body.night-mode #search .result-id a:hover{color:#9768d1}body.night-mode .content-inner{color:#B4B4B4}body.night-mode .content-inner a:visited,body.night-mode .content-inner h1,body.night-mode .content-inner h2,body.night-mode .content-inner h3,body.night-mode .content-inner h4,body.night-mode .content-inner h5,body.night-mode .content-inner h6{color:#D2D2D2}body.night-mode .content-inner a{color:#D2D2D2;text-decoration:underline}body.night-mode .content-inner .summary h2 a,body.night-mode .content-inner a.no-underline,body.night-mode .content-inner a.view-source,body.night-mode .content-inner pre a{text-decoration:none}body.night-mode .content-inner a.view-source:hover{color:#fff}@keyframes night-blink-background{from{background-color:#3A4152}to{background-color:#660}}body.night-mode .content-inner .detail:target .detail-header{animation-name:night-blink-background}body.night-mode .content-inner .detail-header{background:#3A4152;color:#D2D2D2}body.night-mode .content-inner .footer,body.night-mode .content-inner .footer a{color:#959595}body.night-mode .content-inner code{background-color:#2C2C31;border-color:#44444c}body.night-mode .content-inner pre code.hljs{background-color:#2C2C31}body.night-mode .content-inner .footer .line{display:inline-block}body.night-mode .sidebar-button,body.night-mode .sidebar-closed .sidebar-button{color:#d5dae6}.sr-only{position:absolute;width:1px;height:1px;padding:0;margin:-1px;overflow:hidden;clip:rect(0,0,0,0);border:0}@media print{.main{display:block}.sidebar,.sidebar-button{display:none}.content{padding-left:0;overflow:visible}.summary-row{page-break-inside:avoid}} \ No newline at end of file diff --git a/doc/dist/app-9bd040e5e5.js b/doc/dist/app-9bd040e5e5.js deleted file mode 100644 index e4a0ccf..0000000 --- a/doc/dist/app-9bd040e5e5.js +++ /dev/null @@ -1,8 +0,0 @@ -!function(e){function t(r){if(n[r])return n[r].exports;var i=n[r]={exports:{},id:r,loaded:!1};return e[r].call(i.exports,i,i.exports,t),i.loaded=!0,i.exports}var n={};return t.m=e,t.c=n,t.p="",t(0)}([function(e,t,n){"use strict";var r=n(1)["default"],i=n(2),o=r(i),a=n(3),u=r(a),s=n(4),l=n(86),c=n(89);window.$=o["default"],(0,o["default"])(function(){u["default"].configure({tabReplace:" ",languages:[]}),(0,c.initialize)(),(0,l.initialize)(),(0,s.initialize)(),u["default"].initHighlighting()})},function(e,t){"use strict";t["default"]=function(e){return e&&e.__esModule?e:{"default":e}},t.__esModule=!0},function(e,t,n){var r,i;!function(t,n){"use strict";"object"==typeof e&&"object"==typeof e.exports?e.exports=t.document?n(t,!0):function(e){if(!e.document)throw new Error("jQuery requires a window with a document");return n(e)}:n(t)}("undefined"!=typeof window?window:this,function(n,o){"use strict";function a(e,t,n){t=t||ce;var r,i=t.createElement("script");if(i.text=e,n)for(r in ke)n[r]&&(i[r]=n[r]);t.head.appendChild(i).parentNode.removeChild(i)}function u(e){return null==e?e+"":"object"==typeof e||"function"==typeof e?ve[me.call(e)]||"object":typeof e}function s(e){var t=!!e&&"length"in e&&e.length,n=u(e);return!we(e)&&!Ee(e)&&("array"===n||0===t||"number"==typeof t&&t>0&&t-1 in e)}function l(e,t){return e.nodeName&&e.nodeName.toLowerCase()===t.toLowerCase()}function c(e,t,n){return we(t)?Ne.grep(e,function(e,r){return!!t.call(e,r,e)!==n}):t.nodeType?Ne.grep(e,function(e){return e===t!==n}):"string"!=typeof t?Ne.grep(e,function(e){return ge.call(t,e)>-1!==n}):Ne.filter(t,e,n)}function f(e,t){for(;(e=e[t])&&1!==e.nodeType;);return e}function p(e){var t={};return Ne.each(e.match(qe)||[],function(e,n){t[n]=!0}),t}function d(e){return e}function h(e){throw e}function g(e,t,n,r){var i;try{e&&we(i=e.promise)?i.call(e).done(t).fail(n):e&&we(i=e.then)?i.call(e,t,n):t.apply(void 0,[e].slice(r))}catch(e){n.apply(void 0,[e])}}function v(){ce.removeEventListener("DOMContentLoaded",v),n.removeEventListener("load",v),Ne.ready()}function m(e,t){return t.toUpperCase()}function y(e){return e.replace(ze,"ms-").replace(Fe,m)}function b(){this.expando=Ne.expando+b.uid++}function _(e){return"true"===e||"false"!==e&&("null"===e?null:e===+e+""?+e:Ge.test(e)?JSON.parse(e):e)}function x(e,t,n){var r;if(void 0===n&&1===e.nodeType)if(r="data-"+t.replace(Ve,"-$&").toLowerCase(),n=e.getAttribute(r),"string"==typeof n){try{n=_(n)}catch(i){}Ke.set(e,t,n)}else n=void 0;return n}function w(e,t,n,r){var i,o,a=20,u=r?function(){return r.cur()}:function(){return Ne.css(e,t,"")},s=u(),l=n&&n[3]||(Ne.cssNumber[t]?"":"px"),c=(Ne.cssNumber[t]||"px"!==l&&+s)&&Ze.exec(Ne.css(e,t));if(c&&c[3]!==l){for(s/=2,l=l||c[3],c=+s||1;a--;)Ne.style(e,t,c+l),(1-o)*(1-(o=u()/s||.5))<=0&&(a=0),c/=o;c=2*c,Ne.style(e,t,c+l),n=n||[]}return n&&(c=+c||+s||0,i=n[1]?c+(n[1]+1)*n[2]:+n[2],r&&(r.unit=l,r.start=c,r.end=i)),i}function E(e){var t,n=e.ownerDocument,r=e.nodeName,i=et[r];return i?i:(t=n.body.appendChild(n.createElement(r)),i=Ne.css(t,"display"),t.parentNode.removeChild(t),"none"===i&&(i="block"),et[r]=i,i)}function k(e,t){for(var n,r,i=[],o=0,a=e.length;o-1)i&&i.push(o);else if(c=Ne.contains(o.ownerDocument,o),a=C(p.appendChild(o),"script"),c&&N(a),n)for(f=0;o=a[f++];)rt.test(o.type||"")&&n.push(o);return p}function A(){return!0}function S(){return!1}function j(){try{return ce.activeElement}catch(e){}}function O(e,t,n,r,i,o){var a,u;if("object"==typeof t){"string"!=typeof n&&(r=r||n,n=void 0);for(u in t)O(e,u,n,r,t[u],o);return e}if(null==r&&null==i?(i=n,r=n=void 0):null==i&&("string"==typeof n?(i=r,r=void 0):(i=r,r=n,n=void 0)),i===!1)i=S;else if(!i)return e;return 1===o&&(a=i,i=function(e){return Ne().off(e),a.apply(this,arguments)},i.guid=a.guid||(a.guid=Ne.guid++)),e.each(function(){Ne.event.add(this,t,i,r,n)})}function M(e,t){return l(e,"table")&&l(11!==t.nodeType?t:t.firstChild,"tr")?Ne(e).children("tbody")[0]||e:e}function D(e){return e.type=(null!==e.getAttribute("type"))+"/"+e.type,e}function R(e){return"true/"===(e.type||"").slice(0,5)?e.type=e.type.slice(5):e.removeAttribute("type"),e}function P(e,t){var n,r,i,o,a,u,s,l;if(1===t.nodeType){if(Ue.hasData(e)&&(o=Ue.access(e),a=Ue.set(t,o),l=o.events)){delete a.handle,a.events={};for(i in l)for(n=0,r=l[i].length;n1&&"string"==typeof h&&!xe.checkClone&&pt.test(h))return e.each(function(i){var o=e.eq(i);g&&(t[0]=h.call(this,i,o.html())),I(o,t,n,r)});if(p&&(i=T(t,e[0].ownerDocument,!1,e,r),o=i.firstChild,1===i.childNodes.length&&(i=o),o||r)){for(u=Ne.map(C(i,"script"),D),s=u.length;f=0&&(s+=Math.max(0,Math.ceil(e["offset"+t[0].toUpperCase()+t.slice(1)]-o-s-u-.5))),s}function U(e,t,n){var r=gt(e),i=H(e,t,r),o="border-box"===Ne.css(e,"boxSizing",!1,r),a=o;if(ht.test(i)){if(!n)return i;i="auto"}return a=a&&(xe.boxSizingReliable()||i===e.style[t]),("auto"===i||!parseFloat(i)&&"inline"===Ne.css(e,"display",!1,r))&&(i=e["offset"+t[0].toUpperCase()+t.slice(1)],a=!0),i=parseFloat(i)||0,i+W(e,t,n||(o?"border":"content"),a,r,i)+"px"}function K(e,t,n,r,i){return new K.prototype.init(e,t,n,r,i)}function G(){kt&&(ce.hidden===!1&&n.requestAnimationFrame?n.requestAnimationFrame(G):n.setTimeout(G,Ne.fx.interval),Ne.fx.tick())}function V(){return n.setTimeout(function(){Et=void 0}),Et=Date.now()}function X(e,t){var n,r=0,i={height:e};for(t=t?1:0;r<4;r+=2-t)n=Qe[r],i["margin"+n]=i["padding"+n]=e;return t&&(i.opacity=i.width=e),i}function Z(e,t,n){for(var r,i=(J.tweeners[t]||[]).concat(J.tweeners["*"]),o=0,a=i.length;o=0&&nE.cacheLength&&delete e[t.shift()],e[n+" "]=r}var t=[];return e}function r(e){return e[B]=!0,e}function i(e){var t=D.createElement("fieldset");try{return!!e(t)}catch(n){return!1}finally{t.parentNode&&t.parentNode.removeChild(t),t=null}}function o(e,t){for(var n=e.split("|"),r=n.length;r--;)E.attrHandle[n[r]]=t}function a(e,t){var n=t&&e,r=n&&1===e.nodeType&&1===t.nodeType&&e.sourceIndex-t.sourceIndex;if(r)return r;if(n)for(;n=n.nextSibling;)if(n===t)return-1;return e?1:-1}function u(e){return function(t){var n=t.nodeName.toLowerCase();return"input"===n&&t.type===e}}function s(e){return function(t){var n=t.nodeName.toLowerCase();return("input"===n||"button"===n)&&t.type===e}}function l(e){return function(t){return"form"in t?t.parentNode&&t.disabled===!1?"label"in t?"label"in t.parentNode?t.parentNode.disabled===e:t.disabled===e:t.isDisabled===e||t.isDisabled!==!e&&ke(t)===e:t.disabled===e:"label"in t&&t.disabled===e}}function c(e){return r(function(t){return t=+t,r(function(n,r){for(var i,o=e([],n.length,t),a=o.length;a--;)n[i=o[a]]&&(n[i]=!(r[i]=n[i]))})})}function f(e){return e&&"undefined"!=typeof e.getElementsByTagName&&e}function p(){}function d(e){for(var t=0,n=e.length,r="";t1?function(t,n,r){for(var i=e.length;i--;)if(!e[i](t,n,r))return!1;return!0}:e[0]}function v(e,n,r){for(var i=0,o=n.length;i-1&&(r[l]=!(a[l]=f))}}else b=m(b===a?b.splice(h,b.length):b),o?o(null,a,b,s):Y.apply(a,b)})}function b(e){for(var t,n,r,i=e.length,o=E.relative[e[0].type],a=o||E.relative[" "],u=o?1:0,s=h(function(e){return e===t},a,!0),l=h(function(e){return ee(t,e)>-1},a,!0),c=[function(e,n,r){var i=!o&&(r||n!==S)||((t=n).nodeType?s(e,n,r):l(e,n,r));return t=null,i}];u1&&g(c),u>1&&d(e.slice(0,u-1).concat({value:" "===e[u-2].type?"*":""})).replace(ue,"$1"),n,u0,o=e.length>0,a=function(r,a,u,s,l){var c,f,p,d=0,h="0",g=r&&[],v=[],y=S,b=r||o&&E.find.TAG("*",l),_=z+=null==y?1:Math.random()||.1,x=b.length;for(l&&(S=a===D||a||l);h!==x&&null!=(c=b[h]);h++){if(o&&c){for(f=0,a||c.ownerDocument===D||(M(c),u=!P);p=e[f++];)if(p(c,a||D,u)){s.push(c);break}l&&(z=_)}i&&((c=!p&&c)&&d--,r&&g.push(c))}if(d+=h,i&&h!==d){for(f=0;p=n[f++];)p(g,v,a,u);if(r){if(d>0)for(;h--;)g[h]||v[h]||(v[h]=Z.call(s));v=m(v)}Y.apply(s,v),l&&!r&&v.length>0&&d+n.length>1&&t.uniqueSort(s)}return l&&(z=_,S=y),g};return i?r(a):a}var x,w,E,k,C,N,T,A,S,j,O,M,D,R,P,L,I,q,H,B="sizzle"+1*new Date,$=e.document,z=0,F=0,W=n(),U=n(),K=n(),G=function(e,t){return e===t&&(O=!0),0},V={}.hasOwnProperty,X=[],Z=X.pop,Q=X.push,Y=X.push,J=X.slice,ee=function(e,t){for(var n=0,r=e.length;n+~]|"+ne+")"+ne+"*"),ce=new RegExp("="+ne+"*([^\\]'\"]*?)"+ne+"*\\]","g"),fe=new RegExp(oe),pe=new RegExp("^"+re+"$"),de={ID:new RegExp("^#("+re+")"),CLASS:new RegExp("^\\.("+re+")"),TAG:new RegExp("^("+re+"|[*])"),ATTR:new RegExp("^"+ie),PSEUDO:new RegExp("^"+oe),CHILD:new RegExp("^:(only|first|last|nth|nth-last)-(child|of-type)(?:\\("+ne+"*(even|odd|(([+-]|)(\\d*)n|)"+ne+"*(?:([+-]|)"+ne+"*(\\d+)|))"+ne+"*\\)|)","i"),bool:new RegExp("^(?:"+te+")$","i"),needsContext:new RegExp("^"+ne+"*[>+~]|:(even|odd|eq|gt|lt|nth|first|last)(?:\\("+ne+"*((?:-\\d)?\\d*)"+ne+"*\\)|)(?=[^-]|$)","i")},he=/^(?:input|select|textarea|button)$/i,ge=/^h\d$/i,ve=/^[^{]+\{\s*\[native \w/,me=/^(?:#([\w-]+)|(\w+)|\.([\w-]+))$/,ye=/[+~]/,be=new RegExp("\\\\([\\da-f]{1,6}"+ne+"?|("+ne+")|.)","ig"),_e=function(e,t,n){var r="0x"+t-65536;return r!==r||n?t:r<0?String.fromCharCode(r+65536):String.fromCharCode(r>>10|55296,1023&r|56320)},xe=/([\0-\x1f\x7f]|^-?\d)|^-$|[^\0-\x1f\x7f-\uFFFF\w-]/g,we=function(e,t){return t?"\0"===e?"�":e.slice(0,-1)+"\\"+e.charCodeAt(e.length-1).toString(16)+" ":"\\"+e},Ee=function(){M()},ke=h(function(e){return e.disabled===!0&&("form"in e||"label"in e)},{dir:"parentNode",next:"legend"});try{Y.apply(X=J.call($.childNodes),$.childNodes),X[$.childNodes.length].nodeType}catch(Ce){Y={apply:X.length?function(e,t){Q.apply(e,J.call(t))}:function(e,t){for(var n=e.length,r=0;e[n++]=t[r++];);e.length=n-1}}}w=t.support={},C=t.isXML=function(e){var t=e&&(e.ownerDocument||e).documentElement;return!!t&&"HTML"!==t.nodeName},M=t.setDocument=function(e){var t,n,r=e?e.ownerDocument||e:$;return r!==D&&9===r.nodeType&&r.documentElement?(D=r,R=D.documentElement,P=!C(D),$!==D&&(n=D.defaultView)&&n.top!==n&&(n.addEventListener?n.addEventListener("unload",Ee,!1):n.attachEvent&&n.attachEvent("onunload",Ee)),w.attributes=i(function(e){return e.className="i",!e.getAttribute("className")}),w.getElementsByTagName=i(function(e){return e.appendChild(D.createComment("")),!e.getElementsByTagName("*").length}),w.getElementsByClassName=ve.test(D.getElementsByClassName),w.getById=i(function(e){return R.appendChild(e).id=B,!D.getElementsByName||!D.getElementsByName(B).length}),w.getById?(E.filter.ID=function(e){var t=e.replace(be,_e);return function(e){return e.getAttribute("id")===t}},E.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&P){var n=t.getElementById(e);return n?[n]:[]}}):(E.filter.ID=function(e){var t=e.replace(be,_e);return function(e){var n="undefined"!=typeof e.getAttributeNode&&e.getAttributeNode("id");return n&&n.value===t}},E.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&P){var n,r,i,o=t.getElementById(e);if(o){if(n=o.getAttributeNode("id"),n&&n.value===e)return[o];for(i=t.getElementsByName(e),r=0;o=i[r++];)if(n=o.getAttributeNode("id"),n&&n.value===e)return[o]}return[]}}),E.find.TAG=w.getElementsByTagName?function(e,t){return"undefined"!=typeof t.getElementsByTagName?t.getElementsByTagName(e):w.qsa?t.querySelectorAll(e):void 0}:function(e,t){var n,r=[],i=0,o=t.getElementsByTagName(e);if("*"===e){for(;n=o[i++];)1===n.nodeType&&r.push(n);return r}return o},E.find.CLASS=w.getElementsByClassName&&function(e,t){if("undefined"!=typeof t.getElementsByClassName&&P)return t.getElementsByClassName(e)},I=[],L=[],(w.qsa=ve.test(D.querySelectorAll))&&(i(function(e){R.appendChild(e).innerHTML="",e.querySelectorAll("[msallowcapture^='']").length&&L.push("[*^$]="+ne+"*(?:''|\"\")"),e.querySelectorAll("[selected]").length||L.push("\\["+ne+"*(?:value|"+te+")"),e.querySelectorAll("[id~="+B+"-]").length||L.push("~="),e.querySelectorAll(":checked").length||L.push(":checked"),e.querySelectorAll("a#"+B+"+*").length||L.push(".#.+[+~]")}),i(function(e){e.innerHTML="";var t=D.createElement("input");t.setAttribute("type","hidden"),e.appendChild(t).setAttribute("name","D"),e.querySelectorAll("[name=d]").length&&L.push("name"+ne+"*[*^$|!~]?="),2!==e.querySelectorAll(":enabled").length&&L.push(":enabled",":disabled"),R.appendChild(e).disabled=!0,2!==e.querySelectorAll(":disabled").length&&L.push(":enabled",":disabled"),e.querySelectorAll("*,:x"),L.push(",.*:")})),(w.matchesSelector=ve.test(q=R.matches||R.webkitMatchesSelector||R.mozMatchesSelector||R.oMatchesSelector||R.msMatchesSelector))&&i(function(e){w.disconnectedMatch=q.call(e,"*"),q.call(e,"[s!='']:x"),I.push("!=",oe)}),L=L.length&&new RegExp(L.join("|")),I=I.length&&new RegExp(I.join("|")),t=ve.test(R.compareDocumentPosition),H=t||ve.test(R.contains)?function(e,t){var n=9===e.nodeType?e.documentElement:e,r=t&&t.parentNode;return e===r||!(!r||1!==r.nodeType||!(n.contains?n.contains(r):e.compareDocumentPosition&&16&e.compareDocumentPosition(r)))}:function(e,t){if(t)for(;t=t.parentNode;)if(t===e)return!0;return!1},G=t?function(e,t){if(e===t)return O=!0,0;var n=!e.compareDocumentPosition-!t.compareDocumentPosition;return n?n:(n=(e.ownerDocument||e)===(t.ownerDocument||t)?e.compareDocumentPosition(t):1,1&n||!w.sortDetached&&t.compareDocumentPosition(e)===n?e===D||e.ownerDocument===$&&H($,e)?-1:t===D||t.ownerDocument===$&&H($,t)?1:j?ee(j,e)-ee(j,t):0:4&n?-1:1)}:function(e,t){if(e===t)return O=!0,0;var n,r=0,i=e.parentNode,o=t.parentNode,u=[e],s=[t];if(!i||!o)return e===D?-1:t===D?1:i?-1:o?1:j?ee(j,e)-ee(j,t):0;if(i===o)return a(e,t);for(n=e;n=n.parentNode;)u.unshift(n);for(n=t;n=n.parentNode;)s.unshift(n);for(;u[r]===s[r];)r++;return r?a(u[r],s[r]):u[r]===$?-1:s[r]===$?1:0},D):D},t.matches=function(e,n){return t(e,null,null,n)},t.matchesSelector=function(e,n){if((e.ownerDocument||e)!==D&&M(e),n=n.replace(ce,"='$1']"),w.matchesSelector&&P&&!K[n+" "]&&(!I||!I.test(n))&&(!L||!L.test(n)))try{var r=q.call(e,n);if(r||w.disconnectedMatch||e.document&&11!==e.document.nodeType)return r}catch(i){}return t(n,D,null,[e]).length>0},t.contains=function(e,t){return(e.ownerDocument||e)!==D&&M(e),H(e,t)},t.attr=function(e,t){(e.ownerDocument||e)!==D&&M(e);var n=E.attrHandle[t.toLowerCase()],r=n&&V.call(E.attrHandle,t.toLowerCase())?n(e,t,!P):void 0;return void 0!==r?r:w.attributes||!P?e.getAttribute(t):(r=e.getAttributeNode(t))&&r.specified?r.value:null},t.escape=function(e){return(e+"").replace(xe,we)},t.error=function(e){throw new Error("Syntax error, unrecognized expression: "+e)},t.uniqueSort=function(e){var t,n=[],r=0,i=0;if(O=!w.detectDuplicates,j=!w.sortStable&&e.slice(0),e.sort(G),O){for(;t=e[i++];)t===e[i]&&(r=n.push(i));for(;r--;)e.splice(n[r],1)}return j=null,e},k=t.getText=function(e){var t,n="",r=0,i=e.nodeType;if(i){if(1===i||9===i||11===i){if("string"==typeof e.textContent)return e.textContent;for(e=e.firstChild;e;e=e.nextSibling)n+=k(e)}else if(3===i||4===i)return e.nodeValue}else for(;t=e[r++];)n+=k(t);return n},E=t.selectors={cacheLength:50,createPseudo:r,match:de,attrHandle:{},find:{},relative:{">":{dir:"parentNode",first:!0}," ":{dir:"parentNode"},"+":{dir:"previousSibling",first:!0},"~":{dir:"previousSibling"}},preFilter:{ATTR:function(e){return e[1]=e[1].replace(be,_e),e[3]=(e[3]||e[4]||e[5]||"").replace(be,_e),"~="===e[2]&&(e[3]=" "+e[3]+" "),e.slice(0,4)},CHILD:function(e){return e[1]=e[1].toLowerCase(),"nth"===e[1].slice(0,3)?(e[3]||t.error(e[0]),e[4]=+(e[4]?e[5]+(e[6]||1):2*("even"===e[3]||"odd"===e[3])),e[5]=+(e[7]+e[8]||"odd"===e[3])):e[3]&&t.error(e[0]),e},PSEUDO:function(e){var t,n=!e[6]&&e[2];return de.CHILD.test(e[0])?null:(e[3]?e[2]=e[4]||e[5]||"":n&&fe.test(n)&&(t=N(n,!0))&&(t=n.indexOf(")",n.length-t)-n.length)&&(e[0]=e[0].slice(0,t),e[2]=n.slice(0,t)),e.slice(0,3))}},filter:{TAG:function(e){var t=e.replace(be,_e).toLowerCase();return"*"===e?function(){return!0}:function(e){return e.nodeName&&e.nodeName.toLowerCase()===t}},CLASS:function(e){var t=W[e+" "];return t||(t=new RegExp("(^|"+ne+")"+e+"("+ne+"|$)"))&&W(e,function(e){return t.test("string"==typeof e.className&&e.className||"undefined"!=typeof e.getAttribute&&e.getAttribute("class")||"")})},ATTR:function(e,n,r){return function(i){var o=t.attr(i,e);return null==o?"!="===n:!n||(o+="","="===n?o===r:"!="===n?o!==r:"^="===n?r&&0===o.indexOf(r):"*="===n?r&&o.indexOf(r)>-1:"$="===n?r&&o.slice(-r.length)===r:"~="===n?(" "+o.replace(ae," ")+" ").indexOf(r)>-1:"|="===n&&(o===r||o.slice(0,r.length+1)===r+"-"))}},CHILD:function(e,t,n,r,i){var o="nth"!==e.slice(0,3),a="last"!==e.slice(-4),u="of-type"===t;return 1===r&&0===i?function(e){return!!e.parentNode}:function(t,n,s){var l,c,f,p,d,h,g=o!==a?"nextSibling":"previousSibling",v=t.parentNode,m=u&&t.nodeName.toLowerCase(),y=!s&&!u,b=!1;if(v){if(o){for(;g;){for(p=t;p=p[g];)if(u?p.nodeName.toLowerCase()===m:1===p.nodeType)return!1;h=g="only"===e&&!h&&"nextSibling"}return!0}if(h=[a?v.firstChild:v.lastChild],a&&y){for(p=v,f=p[B]||(p[B]={}),c=f[p.uniqueID]||(f[p.uniqueID]={}),l=c[e]||[],d=l[0]===z&&l[1],b=d&&l[2],p=d&&v.childNodes[d];p=++d&&p&&p[g]||(b=d=0)||h.pop();)if(1===p.nodeType&&++b&&p===t){c[e]=[z,d,b];break}}else if(y&&(p=t,f=p[B]||(p[B]={}),c=f[p.uniqueID]||(f[p.uniqueID]={}),l=c[e]||[],d=l[0]===z&&l[1],b=d),b===!1)for(;(p=++d&&p&&p[g]||(b=d=0)||h.pop())&&((u?p.nodeName.toLowerCase()!==m:1!==p.nodeType)||!++b||(y&&(f=p[B]||(p[B]={}),c=f[p.uniqueID]||(f[p.uniqueID]={}),c[e]=[z,b]),p!==t)););return b-=i,b===r||b%r===0&&b/r>=0}}},PSEUDO:function(e,n){var i,o=E.pseudos[e]||E.setFilters[e.toLowerCase()]||t.error("unsupported pseudo: "+e);return o[B]?o(n):o.length>1?(i=[e,e,"",n],E.setFilters.hasOwnProperty(e.toLowerCase())?r(function(e,t){for(var r,i=o(e,n),a=i.length;a--;)r=ee(e,i[a]),e[r]=!(t[r]=i[a])}):function(e){return o(e,0,i)}):o}},pseudos:{not:r(function(e){var t=[],n=[],i=T(e.replace(ue,"$1"));return i[B]?r(function(e,t,n,r){for(var o,a=i(e,null,r,[]),u=e.length;u--;)(o=a[u])&&(e[u]=!(t[u]=o))}):function(e,r,o){return t[0]=e,i(t,null,o,n), -t[0]=null,!n.pop()}}),has:r(function(e){return function(n){return t(e,n).length>0}}),contains:r(function(e){return e=e.replace(be,_e),function(t){return(t.textContent||t.innerText||k(t)).indexOf(e)>-1}}),lang:r(function(e){return pe.test(e||"")||t.error("unsupported lang: "+e),e=e.replace(be,_e).toLowerCase(),function(t){var n;do if(n=P?t.lang:t.getAttribute("xml:lang")||t.getAttribute("lang"))return n=n.toLowerCase(),n===e||0===n.indexOf(e+"-");while((t=t.parentNode)&&1===t.nodeType);return!1}}),target:function(t){var n=e.location&&e.location.hash;return n&&n.slice(1)===t.id},root:function(e){return e===R},focus:function(e){return e===D.activeElement&&(!D.hasFocus||D.hasFocus())&&!!(e.type||e.href||~e.tabIndex)},enabled:l(!1),disabled:l(!0),checked:function(e){var t=e.nodeName.toLowerCase();return"input"===t&&!!e.checked||"option"===t&&!!e.selected},selected:function(e){return e.parentNode&&e.parentNode.selectedIndex,e.selected===!0},empty:function(e){for(e=e.firstChild;e;e=e.nextSibling)if(e.nodeType<6)return!1;return!0},parent:function(e){return!E.pseudos.empty(e)},header:function(e){return ge.test(e.nodeName)},input:function(e){return he.test(e.nodeName)},button:function(e){var t=e.nodeName.toLowerCase();return"input"===t&&"button"===e.type||"button"===t},text:function(e){var t;return"input"===e.nodeName.toLowerCase()&&"text"===e.type&&(null==(t=e.getAttribute("type"))||"text"===t.toLowerCase())},first:c(function(){return[0]}),last:c(function(e,t){return[t-1]}),eq:c(function(e,t,n){return[n<0?n+t:n]}),even:c(function(e,t){for(var n=0;n=0;)e.push(r);return e}),gt:c(function(e,t,n){for(var r=n<0?n+t:n;++r2&&"ID"===(a=o[0]).type&&9===t.nodeType&&P&&E.relative[o[1].type]){if(t=(E.find.ID(a.matches[0].replace(be,_e),t)||[])[0],!t)return n;l&&(t=t.parentNode),e=e.slice(o.shift().value.length)}for(i=de.needsContext.test(e)?0:o.length;i--&&(a=o[i],!E.relative[u=a.type]);)if((s=E.find[u])&&(r=s(a.matches[0].replace(be,_e),ye.test(o[0].type)&&f(t.parentNode)||t))){if(o.splice(i,1),e=r.length&&d(o),!e)return Y.apply(n,r),n;break}}return(l||T(e,c))(r,t,!P,n,!t||ye.test(e)&&f(t.parentNode)||t),n},w.sortStable=B.split("").sort(G).join("")===B,w.detectDuplicates=!!O,M(),w.sortDetached=i(function(e){return 1&e.compareDocumentPosition(D.createElement("fieldset"))}),i(function(e){return e.innerHTML="","#"===e.firstChild.getAttribute("href")})||o("type|href|height|width",function(e,t,n){if(!n)return e.getAttribute(t,"type"===t.toLowerCase()?1:2)}),w.attributes&&i(function(e){return e.innerHTML="",e.firstChild.setAttribute("value",""),""===e.firstChild.getAttribute("value")})||o("value",function(e,t,n){if(!n&&"input"===e.nodeName.toLowerCase())return e.defaultValue}),i(function(e){return null==e.getAttribute("disabled")})||o(te,function(e,t,n){var r;if(!n)return e[t]===!0?t.toLowerCase():(r=e.getAttributeNode(t))&&r.specified?r.value:null}),t}(n);Ne.find=Ae,Ne.expr=Ae.selectors,Ne.expr[":"]=Ne.expr.pseudos,Ne.uniqueSort=Ne.unique=Ae.uniqueSort,Ne.text=Ae.getText,Ne.isXMLDoc=Ae.isXML,Ne.contains=Ae.contains,Ne.escapeSelector=Ae.escape;var Se=function(e,t,n){for(var r=[],i=void 0!==n;(e=e[t])&&9!==e.nodeType;)if(1===e.nodeType){if(i&&Ne(e).is(n))break;r.push(e)}return r},je=function(e,t){for(var n=[];e;e=e.nextSibling)1===e.nodeType&&e!==t&&n.push(e);return n},Oe=Ne.expr.match.needsContext,Me=/^<([a-z][^\/\0>:\x20\t\r\n\f]*)[\x20\t\r\n\f]*\/?>(?:<\/\1>|)$/i;Ne.filter=function(e,t,n){var r=t[0];return n&&(e=":not("+e+")"),1===t.length&&1===r.nodeType?Ne.find.matchesSelector(r,e)?[r]:[]:Ne.find.matches(e,Ne.grep(t,function(e){return 1===e.nodeType}))},Ne.fn.extend({find:function(e){var t,n,r=this.length,i=this;if("string"!=typeof e)return this.pushStack(Ne(e).filter(function(){for(t=0;t1?Ne.uniqueSort(n):n},filter:function(e){return this.pushStack(c(this,e||[],!1))},not:function(e){return this.pushStack(c(this,e||[],!0))},is:function(e){return!!c(this,"string"==typeof e&&Oe.test(e)?Ne(e):e||[],!1).length}});var De,Re=/^(?:\s*(<[\w\W]+>)[^>]*|#([\w-]+))$/,Pe=Ne.fn.init=function(e,t,n){var r,i;if(!e)return this;if(n=n||De,"string"==typeof e){if(r="<"===e[0]&&">"===e[e.length-1]&&e.length>=3?[null,e,null]:Re.exec(e),!r||!r[1]&&t)return!t||t.jquery?(t||n).find(e):this.constructor(t).find(e);if(r[1]){if(t=t instanceof Ne?t[0]:t,Ne.merge(this,Ne.parseHTML(r[1],t&&t.nodeType?t.ownerDocument||t:ce,!0)),Me.test(r[1])&&Ne.isPlainObject(t))for(r in t)we(this[r])?this[r](t[r]):this.attr(r,t[r]);return this}return i=ce.getElementById(r[2]),i&&(this[0]=i,this.length=1),this}return e.nodeType?(this[0]=e,this.length=1,this):we(e)?void 0!==n.ready?n.ready(e):e(Ne):Ne.makeArray(e,this)};Pe.prototype=Ne.fn,De=Ne(ce);var Le=/^(?:parents|prev(?:Until|All))/,Ie={children:!0,contents:!0,next:!0,prev:!0};Ne.fn.extend({has:function(e){var t=Ne(e,this),n=t.length;return this.filter(function(){for(var e=0;e-1:1===n.nodeType&&Ne.find.matchesSelector(n,e))){o.push(n);break}return this.pushStack(o.length>1?Ne.uniqueSort(o):o)},index:function(e){return e?"string"==typeof e?ge.call(Ne(e),this[0]):ge.call(this,e.jquery?e[0]:e):this[0]&&this[0].parentNode?this.first().prevAll().length:-1},add:function(e,t){return this.pushStack(Ne.uniqueSort(Ne.merge(this.get(),Ne(e,t))))},addBack:function(e){return this.add(null==e?this.prevObject:this.prevObject.filter(e))}}),Ne.each({parent:function(e){var t=e.parentNode;return t&&11!==t.nodeType?t:null},parents:function(e){return Se(e,"parentNode")},parentsUntil:function(e,t,n){return Se(e,"parentNode",n)},next:function(e){return f(e,"nextSibling")},prev:function(e){return f(e,"previousSibling")},nextAll:function(e){return Se(e,"nextSibling")},prevAll:function(e){return Se(e,"previousSibling")},nextUntil:function(e,t,n){return Se(e,"nextSibling",n)},prevUntil:function(e,t,n){return Se(e,"previousSibling",n)},siblings:function(e){return je((e.parentNode||{}).firstChild,e)},children:function(e){return je(e.firstChild)},contents:function(e){return l(e,"iframe")?e.contentDocument:(l(e,"template")&&(e=e.content||e),Ne.merge([],e.childNodes))}},function(e,t){Ne.fn[e]=function(n,r){var i=Ne.map(this,t,n);return"Until"!==e.slice(-5)&&(r=n),r&&"string"==typeof r&&(i=Ne.filter(r,i)),this.length>1&&(Ie[e]||Ne.uniqueSort(i),Le.test(e)&&i.reverse()),this.pushStack(i)}});var qe=/[^\x20\t\r\n\f]+/g;Ne.Callbacks=function(e){e="string"==typeof e?p(e):Ne.extend({},e);var t,n,r,i,o=[],a=[],s=-1,l=function(){for(i=i||e.once,r=t=!0;a.length;s=-1)for(n=a.shift();++s-1;)o.splice(n,1),n<=s&&s--}),this},has:function(e){return e?Ne.inArray(e,o)>-1:o.length>0},empty:function(){return o&&(o=[]),this},disable:function(){return i=a=[],o=n="",this},disabled:function(){return!o},lock:function(){return i=a=[],n||t||(o=n=""),this},locked:function(){return!!i},fireWith:function(e,n){return i||(n=n||[],n=[e,n.slice?n.slice():n],a.push(n),t||l()),this},fire:function(){return c.fireWith(this,arguments),this},fired:function(){return!!r}};return c},Ne.extend({Deferred:function(e){var t=[["notify","progress",Ne.Callbacks("memory"),Ne.Callbacks("memory"),2],["resolve","done",Ne.Callbacks("once memory"),Ne.Callbacks("once memory"),0,"resolved"],["reject","fail",Ne.Callbacks("once memory"),Ne.Callbacks("once memory"),1,"rejected"]],r="pending",i={state:function(){return r},always:function(){return o.done(arguments).fail(arguments),this},"catch":function(e){return i.then(null,e)},pipe:function(){var e=arguments;return Ne.Deferred(function(n){Ne.each(t,function(t,r){var i=we(e[r[4]])&&e[r[4]];o[r[1]](function(){var e=i&&i.apply(this,arguments);e&&we(e.promise)?e.promise().progress(n.notify).done(n.resolve).fail(n.reject):n[r[0]+"With"](this,i?[e]:arguments)})}),e=null}).promise()},then:function(e,r,i){function o(e,t,r,i){return function(){var u=this,s=arguments,l=function(){var n,l;if(!(e=a&&(r!==h&&(u=void 0,s=[n]),t.rejectWith(u,s))}};e?c():(Ne.Deferred.getStackHook&&(c.stackTrace=Ne.Deferred.getStackHook()),n.setTimeout(c))}}var a=0;return Ne.Deferred(function(n){t[0][3].add(o(0,n,we(i)?i:d,n.notifyWith)),t[1][3].add(o(0,n,we(e)?e:d)),t[2][3].add(o(0,n,we(r)?r:h))}).promise()},promise:function(e){return null!=e?Ne.extend(e,i):i}},o={};return Ne.each(t,function(e,n){var a=n[2],u=n[5];i[n[1]]=a.add,u&&a.add(function(){r=u},t[3-e][2].disable,t[3-e][3].disable,t[0][2].lock,t[0][3].lock),a.add(n[3].fire),o[n[0]]=function(){return o[n[0]+"With"](this===o?void 0:this,arguments),this},o[n[0]+"With"]=a.fireWith}),i.promise(o),e&&e.call(o,o),o},when:function(e){var t=arguments.length,n=t,r=Array(n),i=pe.call(arguments),o=Ne.Deferred(),a=function(e){return function(n){r[e]=this,i[e]=arguments.length>1?pe.call(arguments):n,--t||o.resolveWith(r,i)}};if(t<=1&&(g(e,o.done(a(n)).resolve,o.reject,!t),"pending"===o.state()||we(i[n]&&i[n].then)))return o.then();for(;n--;)g(i[n],a(n),o.reject);return o.promise()}});var He=/^(Eval|Internal|Range|Reference|Syntax|Type|URI)Error$/;Ne.Deferred.exceptionHook=function(e,t){n.console&&n.console.warn&&e&&He.test(e.name)&&n.console.warn("jQuery.Deferred exception: "+e.message,e.stack,t)},Ne.readyException=function(e){n.setTimeout(function(){throw e})};var Be=Ne.Deferred();Ne.fn.ready=function(e){return Be.then(e)["catch"](function(e){Ne.readyException(e)}),this},Ne.extend({isReady:!1,readyWait:1,ready:function(e){(e===!0?--Ne.readyWait:Ne.isReady)||(Ne.isReady=!0,e!==!0&&--Ne.readyWait>0||Be.resolveWith(ce,[Ne]))}}),Ne.ready.then=Be.then,"complete"===ce.readyState||"loading"!==ce.readyState&&!ce.documentElement.doScroll?n.setTimeout(Ne.ready):(ce.addEventListener("DOMContentLoaded",v),n.addEventListener("load",v));var $e=function(e,t,n,r,i,o,a){var s=0,l=e.length,c=null==n;if("object"===u(n)){i=!0;for(s in n)$e(e,t,s,n[s],!0,o,a)}else if(void 0!==r&&(i=!0,we(r)||(a=!0),c&&(a?(t.call(e,r),t=null):(c=t,t=function(e,t,n){return c.call(Ne(e),n)})),t))for(;s1,null,!0)},removeData:function(e){return this.each(function(){Ke.remove(this,e)})}}),Ne.extend({queue:function(e,t,n){var r;if(e)return t=(t||"fx")+"queue",r=Ue.get(e,t),n&&(!r||Array.isArray(n)?r=Ue.access(e,t,Ne.makeArray(n)):r.push(n)),r||[]},dequeue:function(e,t){t=t||"fx";var n=Ne.queue(e,t),r=n.length,i=n.shift(),o=Ne._queueHooks(e,t),a=function(){Ne.dequeue(e,t)};"inprogress"===i&&(i=n.shift(),r--),i&&("fx"===t&&n.unshift("inprogress"),delete o.stop,i.call(e,a,o)),!r&&o&&o.empty.fire()},_queueHooks:function(e,t){var n=t+"queueHooks";return Ue.get(e,n)||Ue.access(e,n,{empty:Ne.Callbacks("once memory").add(function(){Ue.remove(e,[t+"queue",n])})})}}),Ne.fn.extend({queue:function(e,t){var n=2;return"string"!=typeof e&&(t=e,e="fx",n--),arguments.length\x20\t\r\n\f]+)/i,rt=/^$|^module$|\/(?:java|ecma)script/i,it={option:[1,""],thead:[1,"","
    "],col:[2,"","
    "],tr:[2,"","
    "],td:[3,"","
    "],_default:[0,"",""]};it.optgroup=it.option,it.tbody=it.tfoot=it.colgroup=it.caption=it.thead,it.th=it.td;var ot=/<|&#?\w+;/;!function(){var e=ce.createDocumentFragment(),t=e.appendChild(ce.createElement("div")),n=ce.createElement("input");n.setAttribute("type","radio"),n.setAttribute("checked","checked"),n.setAttribute("name","t"),t.appendChild(n),xe.checkClone=t.cloneNode(!0).cloneNode(!0).lastChild.checked,t.innerHTML="",xe.noCloneChecked=!!t.cloneNode(!0).lastChild.defaultValue}();var at=ce.documentElement,ut=/^key/,st=/^(?:mouse|pointer|contextmenu|drag|drop)|click/,lt=/^([^.]*)(?:\.(.+)|)/;Ne.event={global:{},add:function(e,t,n,r,i){var o,a,u,s,l,c,f,p,d,h,g,v=Ue.get(e);if(v)for(n.handler&&(o=n,n=o.handler,i=o.selector),i&&Ne.find.matchesSelector(at,i),n.guid||(n.guid=Ne.guid++),(s=v.events)||(s=v.events={}),(a=v.handle)||(a=v.handle=function(t){return"undefined"!=typeof Ne&&Ne.event.triggered!==t.type?Ne.event.dispatch.apply(e,arguments):void 0}),t=(t||"").match(qe)||[""],l=t.length;l--;)u=lt.exec(t[l])||[],d=g=u[1],h=(u[2]||"").split(".").sort(),d&&(f=Ne.event.special[d]||{},d=(i?f.delegateType:f.bindType)||d,f=Ne.event.special[d]||{},c=Ne.extend({type:d,origType:g,data:r,handler:n,guid:n.guid,selector:i,needsContext:i&&Ne.expr.match.needsContext.test(i),namespace:h.join(".")},o),(p=s[d])||(p=s[d]=[],p.delegateCount=0,f.setup&&f.setup.call(e,r,h,a)!==!1||e.addEventListener&&e.addEventListener(d,a)),f.add&&(f.add.call(e,c),c.handler.guid||(c.handler.guid=n.guid)),i?p.splice(p.delegateCount++,0,c):p.push(c),Ne.event.global[d]=!0)},remove:function(e,t,n,r,i){var o,a,u,s,l,c,f,p,d,h,g,v=Ue.hasData(e)&&Ue.get(e);if(v&&(s=v.events)){for(t=(t||"").match(qe)||[""],l=t.length;l--;)if(u=lt.exec(t[l])||[],d=g=u[1],h=(u[2]||"").split(".").sort(),d){for(f=Ne.event.special[d]||{},d=(r?f.delegateType:f.bindType)||d,p=s[d]||[],u=u[2]&&new RegExp("(^|\\.)"+h.join("\\.(?:.*\\.|)")+"(\\.|$)"),a=o=p.length;o--;)c=p[o],!i&&g!==c.origType||n&&n.guid!==c.guid||u&&!u.test(c.namespace)||r&&r!==c.selector&&("**"!==r||!c.selector)||(p.splice(o,1),c.selector&&p.delegateCount--,f.remove&&f.remove.call(e,c));a&&!p.length&&(f.teardown&&f.teardown.call(e,h,v.handle)!==!1||Ne.removeEvent(e,d,v.handle),delete s[d])}else for(d in s)Ne.event.remove(e,d+t[l],n,r,!0);Ne.isEmptyObject(s)&&Ue.remove(e,"handle events")}},dispatch:function(e){var t,n,r,i,o,a,u=Ne.event.fix(e),s=new Array(arguments.length),l=(Ue.get(this,"events")||{})[u.type]||[],c=Ne.event.special[u.type]||{};for(s[0]=u,t=1;t=1))for(;l!==this;l=l.parentNode||this)if(1===l.nodeType&&("click"!==e.type||l.disabled!==!0)){for(o=[],a={},n=0;n-1:Ne.find(i,this,null,[l]).length),a[i]&&o.push(r);o.length&&u.push({elem:l,handlers:o})}return l=this,s\x20\t\r\n\f]*)[^>]*)\/>/gi,ft=/\s*$/g;Ne.extend({htmlPrefilter:function(e){return e.replace(ct,"<$1>")},clone:function(e,t,n){var r,i,o,a,u=e.cloneNode(!0),s=Ne.contains(e.ownerDocument,e);if(!(xe.noCloneChecked||1!==e.nodeType&&11!==e.nodeType||Ne.isXMLDoc(e)))for(a=C(u),o=C(e),r=0,i=o.length;r0&&N(a,!s&&C(e,"script")),u},cleanData:function(e){for(var t,n,r,i=Ne.event.special,o=0;void 0!==(n=e[o]);o++)if(We(n)){if(t=n[Ue.expando]){if(t.events)for(r in t.events)i[r]?Ne.event.remove(n,r):Ne.removeEvent(n,r,t.handle);n[Ue.expando]=void 0}n[Ke.expando]&&(n[Ke.expando]=void 0)}}}),Ne.fn.extend({detach:function(e){return q(this,e,!0)},remove:function(e){return q(this,e)},text:function(e){return $e(this,function(e){return void 0===e?Ne.text(this):this.empty().each(function(){1!==this.nodeType&&11!==this.nodeType&&9!==this.nodeType||(this.textContent=e)})},null,e,arguments.length)},append:function(){return I(this,arguments,function(e){if(1===this.nodeType||11===this.nodeType||9===this.nodeType){var t=M(this,e);t.appendChild(e)}})},prepend:function(){return I(this,arguments,function(e){if(1===this.nodeType||11===this.nodeType||9===this.nodeType){var t=M(this,e);t.insertBefore(e,t.firstChild)}})},before:function(){return I(this,arguments,function(e){this.parentNode&&this.parentNode.insertBefore(e,this)})},after:function(){return I(this,arguments,function(e){this.parentNode&&this.parentNode.insertBefore(e,this.nextSibling)})},empty:function(){for(var e,t=0;null!=(e=this[t]);t++)1===e.nodeType&&(Ne.cleanData(C(e,!1)),e.textContent="");return this},clone:function(e,t){return e=null!=e&&e,t=null==t?e:t,this.map(function(){return Ne.clone(this,e,t)})},html:function(e){return $e(this,function(e){var t=this[0]||{},n=0,r=this.length;if(void 0===e&&1===t.nodeType)return t.innerHTML;if("string"==typeof e&&!ft.test(e)&&!it[(nt.exec(e)||["",""])[1].toLowerCase()]){e=Ne.htmlPrefilter(e);try{for(;n1)}}),Ne.Tween=K,K.prototype={constructor:K,init:function(e,t,n,r,i,o){this.elem=e,this.prop=n,this.easing=i||Ne.easing._default,this.options=t,this.start=this.now=this.cur(),this.end=r,this.unit=o||(Ne.cssNumber[n]?"":"px")},cur:function(){var e=K.propHooks[this.prop];return e&&e.get?e.get(this):K.propHooks._default.get(this)},run:function(e){var t,n=K.propHooks[this.prop];return this.options.duration?this.pos=t=Ne.easing[this.easing](e,this.options.duration*e,0,1,this.options.duration):this.pos=t=e,this.now=(this.end-this.start)*t+this.start,this.options.step&&this.options.step.call(this.elem,this.now,this),n&&n.set?n.set(this):K.propHooks._default.set(this),this}},K.prototype.init.prototype=K.prototype,K.propHooks={_default:{get:function(e){var t;return 1!==e.elem.nodeType||null!=e.elem[e.prop]&&null==e.elem.style[e.prop]?e.elem[e.prop]:(t=Ne.css(e.elem,e.prop,""),t&&"auto"!==t?t:0)},set:function(e){Ne.fx.step[e.prop]?Ne.fx.step[e.prop](e):1!==e.elem.nodeType||null==e.elem.style[Ne.cssProps[e.prop]]&&!Ne.cssHooks[e.prop]?e.elem[e.prop]=e.now:Ne.style(e.elem,e.prop,e.now+e.unit)}}},K.propHooks.scrollTop=K.propHooks.scrollLeft={set:function(e){e.elem.nodeType&&e.elem.parentNode&&(e.elem[e.prop]=e.now)}},Ne.easing={linear:function(e){return e},swing:function(e){return.5-Math.cos(e*Math.PI)/2},_default:"swing"},Ne.fx=K.prototype.init,Ne.fx.step={};var Et,kt,Ct=/^(?:toggle|show|hide)$/,Nt=/queueHooks$/;Ne.Animation=Ne.extend(J,{tweeners:{"*":[function(e,t){var n=this.createTween(e,t);return w(n.elem,e,Ze.exec(t),n),n}]},tweener:function(e,t){we(e)?(t=e,e=["*"]):e=e.match(qe);for(var n,r=0,i=e.length;r1)},removeAttr:function(e){return this.each(function(){Ne.removeAttr(this,e)})}}),Ne.extend({attr:function(e,t,n){var r,i,o=e.nodeType;if(3!==o&&8!==o&&2!==o)return"undefined"==typeof e.getAttribute?Ne.prop(e,t,n):(1===o&&Ne.isXMLDoc(e)||(i=Ne.attrHooks[t.toLowerCase()]||(Ne.expr.match.bool.test(t)?Tt:void 0)),void 0!==n?null===n?void Ne.removeAttr(e,t):i&&"set"in i&&void 0!==(r=i.set(e,n,t))?r:(e.setAttribute(t,n+""),n):i&&"get"in i&&null!==(r=i.get(e,t))?r:(r=Ne.find.attr(e,t),null==r?void 0:r))},attrHooks:{type:{set:function(e,t){if(!xe.radioValue&&"radio"===t&&l(e,"input")){var n=e.value;return e.setAttribute("type",t),n&&(e.value=n),t}}}},removeAttr:function(e,t){var n,r=0,i=t&&t.match(qe);if(i&&1===e.nodeType)for(;n=i[r++];)e.removeAttribute(n)}}),Tt={set:function(e,t,n){return t===!1?Ne.removeAttr(e,n):e.setAttribute(n,n),n}},Ne.each(Ne.expr.match.bool.source.match(/\w+/g),function(e,t){var n=At[t]||Ne.find.attr;At[t]=function(e,t,r){var i,o,a=t.toLowerCase();return r||(o=At[a],At[a]=i,i=null!=n(e,t,r)?a:null,At[a]=o),i}});var St=/^(?:input|select|textarea|button)$/i,jt=/^(?:a|area)$/i;Ne.fn.extend({prop:function(e,t){return $e(this,Ne.prop,e,t,arguments.length>1)},removeProp:function(e){return this.each(function(){delete this[Ne.propFix[e]||e]})}}),Ne.extend({prop:function(e,t,n){var r,i,o=e.nodeType;if(3!==o&&8!==o&&2!==o)return 1===o&&Ne.isXMLDoc(e)||(t=Ne.propFix[t]||t,i=Ne.propHooks[t]),void 0!==n?i&&"set"in i&&void 0!==(r=i.set(e,n,t))?r:e[t]=n:i&&"get"in i&&null!==(r=i.get(e,t))?r:e[t]},propHooks:{tabIndex:{get:function(e){var t=Ne.find.attr(e,"tabindex");return t?parseInt(t,10):St.test(e.nodeName)||jt.test(e.nodeName)&&e.href?0:-1}}},propFix:{"for":"htmlFor","class":"className"}}),xe.optSelected||(Ne.propHooks.selected={get:function(e){var t=e.parentNode;return t&&t.parentNode&&t.parentNode.selectedIndex,null},set:function(e){var t=e.parentNode;t&&(t.selectedIndex,t.parentNode&&t.parentNode.selectedIndex)}}),Ne.each(["tabIndex","readOnly","maxLength","cellSpacing","cellPadding","rowSpan","colSpan","useMap","frameBorder","contentEditable"],function(){Ne.propFix[this.toLowerCase()]=this}),Ne.fn.extend({addClass:function(e){var t,n,r,i,o,a,u,s=0;if(we(e))return this.each(function(t){Ne(this).addClass(e.call(this,t,te(this)))});if(t=ne(e),t.length)for(;n=this[s++];)if(i=te(n),r=1===n.nodeType&&" "+ee(i)+" "){for(a=0;o=t[a++];)r.indexOf(" "+o+" ")<0&&(r+=o+" ");u=ee(r),i!==u&&n.setAttribute("class",u)}return this},removeClass:function(e){var t,n,r,i,o,a,u,s=0;if(we(e))return this.each(function(t){Ne(this).removeClass(e.call(this,t,te(this)))});if(!arguments.length)return this.attr("class","");if(t=ne(e),t.length)for(;n=this[s++];)if(i=te(n),r=1===n.nodeType&&" "+ee(i)+" "){for(a=0;o=t[a++];)for(;r.indexOf(" "+o+" ")>-1;)r=r.replace(" "+o+" "," ");u=ee(r),i!==u&&n.setAttribute("class",u)}return this},toggleClass:function(e,t){var n=typeof e,r="string"===n||Array.isArray(e);return"boolean"==typeof t&&r?t?this.addClass(e):this.removeClass(e):we(e)?this.each(function(n){Ne(this).toggleClass(e.call(this,n,te(this),t),t)}):this.each(function(){var t,i,o,a;if(r)for(i=0,o=Ne(this),a=ne(e);t=a[i++];)o.hasClass(t)?o.removeClass(t):o.addClass(t);else void 0!==e&&"boolean"!==n||(t=te(this),t&&Ue.set(this,"__className__",t),this.setAttribute&&this.setAttribute("class",t||e===!1?"":Ue.get(this,"__className__")||""))})},hasClass:function(e){var t,n,r=0;for(t=" "+e+" ";n=this[r++];)if(1===n.nodeType&&(" "+ee(te(n))+" ").indexOf(t)>-1)return!0;return!1}});var Ot=/\r/g;Ne.fn.extend({val:function(e){var t,n,r,i=this[0];{if(arguments.length)return r=we(e),this.each(function(n){var i;1===this.nodeType&&(i=r?e.call(this,n,Ne(this).val()):e,null==i?i="":"number"==typeof i?i+="":Array.isArray(i)&&(i=Ne.map(i,function(e){return null==e?"":e+""})),t=Ne.valHooks[this.type]||Ne.valHooks[this.nodeName.toLowerCase()],t&&"set"in t&&void 0!==t.set(this,i,"value")||(this.value=i))});if(i)return t=Ne.valHooks[i.type]||Ne.valHooks[i.nodeName.toLowerCase()],t&&"get"in t&&void 0!==(n=t.get(i,"value"))?n:(n=i.value,"string"==typeof n?n.replace(Ot,""):null==n?"":n)}}}),Ne.extend({valHooks:{option:{get:function(e){var t=Ne.find.attr(e,"value");return null!=t?t:ee(Ne.text(e))}},select:{get:function(e){var t,n,r,i=e.options,o=e.selectedIndex,a="select-one"===e.type,u=a?null:[],s=a?o+1:i.length;for(r=o<0?s:a?o:0;r-1)&&(n=!0);return n||(e.selectedIndex=-1),o}}}}),Ne.each(["radio","checkbox"],function(){Ne.valHooks[this]={set:function(e,t){if(Array.isArray(t))return e.checked=Ne.inArray(Ne(e).val(),t)>-1}},xe.checkOn||(Ne.valHooks[this].get=function(e){return null===e.getAttribute("value")?"on":e.value})}),xe.focusin="onfocusin"in n;var Mt=/^(?:focusinfocus|focusoutblur)$/,Dt=function(e){e.stopPropagation()};Ne.extend(Ne.event,{trigger:function(e,t,r,i){var o,a,u,s,l,c,f,p,d=[r||ce],h=ye.call(e,"type")?e.type:e,g=ye.call(e,"namespace")?e.namespace.split("."):[];if(a=p=u=r=r||ce,3!==r.nodeType&&8!==r.nodeType&&!Mt.test(h+Ne.event.triggered)&&(h.indexOf(".")>-1&&(g=h.split("."),h=g.shift(),g.sort()),l=h.indexOf(":")<0&&"on"+h,e=e[Ne.expando]?e:new Ne.Event(h,"object"==typeof e&&e),e.isTrigger=i?2:3,e.namespace=g.join("."),e.rnamespace=e.namespace?new RegExp("(^|\\.)"+g.join("\\.(?:.*\\.|)")+"(\\.|$)"):null,e.result=void 0,e.target||(e.target=r),t=null==t?[e]:Ne.makeArray(t,[e]),f=Ne.event.special[h]||{},i||!f.trigger||f.trigger.apply(r,t)!==!1)){if(!i&&!f.noBubble&&!Ee(r)){for(s=f.delegateType||h,Mt.test(s+h)||(a=a.parentNode);a;a=a.parentNode)d.push(a),u=a;u===(r.ownerDocument||ce)&&d.push(u.defaultView||u.parentWindow||n)}for(o=0;(a=d[o++])&&!e.isPropagationStopped();)p=a,e.type=o>1?s:f.bindType||h,c=(Ue.get(a,"events")||{})[e.type]&&Ue.get(a,"handle"),c&&c.apply(a,t),c=l&&a[l],c&&c.apply&&We(a)&&(e.result=c.apply(a,t),e.result===!1&&e.preventDefault());return e.type=h,i||e.isDefaultPrevented()||f._default&&f._default.apply(d.pop(),t)!==!1||!We(r)||l&&we(r[h])&&!Ee(r)&&(u=r[l],u&&(r[l]=null),Ne.event.triggered=h,e.isPropagationStopped()&&p.addEventListener(h,Dt),r[h](),e.isPropagationStopped()&&p.removeEventListener(h,Dt),Ne.event.triggered=void 0,u&&(r[l]=u)),e.result}},simulate:function(e,t,n){var r=Ne.extend(new Ne.Event,n,{type:e,isSimulated:!0});Ne.event.trigger(r,null,t)}}),Ne.fn.extend({trigger:function(e,t){return this.each(function(){Ne.event.trigger(e,t,this)})},triggerHandler:function(e,t){var n=this[0];if(n)return Ne.event.trigger(e,t,n,!0)}}),xe.focusin||Ne.each({focus:"focusin",blur:"focusout"},function(e,t){var n=function(e){Ne.event.simulate(t,e.target,Ne.event.fix(e))};Ne.event.special[t]={setup:function(){var r=this.ownerDocument||this,i=Ue.access(r,t);i||r.addEventListener(e,n,!0),Ue.access(r,t,(i||0)+1)},teardown:function(){var r=this.ownerDocument||this,i=Ue.access(r,t)-1;i?Ue.access(r,t,i):(r.removeEventListener(e,n,!0),Ue.remove(r,t))}}});var Rt=n.location,Pt=Date.now(),Lt=/\?/;Ne.parseXML=function(e){var t;if(!e||"string"!=typeof e)return null;try{t=(new n.DOMParser).parseFromString(e,"text/xml")}catch(r){t=void 0}return t&&!t.getElementsByTagName("parsererror").length||Ne.error("Invalid XML: "+e),t};var It=/\[\]$/,qt=/\r?\n/g,Ht=/^(?:submit|button|image|reset|file)$/i,Bt=/^(?:input|select|textarea|keygen)/i;Ne.param=function(e,t){var n,r=[],i=function(e,t){var n=we(t)?t():t;r[r.length]=encodeURIComponent(e)+"="+encodeURIComponent(null==n?"":n)};if(Array.isArray(e)||e.jquery&&!Ne.isPlainObject(e))Ne.each(e,function(){i(this.name,this.value)});else for(n in e)re(n,e[n],t,i);return r.join("&")},Ne.fn.extend({serialize:function(){return Ne.param(this.serializeArray())},serializeArray:function(){return this.map(function(){var e=Ne.prop(this,"elements");return e?Ne.makeArray(e):this}).filter(function(){var e=this.type;return this.name&&!Ne(this).is(":disabled")&&Bt.test(this.nodeName)&&!Ht.test(e)&&(this.checked||!tt.test(e))}).map(function(e,t){var n=Ne(this).val();return null==n?null:Array.isArray(n)?Ne.map(n,function(e){return{name:t.name,value:e.replace(qt,"\r\n")}}):{name:t.name,value:n.replace(qt,"\r\n")}}).get()}});var $t=/%20/g,zt=/#.*$/,Ft=/([?&])_=[^&]*/,Wt=/^(.*?):[ \t]*([^\r\n]*)$/gm,Ut=/^(?:about|app|app-storage|.+-extension|file|res|widget):$/,Kt=/^(?:GET|HEAD)$/,Gt=/^\/\//,Vt={},Xt={},Zt="*/".concat("*"),Qt=ce.createElement("a");Qt.href=Rt.href,Ne.extend({active:0,lastModified:{},etag:{},ajaxSettings:{url:Rt.href,type:"GET",isLocal:Ut.test(Rt.protocol),global:!0,processData:!0,async:!0,contentType:"application/x-www-form-urlencoded; charset=UTF-8",accepts:{"*":Zt,text:"text/plain",html:"text/html",xml:"application/xml, text/xml",json:"application/json, text/javascript"},contents:{xml:/\bxml\b/,html:/\bhtml/,json:/\bjson\b/},responseFields:{xml:"responseXML",text:"responseText",json:"responseJSON"},converters:{"* text":String,"text html":!0,"text json":JSON.parse,"text xml":Ne.parseXML},flatOptions:{url:!0,context:!0}},ajaxSetup:function(e,t){return t?ae(ae(e,Ne.ajaxSettings),t):ae(Ne.ajaxSettings,e)},ajaxPrefilter:ie(Vt),ajaxTransport:ie(Xt),ajax:function(e,t){function r(e,t,r,u){var l,p,d,_,x,w=t;c||(c=!0,s&&n.clearTimeout(s),i=void 0,a=u||"",E.readyState=e>0?4:0,l=e>=200&&e<300||304===e,r&&(_=ue(h,E,r)),_=se(h,_,E,l),l?(h.ifModified&&(x=E.getResponseHeader("Last-Modified"),x&&(Ne.lastModified[o]=x),x=E.getResponseHeader("etag"),x&&(Ne.etag[o]=x)),204===e||"HEAD"===h.type?w="nocontent":304===e?w="notmodified":(w=_.state,p=_.data,d=_.error,l=!d)):(d=w,!e&&w||(w="error",e<0&&(e=0))),E.status=e,E.statusText=(t||w)+"",l?m.resolveWith(g,[p,w,E]):m.rejectWith(g,[E,w,d]),E.statusCode(b),b=void 0,f&&v.trigger(l?"ajaxSuccess":"ajaxError",[E,h,l?p:d]),y.fireWith(g,[E,w]),f&&(v.trigger("ajaxComplete",[E,h]),--Ne.active||Ne.event.trigger("ajaxStop")))}"object"==typeof e&&(t=e,e=void 0),t=t||{};var i,o,a,u,s,l,c,f,p,d,h=Ne.ajaxSetup({},t),g=h.context||h,v=h.context&&(g.nodeType||g.jquery)?Ne(g):Ne.event,m=Ne.Deferred(),y=Ne.Callbacks("once memory"),b=h.statusCode||{},_={},x={},w="canceled",E={readyState:0,getResponseHeader:function(e){var t;if(c){if(!u)for(u={};t=Wt.exec(a);)u[t[1].toLowerCase()]=t[2];t=u[e.toLowerCase()]}return null==t?null:t},getAllResponseHeaders:function(){return c?a:null},setRequestHeader:function(e,t){return null==c&&(e=x[e.toLowerCase()]=x[e.toLowerCase()]||e,_[e]=t),this},overrideMimeType:function(e){return null==c&&(h.mimeType=e),this},statusCode:function(e){var t;if(e)if(c)E.always(e[E.status]);else for(t in e)b[t]=[b[t],e[t]];return this},abort:function(e){var t=e||w;return i&&i.abort(t),r(0,t),this}};if(m.promise(E),h.url=((e||h.url||Rt.href)+"").replace(Gt,Rt.protocol+"//"),h.type=t.method||t.type||h.method||h.type,h.dataTypes=(h.dataType||"*").toLowerCase().match(qe)||[""],null==h.crossDomain){l=ce.createElement("a");try{l.href=h.url,l.href=l.href,h.crossDomain=Qt.protocol+"//"+Qt.host!=l.protocol+"//"+l.host}catch(k){h.crossDomain=!0}}if(h.data&&h.processData&&"string"!=typeof h.data&&(h.data=Ne.param(h.data,h.traditional)),oe(Vt,h,t,E),c)return E;f=Ne.event&&h.global,f&&0===Ne.active++&&Ne.event.trigger("ajaxStart"),h.type=h.type.toUpperCase(),h.hasContent=!Kt.test(h.type),o=h.url.replace(zt,""),h.hasContent?h.data&&h.processData&&0===(h.contentType||"").indexOf("application/x-www-form-urlencoded")&&(h.data=h.data.replace($t,"+")):(d=h.url.slice(o.length),h.data&&(h.processData||"string"==typeof h.data)&&(o+=(Lt.test(o)?"&":"?")+h.data,delete h.data),h.cache===!1&&(o=o.replace(Ft,"$1"),d=(Lt.test(o)?"&":"?")+"_="+Pt++ +d),h.url=o+d),h.ifModified&&(Ne.lastModified[o]&&E.setRequestHeader("If-Modified-Since",Ne.lastModified[o]),Ne.etag[o]&&E.setRequestHeader("If-None-Match",Ne.etag[o])),(h.data&&h.hasContent&&h.contentType!==!1||t.contentType)&&E.setRequestHeader("Content-Type",h.contentType),E.setRequestHeader("Accept",h.dataTypes[0]&&h.accepts[h.dataTypes[0]]?h.accepts[h.dataTypes[0]]+("*"!==h.dataTypes[0]?", "+Zt+"; q=0.01":""):h.accepts["*"]);for(p in h.headers)E.setRequestHeader(p,h.headers[p]);if(h.beforeSend&&(h.beforeSend.call(g,E,h)===!1||c))return E.abort();if(w="abort",y.add(h.complete),E.done(h.success),E.fail(h.error),i=oe(Xt,h,t,E)){if(E.readyState=1,f&&v.trigger("ajaxSend",[E,h]),c)return E;h.async&&h.timeout>0&&(s=n.setTimeout(function(){E.abort("timeout")},h.timeout));try{c=!1,i.send(_,r)}catch(k){if(c)throw k;r(-1,k)}}else r(-1,"No Transport");return E},getJSON:function(e,t,n){return Ne.get(e,t,n,"json")},getScript:function(e,t){return Ne.get(e,void 0,t,"script")}}),Ne.each(["get","post"],function(e,t){Ne[t]=function(e,n,r,i){return we(n)&&(i=i||r,r=n,n=void 0),Ne.ajax(Ne.extend({url:e,type:t,dataType:i,data:n,success:r},Ne.isPlainObject(e)&&e))}}),Ne._evalUrl=function(e){return Ne.ajax({url:e,type:"GET",dataType:"script",cache:!0,async:!1,global:!1,"throws":!0})},Ne.fn.extend({wrapAll:function(e){var t;return this[0]&&(we(e)&&(e=e.call(this[0])),t=Ne(e,this[0].ownerDocument).eq(0).clone(!0),this[0].parentNode&&t.insertBefore(this[0]),t.map(function(){for(var e=this;e.firstElementChild;)e=e.firstElementChild;return e}).append(this)),this},wrapInner:function(e){return we(e)?this.each(function(t){Ne(this).wrapInner(e.call(this,t))}):this.each(function(){var t=Ne(this),n=t.contents();n.length?n.wrapAll(e):t.append(e)})},wrap:function(e){var t=we(e);return this.each(function(n){Ne(this).wrapAll(t?e.call(this,n):e)})},unwrap:function(e){return this.parent(e).not("body").each(function(){Ne(this).replaceWith(this.childNodes)}),this}}),Ne.expr.pseudos.hidden=function(e){return!Ne.expr.pseudos.visible(e)},Ne.expr.pseudos.visible=function(e){return!!(e.offsetWidth||e.offsetHeight||e.getClientRects().length)},Ne.ajaxSettings.xhr=function(){try{return new n.XMLHttpRequest}catch(e){}};var Yt={0:200,1223:204},Jt=Ne.ajaxSettings.xhr();xe.cors=!!Jt&&"withCredentials"in Jt,xe.ajax=Jt=!!Jt,Ne.ajaxTransport(function(e){var t,r;if(xe.cors||Jt&&!e.crossDomain)return{send:function(i,o){var a,u=e.xhr();if(u.open(e.type,e.url,e.async,e.username,e.password),e.xhrFields)for(a in e.xhrFields)u[a]=e.xhrFields[a];e.mimeType&&u.overrideMimeType&&u.overrideMimeType(e.mimeType),e.crossDomain||i["X-Requested-With"]||(i["X-Requested-With"]="XMLHttpRequest");for(a in i)u.setRequestHeader(a,i[a]);t=function(e){return function(){t&&(t=r=u.onload=u.onerror=u.onabort=u.ontimeout=u.onreadystatechange=null,"abort"===e?u.abort():"error"===e?"number"!=typeof u.status?o(0,"error"):o(u.status,u.statusText):o(Yt[u.status]||u.status,u.statusText,"text"!==(u.responseType||"text")||"string"!=typeof u.responseText?{binary:u.response}:{text:u.responseText},u.getAllResponseHeaders()))}},u.onload=t(),r=u.onerror=u.ontimeout=t("error"),void 0!==u.onabort?u.onabort=r:u.onreadystatechange=function(){4===u.readyState&&n.setTimeout(function(){t&&r()})},t=t("abort");try{u.send(e.hasContent&&e.data||null)}catch(s){if(t)throw s}},abort:function(){t&&t()}}}),Ne.ajaxPrefilter(function(e){e.crossDomain&&(e.contents.script=!1)}),Ne.ajaxSetup({accepts:{script:"text/javascript, application/javascript, application/ecmascript, application/x-ecmascript"},contents:{script:/\b(?:java|ecma)script\b/},converters:{"text script":function(e){return Ne.globalEval(e),e}}}),Ne.ajaxPrefilter("script",function(e){void 0===e.cache&&(e.cache=!1),e.crossDomain&&(e.type="GET")}),Ne.ajaxTransport("script",function(e){if(e.crossDomain){var t,n;return{send:function(r,i){t=Ne("",returnEnd:!0,subLanguage:["actionscript","javascript","handlebars","xml"]}},{className:"meta",variants:[{begin:/<\?xml/,end:/\?>/,relevance:10},{begin:/<\?\w+/,end:/\?>/}]},{className:"tag",begin:"",contains:[{className:"name",begin:/[^\/><\s]+/,relevance:0},n]}]}}),e.registerLanguage("markdown",function(e){return{aliases:["md","mkdown","mkd"],contains:[{className:"section",variants:[{begin:"^#{1,6}",end:"$"},{begin:"^.+?\\n[=-]{2,}$"}]},{begin:"<",end:">",subLanguage:"xml",relevance:0},{className:"bullet",begin:"^([*+-]|(\\d+\\.))\\s+"},{className:"strong",begin:"[*_]{2}.+?[*_]{2}"},{className:"emphasis",variants:[{begin:"\\*.+?\\*"},{begin:"_.+?_",relevance:0}]},{className:"quote",begin:"^>\\s+",end:"$"},{className:"code",variants:[{begin:"^```w*s*$",end:"^```s*$"},{begin:"`.+?`"},{begin:"^( {4}|\t)",end:"$",relevance:0}]},{begin:"^[-\\*]{3,}",end:"$"},{begin:"\\[.+?\\][\\(\\[].*?[\\)\\]]",returnBegin:!0,contains:[{className:"string",begin:"\\[",end:"\\]",excludeBegin:!0,returnEnd:!0,relevance:0},{className:"link",begin:"\\]\\(",end:"\\)",excludeBegin:!0,excludeEnd:!0},{className:"symbol",begin:"\\]\\[",end:"\\]",excludeBegin:!0,excludeEnd:!0}],relevance:10},{begin:/^\[[^\n]+\]:/,returnBegin:!0,contains:[{className:"symbol",begin:/\[/,end:/\]/,excludeBegin:!0,excludeEnd:!0},{className:"link",begin:/:\s*/,end:/$/,excludeBegin:!0}]}]}}),e.registerLanguage("sql",function(e){var t=e.COMMENT("--","$");return{case_insensitive:!0,illegal:/[<>{}*#]/,contains:[{beginKeywords:"begin end start commit rollback savepoint lock alter create drop rename call delete do handler insert load replace select truncate update set show pragma grant merge describe use explain help declare prepare execute deallocate release unlock purge reset change stop analyze cache flush optimize repair kill install uninstall checksum restore check backup revoke comment",end:/;/,endsWithParent:!0,lexemes:/[\w\.]+/,keywords:{keyword:"abort abs absolute acc acce accep accept access accessed accessible account acos action activate add addtime admin administer advanced advise aes_decrypt aes_encrypt after agent aggregate ali alia alias allocate allow alter always analyze ancillary and any anydata anydataset anyschema anytype apply archive archived archivelog are as asc ascii asin assembly assertion associate asynchronous at atan atn2 attr attri attrib attribu attribut attribute attributes audit authenticated authentication authid authors auto autoallocate autodblink autoextend automatic availability avg backup badfile basicfile before begin beginning benchmark between bfile bfile_base big bigfile bin binary_double binary_float binlog bit_and bit_count bit_length bit_or bit_xor bitmap blob_base block blocksize body both bound buffer_cache buffer_pool build bulk by byte byteordermark bytes cache caching call calling cancel capacity cascade cascaded case cast catalog category ceil ceiling chain change changed char_base char_length character_length characters characterset charindex charset charsetform charsetid check checksum checksum_agg child choose chr chunk class cleanup clear client clob clob_base clone close cluster_id cluster_probability cluster_set clustering coalesce coercibility col collate collation collect colu colum column column_value columns columns_updated comment commit compact compatibility compiled complete composite_limit compound compress compute concat concat_ws concurrent confirm conn connec connect connect_by_iscycle connect_by_isleaf connect_by_root connect_time connection consider consistent constant constraint constraints constructor container content contents context contributors controlfile conv convert convert_tz corr corr_k corr_s corresponding corruption cos cost count count_big counted covar_pop covar_samp cpu_per_call cpu_per_session crc32 create creation critical cross cube cume_dist curdate current current_date current_time current_timestamp current_user cursor curtime customdatum cycle data database databases datafile datafiles datalength date_add date_cache date_format date_sub dateadd datediff datefromparts datename datepart datetime2fromparts day day_to_second dayname dayofmonth dayofweek dayofyear days db_role_change dbtimezone ddl deallocate declare decode decompose decrement decrypt deduplicate def defa defau defaul default defaults deferred defi defin define degrees delayed delegate delete delete_all delimited demand dense_rank depth dequeue des_decrypt des_encrypt des_key_file desc descr descri describ describe descriptor deterministic diagnostics difference dimension direct_load directory disable disable_all disallow disassociate discardfile disconnect diskgroup distinct distinctrow distribute distributed div do document domain dotnet double downgrade drop dumpfile duplicate duration each edition editionable editions element ellipsis else elsif elt empty enable enable_all enclosed encode encoding encrypt end end-exec endian enforced engine engines enqueue enterprise entityescaping eomonth error errors escaped evalname evaluate event eventdata events except exception exceptions exchange exclude excluding execu execut execute exempt exists exit exp expire explain export export_set extended extent external external_1 external_2 externally extract failed failed_login_attempts failover failure far fast feature_set feature_value fetch field fields file file_name_convert filesystem_like_logging final finish first first_value fixed flash_cache flashback floor flush following follows for forall force form forma format found found_rows freelist freelists freepools fresh from from_base64 from_days ftp full function general generated get get_format get_lock getdate getutcdate global global_name globally go goto grant grants greatest group group_concat group_id grouping grouping_id groups gtid_subtract guarantee guard handler hash hashkeys having hea head headi headin heading heap help hex hierarchy high high_priority hosts hour http id ident_current ident_incr ident_seed identified identity idle_time if ifnull ignore iif ilike ilm immediate import in include including increment index indexes indexing indextype indicator indices inet6_aton inet6_ntoa inet_aton inet_ntoa infile initial initialized initially initrans inmemory inner innodb input insert install instance instantiable instr interface interleaved intersect into invalidate invisible is is_free_lock is_ipv4 is_ipv4_compat is_not is_not_null is_used_lock isdate isnull isolation iterate java join json json_exists keep keep_duplicates key keys kill language large last last_day last_insert_id last_value lax lcase lead leading least leaves left len lenght length less level levels library like like2 like4 likec limit lines link list listagg little ln load load_file lob lobs local localtime localtimestamp locate locator lock locked log log10 log2 logfile logfiles logging logical logical_reads_per_call logoff logon logs long loop low low_priority lower lpad lrtrim ltrim main make_set makedate maketime managed management manual map mapping mask master master_pos_wait match matched materialized max maxextents maximize maxinstances maxlen maxlogfiles maxloghistory maxlogmembers maxsize maxtrans md5 measures median medium member memcompress memory merge microsecond mid migration min minextents minimum mining minus minute minvalue missing mod mode model modification modify module monitoring month months mount move movement multiset mutex name name_const names nan national native natural nav nchar nclob nested never new newline next nextval no no_write_to_binlog noarchivelog noaudit nobadfile nocheck nocompress nocopy nocycle nodelay nodiscardfile noentityescaping noguarantee nokeep nologfile nomapping nomaxvalue nominimize nominvalue nomonitoring none noneditionable nonschema noorder nopr nopro noprom nopromp noprompt norely noresetlogs noreverse normal norowdependencies noschemacheck noswitch not nothing notice notrim novalidate now nowait nth_value nullif nulls num numb numbe nvarchar nvarchar2 object ocicoll ocidate ocidatetime ociduration ociinterval ociloblocator ocinumber ociref ocirefcursor ocirowid ocistring ocitype oct octet_length of off offline offset oid oidindex old on online only opaque open operations operator optimal optimize option optionally or oracle oracle_date oradata ord ordaudio orddicom orddoc order ordimage ordinality ordvideo organization orlany orlvary out outer outfile outline output over overflow overriding package pad parallel parallel_enable parameters parent parse partial partition partitions pascal passing password password_grace_time password_lock_time password_reuse_max password_reuse_time password_verify_function patch path patindex pctincrease pctthreshold pctused pctversion percent percent_rank percentile_cont percentile_disc performance period period_add period_diff permanent physical pi pipe pipelined pivot pluggable plugin policy position post_transaction pow power pragma prebuilt precedes preceding precision prediction prediction_cost prediction_details prediction_probability prediction_set prepare present preserve prior priority private private_sga privileges procedural procedure procedure_analyze processlist profiles project prompt protection public publishingservername purge quarter query quick quiesce quota quotename radians raise rand range rank raw read reads readsize rebuild record records recover recovery recursive recycle redo reduced ref reference referenced references referencing refresh regexp_like register regr_avgx regr_avgy regr_count regr_intercept regr_r2 regr_slope regr_sxx regr_sxy reject rekey relational relative relaylog release release_lock relies_on relocate rely rem remainder rename repair repeat replace replicate replication required reset resetlogs resize resource respect restore restricted result result_cache resumable resume retention return returning returns reuse reverse revoke right rlike role roles rollback rolling rollup round row row_count rowdependencies rowid rownum rows rtrim rules safe salt sample save savepoint sb1 sb2 sb4 scan schema schemacheck scn scope scroll sdo_georaster sdo_topo_geometry search sec_to_time second section securefile security seed segment select self sequence sequential serializable server servererror session session_user sessions_per_user set sets settings sha sha1 sha2 share shared shared_pool short show shrink shutdown si_averagecolor si_colorhistogram si_featurelist si_positionalcolor si_stillimage si_texture siblings sid sign sin size size_t sizes skip slave sleep smalldatetimefromparts smallfile snapshot some soname sort soundex source space sparse spfile split sql sql_big_result sql_buffer_result sql_cache sql_calc_found_rows sql_small_result sql_variant_property sqlcode sqldata sqlerror sqlname sqlstate sqrt square standalone standby start starting startup statement static statistics stats_binomial_test stats_crosstab stats_ks_test stats_mode stats_mw_test stats_one_way_anova stats_t_test_ stats_t_test_indep stats_t_test_one stats_t_test_paired stats_wsr_test status std stddev stddev_pop stddev_samp stdev stop storage store stored str str_to_date straight_join strcmp strict string struct stuff style subdate subpartition subpartitions substitutable substr substring subtime subtring_index subtype success sum suspend switch switchoffset switchover sync synchronous synonym sys sys_xmlagg sysasm sysaux sysdate sysdatetimeoffset sysdba sysoper system system_user sysutcdatetime table tables tablespace tan tdo template temporary terminated tertiary_weights test than then thread through tier ties time time_format time_zone timediff timefromparts timeout timestamp timestampadd timestampdiff timezone_abbr timezone_minute timezone_region to to_base64 to_date to_days to_seconds todatetimeoffset trace tracking transaction transactional translate translation treat trigger trigger_nestlevel triggers trim truncate try_cast try_convert try_parse type ub1 ub2 ub4 ucase unarchived unbounded uncompress under undo unhex unicode uniform uninstall union unique unix_timestamp unknown unlimited unlock unpivot unrecoverable unsafe unsigned until untrusted unusable unused update updated upgrade upped upper upsert url urowid usable usage use use_stored_outlines user user_data user_resources users using utc_date utc_timestamp uuid uuid_short validate validate_password_strength validation valist value values var var_samp varcharc vari varia variab variabl variable variables variance varp varraw varrawc varray verify version versions view virtual visible void wait wallet warning warnings week weekday weekofyear wellformed when whene whenev wheneve whenever where while whitespace with within without work wrapped xdb xml xmlagg xmlattributes xmlcast xmlcolattval xmlelement xmlexists xmlforest xmlindex xmlnamespaces xmlpi xmlquery xmlroot xmlschema xmlserialize xmltable xmltype xor year year_to_month years yearweek",literal:"true false null",built_in:"array bigint binary bit blob boolean char character date dec decimal float int int8 integer interval number numeric real record serial serial8 smallint text varchar varying void"},contains:[{className:"string",begin:"'",end:"'",contains:[e.BACKSLASH_ESCAPE,{begin:"''"}]},{className:"string",begin:'"',end:'"',contains:[e.BACKSLASH_ESCAPE,{begin:'""'}]},{className:"string",begin:"`",end:"`",contains:[e.BACKSLASH_ESCAPE]},e.C_NUMBER_MODE,e.C_BLOCK_COMMENT_MODE,t]},e.C_BLOCK_COMMENT_MODE,t]}}),e})},function(e,t,n){"use strict";function r(e){w.forEach(function(t){t===e?(0,v["default"])(t).parent().addClass("selected"):(0,v["default"])(t).parent().removeClass("selected")})}function i(){var e=(0,v["default"])("#full-list"),t=(0,v["default"])("#full-list .clicked");t.length>0&&e.scrollTop(t.offset().top-e.offset().top-40)}function o(e,t){var n=b.getModuleType();t=t||n;var i=e[t]||[],o=(0,v["default"])("#full-list");o.replaceWith((0,x["default"])({nodes:i,group:""})),r(["#",t,"-list"].join("")),(0,v["default"])("#full-list li a").on("click",function(e){var t=(0,v["default"])(e.target);t.hasClass("expand")?(e.preventDefault(),(0,v["default"])(e.target).closest("li").toggleClass("open")):((0,v["default"])("#full-list .clicked li.active").removeClass("active"),(0,v["default"])(e.target).closest("li").addClass("active"))})}function a(e){return function(t){t.preventDefault(),o(sidebarNodes,e),i()}}function u(){E.on("click","#extras-list",a("extras")),E.on("click","#modules-list",a("modules")),E.on("click","#exceptions-list",a("exceptions")),E.on("click","#tasks-list",a("tasks")),(0,v["default"])(".sidebar-search input").on("keydown",function(e){27===e.keyCode?(0,v["default"])(this).val(""):(event.metaKey||event.ctrlKey)&&13===e.keyCode&&((0,v["default"])(this).parent().attr("target","_blank").submit().removeAttr(""),e.preventDefault())});var e=window.location.pathname;"search.html"===e.substr(e.lastIndexOf("/")+1)&&(0,m.search)(s("q"))}function s(e){var t=window.location.href,n=e.replace(/[\[\]]/g,"\\$&"),r=new RegExp("[?&]"+n+"(=([^&#]*)|&|#|$)"),i=r.exec(t);return i&&i[2]?decodeURIComponent(i[2].replace(/\+/g," ")):""}function l(){var e=b.getLocationHash()||"content",t=sidebarNodes[b.getModuleType()],n=b.findSidebarCategory(t,e);(0,v["default"])('#full-list .clicked a.expand[href$="#'+n+'"]').closest("li").addClass("open"),(0,v["default"])('#full-list .clicked a[href$="#'+e+'"]').closest("li").addClass("active")}function c(){k.find("a").has("code").addClass("no-underline"),k.find("a").has("img").addClass("no-underline")}function f(){k.attr("tabindex",-1).focus()}function p(){o(sidebarNodes),u(),i(),l(),c(),f()}var d=n(1)["default"],h=n(5)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.initialize=p;var g=n(2),v=d(g),m=n(6),y=n(7),b=h(y),_=n(81),x=d(_),w=["#extras-list","#modules-list","#exceptions-list","#tasks-list","#search-list"],E=(0,v["default"])(".sidebar-listNav"),k=(0,v["default"])(".content")},function(e,t){"use strict";t["default"]=function(e){if(e&&e.__esModule)return e;var t={};if(null!=e)for(var n in e)Object.prototype.hasOwnProperty.call(e,n)&&(t[n]=e[n]);return t["default"]=e,t},t.__esModule=!0},function(e,t,n){"use strict";function r(e){var t=e.index,n=e.index+e[0].length,r=e.input,i=""+e[0]+"";return r.slice(0,t)+i+r.slice(n)}function i(e){return!!e}function o(e,t,n){return(e||[]).map(function(e){var i=(t+"."+e.id).match(n),o=e.id&&e.id.match(n);if(i||o){var a=JSON.parse(JSON.stringify(e));return a.match=o?r(o):e.id,a}}).filter(i)}function a(e,t,n){t.length>0&&e.push({name:n,results:t})}function u(e,t){return e.map(function(e){var n=e.title,i=n&&n.match(t),a=o(e.functions,n,t),u=o(e.macros,n,t),s=o(e.callbacks,n,t),l=o(e.types,n,t),c={id:e.id,match:i?r(i):e.title};if(a.length>0&&(c.functions=a),u.length>0&&(c.macros=u),s.length>0&&(c.callbacks=s),l.length>0&&(c.types=l),i||a.length>0||u.length>0||s.length>0||l.length>0)return c}).filter(i)}function s(e){var t=sidebarNodes;if(""!==e.replace(/\s/,"")){var n=new RegExp(h.escapeText(e),"i"),r=[],i=u(t.modules,n),o=u(t.exceptions,n),s=u(t.tasks,n);a(r,i,"Modules"),a(r,o,"Exceptions"),a(r,s,"Mix Tasks");var l=(0,v["default"])({value:e,levels:r,empty:0===r.length});y.val(e),m.html(l)}}var l=n(1)["default"],c=n(5)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.findIn=u,t.search=s;var f=n(2),p=l(f),d=n(7),h=c(d),g=n(61),v=l(g),m=(0,p["default"])("#search"),y=(0,p["default"])(".sidebar-search input")},function(e,t,n){"use strict";function r(e){return e.replace(/[\-\[\]{}()*+?.,\\\^$|#\s]/g,"\\$&")}function i(){return(0,c["default"])("body").data("type")}function o(e,t){var n=!0,r=!1,i=void 0;try{for(var o,a=u(e);!(n=(o=a.next()).done);n=!0){var s=o.value,l=(0,h["default"])(s,function(e,n){var r=(0,p["default"])(e,function(e){var n=e.anchor;return n===t});return r});if(l)return l}}catch(c){r=!0,i=c}finally{try{!n&&a["return"]&&a["return"]()}finally{if(r)throw i}}}function a(){return window.location.hash.replace(/^#/,"")}var u=n(8)["default"],s=n(1)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.escapeText=r,t.getModuleType=i,t.findSidebarCategory=o,t.getLocationHash=a;var l=n(2),c=s(l),f=n(46),p=s(f),d=n(59),h=s(d)},function(e,t,n){e.exports={"default":n(9),__esModule:!0}},function(e,t,n){n(10),n(38),e.exports=n(41)},function(e,t,n){n(11);var r=n(14);r.NodeList=r.HTMLCollection=r.Array},function(e,t,n){"use strict";var r=n(12),i=n(13),o=n(14),a=n(15);e.exports=n(19)(Array,"Array",function(e,t){this._t=a(e),this._i=0,this._k=t},function(){var e=this._t,t=this._k,n=this._i++;return!e||n>=e.length?(this._t=void 0,i(1)):"keys"==t?i(0,n):"values"==t?i(0,e[n]):i(0,[n,e[n]])},"values"),o.Arguments=o.Array,r("keys"),r("values"),r("entries")},function(e,t){e.exports=function(){}},function(e,t){e.exports=function(e,t){return{value:t,done:!!e}}},function(e,t){e.exports={}},function(e,t,n){var r=n(16),i=n(18);e.exports=function(e){return r(i(e))}},function(e,t,n){var r=n(17);e.exports=Object("z").propertyIsEnumerable(0)?Object:function(e){return"String"==r(e)?e.split(""):Object(e)}},function(e,t){ -var n={}.toString;e.exports=function(e){return n.call(e).slice(8,-1)}},function(e,t){e.exports=function(e){if(void 0==e)throw TypeError("Can't call method on "+e);return e}},function(e,t,n){"use strict";var r=n(20),i=n(21),o=n(26),a=n(27),u=n(32),s=n(14),l=n(33),c=n(34),f=n(28).getProto,p=n(35)("iterator"),d=!([].keys&&"next"in[].keys()),h="@@iterator",g="keys",v="values",m=function(){return this};e.exports=function(e,t,n,y,b,_,x){l(n,t,y);var w,E,k=function(e){if(!d&&e in A)return A[e];switch(e){case g:return function(){return new n(this,e)};case v:return function(){return new n(this,e)}}return function(){return new n(this,e)}},C=t+" Iterator",N=b==v,T=!1,A=e.prototype,S=A[p]||A[h]||b&&A[b],j=S||k(b);if(S){var O=f(j.call(new e));c(O,C,!0),!r&&u(A,h)&&a(O,p,m),N&&S.name!==v&&(T=!0,j=function(){return S.call(this)})}if(r&&!x||!d&&!T&&A[p]||a(A,p,j),s[t]=j,s[C]=m,b)if(w={values:N?j:k(v),keys:_?j:k(g),entries:N?k("entries"):j},x)for(E in w)E in A||o(A,E,w[E]);else i(i.P+i.F*(d||T),t,w);return w}},function(e,t){e.exports=!0},function(e,t,n){var r=n(22),i=n(23),o=n(24),a="prototype",u=function(e,t,n){var s,l,c,f=e&u.F,p=e&u.G,d=e&u.S,h=e&u.P,g=e&u.B,v=e&u.W,m=p?i:i[t]||(i[t]={}),y=p?r:d?r[t]:(r[t]||{})[a];p&&(n=t);for(s in n)l=!f&&y&&s in y,l&&s in m||(c=l?y[s]:n[s],m[s]=p&&"function"!=typeof y[s]?n[s]:g&&l?o(c,r):v&&y[s]==c?function(e){var t=function(t){return this instanceof e?new e(t):e(t)};return t[a]=e[a],t}(c):h&&"function"==typeof c?o(Function.call,c):c,h&&((m[a]||(m[a]={}))[s]=c))};u.F=1,u.G=2,u.S=4,u.P=8,u.B=16,u.W=32,e.exports=u},function(e,t){var n=e.exports="undefined"!=typeof window&&window.Math==Math?window:"undefined"!=typeof self&&self.Math==Math?self:Function("return this")();"number"==typeof __g&&(__g=n)},function(e,t){var n=e.exports={version:"1.2.6"};"number"==typeof __e&&(__e=n)},function(e,t,n){var r=n(25);e.exports=function(e,t,n){if(r(e),void 0===t)return e;switch(n){case 1:return function(n){return e.call(t,n)};case 2:return function(n,r){return e.call(t,n,r)};case 3:return function(n,r,i){return e.call(t,n,r,i)}}return function(){return e.apply(t,arguments)}}},function(e,t){e.exports=function(e){if("function"!=typeof e)throw TypeError(e+" is not a function!");return e}},function(e,t,n){e.exports=n(27)},function(e,t,n){var r=n(28),i=n(29);e.exports=n(30)?function(e,t,n){return r.setDesc(e,t,i(1,n))}:function(e,t,n){return e[t]=n,e}},function(e,t){var n=Object;e.exports={create:n.create,getProto:n.getPrototypeOf,isEnum:{}.propertyIsEnumerable,getDesc:n.getOwnPropertyDescriptor,setDesc:n.defineProperty,setDescs:n.defineProperties,getKeys:n.keys,getNames:n.getOwnPropertyNames,getSymbols:n.getOwnPropertySymbols,each:[].forEach}},function(e,t){e.exports=function(e,t){return{enumerable:!(1&e),configurable:!(2&e),writable:!(4&e),value:t}}},function(e,t,n){e.exports=!n(31)(function(){return 7!=Object.defineProperty({},"a",{get:function(){return 7}}).a})},function(e,t){e.exports=function(e){try{return!!e()}catch(t){return!0}}},function(e,t){var n={}.hasOwnProperty;e.exports=function(e,t){return n.call(e,t)}},function(e,t,n){"use strict";var r=n(28),i=n(29),o=n(34),a={};n(27)(a,n(35)("iterator"),function(){return this}),e.exports=function(e,t,n){e.prototype=r.create(a,{next:i(1,n)}),o(e,t+" Iterator")}},function(e,t,n){var r=n(28).setDesc,i=n(32),o=n(35)("toStringTag");e.exports=function(e,t,n){e&&!i(e=n?e:e.prototype,o)&&r(e,o,{configurable:!0,value:t})}},function(e,t,n){var r=n(36)("wks"),i=n(37),o=n(22).Symbol;e.exports=function(e){return r[e]||(r[e]=o&&o[e]||(o||i)("Symbol."+e))}},function(e,t,n){var r=n(22),i="__core-js_shared__",o=r[i]||(r[i]={});e.exports=function(e){return o[e]||(o[e]={})}},function(e,t){var n=0,r=Math.random();e.exports=function(e){return"Symbol(".concat(void 0===e?"":e,")_",(++n+r).toString(36))}},function(e,t,n){"use strict";var r=n(39)(!0);n(19)(String,"String",function(e){this._t=String(e),this._i=0},function(){var e,t=this._t,n=this._i;return n>=t.length?{value:void 0,done:!0}:(e=r(t,n),this._i+=e.length,{value:e,done:!1})})},function(e,t,n){var r=n(40),i=n(18);e.exports=function(e){return function(t,n){var o,a,u=String(i(t)),s=r(n),l=u.length;return s<0||s>=l?e?"":void 0:(o=u.charCodeAt(s),o<55296||o>56319||s+1===l||(a=u.charCodeAt(s+1))<56320||a>57343?e?u.charAt(s):o:e?u.slice(s,s+2):(o-55296<<10)+(a-56320)+65536)}}},function(e,t){var n=Math.ceil,r=Math.floor;e.exports=function(e){return isNaN(e=+e)?0:(e>0?r:n)(e)}},function(e,t,n){var r=n(42),i=n(44);e.exports=n(23).getIterator=function(e){var t=i(e);if("function"!=typeof t)throw TypeError(e+" is not iterable!");return r(t.call(e))}},function(e,t,n){var r=n(43);e.exports=function(e){if(!r(e))throw TypeError(e+" is not an object!");return e}},function(e,t){e.exports=function(e){return"object"==typeof e?null!==e:"function"==typeof e}},function(e,t,n){var r=n(45),i=n(35)("iterator"),o=n(14);e.exports=n(23).getIteratorMethod=function(e){if(void 0!=e)return e[i]||e["@@iterator"]||o[r(e)]}},function(e,t,n){var r=n(17),i=n(35)("toStringTag"),o="Arguments"==r(function(){return arguments}());e.exports=function(e){var t,n,a;return void 0===e?"Undefined":null===e?"Null":"string"==typeof(n=(t=Object(e))[i])?n:o?r(t):"Object"==(a=r(t))&&"function"==typeof t.callee?"Arguments":a}},function(e,t,n){function r(e,t){return function(n,r,o){if(r=i(r,o,3),s(n)){var l=u(n,r,t);return l>-1?n[l]:void 0}return a(n,r,e)}}var i=n(47),o=n(56),a=n(57),u=n(58),s=n(49),l=r(o);e.exports=l},function(e,t,n){function r(e){return null==e?"":e+""}function i(e,t,n){var r=typeof e;return"function"==r?void 0===t?e:w(e,t,n):null==e?b:"object"==r?u(e):void 0===t?_(e):s(e,t)}function o(e,t,n){if(null!=e){void 0!==n&&n in g(e)&&(t=[n]);for(var r=0,i=t.length;null!=e&&ri?0:i+t),n=void 0===n||n>i?i:+n||0,n<0&&(n+=i),i=t>n?0:n-t>>>0,t>>>=0;for(var o=Array(i);++rl))return!1;for(;++s-1&&e%1==0&&e<=m}function o(e){return a(e)&&h.call(e)==l}function a(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}function u(e){return null!=e&&(o(e)?g.test(p.call(e)):n(e)&&c.test(e))}var s="[object Array]",l="[object Function]",c=/^\[object .+?Constructor\]$/,f=Object.prototype,p=Function.prototype.toString,d=f.hasOwnProperty,h=f.toString,g=RegExp("^"+p.call(d).replace(/[\\^$.*+?()[\]{}|]/g,"\\$&").replace(/hasOwnProperty|(function).*?(?=\\\()| for .+?(?=\\\])/g,"$1.*?")+"$"),v=r(Array,"isArray"),m=9007199254740991,y=v||function(e){return n(e)&&i(e.length)&&h.call(e)==s};e.exports=y},function(e,t){function n(e){return"number"==typeof e&&e>-1&&e%1==0&&e<=o}function r(e){return!!e&&"object"==typeof e}function i(e){return r(e)&&n(e.length)&&!!j[M.call(e)]}var o=9007199254740991,a="[object Arguments]",u="[object Array]",s="[object Boolean]",l="[object Date]",c="[object Error]",f="[object Function]",p="[object Map]",d="[object Number]",h="[object Object]",g="[object RegExp]",v="[object Set]",m="[object String]",y="[object WeakMap]",b="[object ArrayBuffer]",_="[object DataView]",x="[object Float32Array]",w="[object Float64Array]",E="[object Int8Array]",k="[object Int16Array]",C="[object Int32Array]",N="[object Uint8Array]",T="[object Uint8ClampedArray]",A="[object Uint16Array]",S="[object Uint32Array]",j={};j[x]=j[w]=j[E]=j[k]=j[C]=j[N]=j[T]=j[A]=j[S]=!0,j[a]=j[u]=j[b]=j[s]=j[_]=j[l]=j[c]=j[f]=j[p]=j[d]=j[h]=j[g]=j[v]=j[m]=j[y]=!1;var O=Object.prototype,M=O.toString;e.exports=i},function(e,t,n){function r(e){return function(t){return null==t?void 0:t[e]}}function i(e){return null!=e&&a(y(e))}function o(e,t){return e="number"==typeof e||d.test(e)?+e:-1,t=null==t?m:t,e>-1&&e%1==0&&e-1&&e%1==0&&e<=m}function u(e){for(var t=l(e),n=t.length,r=n&&e.length,i=!!r&&a(r)&&(p(e)||f(e)),u=-1,s=[];++u0;++r-1&&e%1==0&&e<=l}function u(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}function s(e){return!!e&&"object"==typeof e}var l=9007199254740991,c="[object Arguments]",f="[object Function]",p="[object GeneratorFunction]",d=Object.prototype,h=d.hasOwnProperty,g=d.toString,v=d.propertyIsEnumerable;e.exports=n},function(e,t){function n(e,t,n){if("function"!=typeof e)return r;if(void 0===t)return e;switch(n){case 1:return function(n){return e.call(t,n)};case 3:return function(n,r,i){return e.call(t,n,r,i)};case 4:return function(n,r,i,o){return e.call(t,n,r,i,o)};case 5:return function(n,r,i,o,a){return e.call(t,n,r,i,o,a)}}return function(){return e.apply(t,arguments)}}function r(e){return e}e.exports=n},function(e,t,n){function r(e){return i(e)?e:Object(e)}function i(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}function o(e){e=r(e);for(var t=-1,n=a(e),i=n.length,o=Array(i);++t-1&&e%1==0&&e<=f}function s(e){return l(e)?e:Object(e)}function l(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}var c=n(51),f=9007199254740991,p=o(r),d=a(),h=i("length");e.exports=p},function(e,t){function n(e,t,n,r){var i;return n(e,function(e,n,o){if(t(e,n,o))return i=r?n:e,!1}),i}e.exports=n},function(e,t){function n(e,t,n){for(var r=e.length,i=n?r:-1;n?i--:++iSorry, we couldn't find anything for "+e.escapeExpression((o=null!=(o=n.value||(null!=t?t.value:t))?o:n.helperMissing,"function"==typeof o?o.call(null!=t?t:e.nullContext||{},{name:"value",hash:{},data:i}):o))+".

    \n"},3:function(e,t,n,r,i,o,a){var u;return null!=(u=n.each.call(null!=t?t:e.nullContext||{},null!=t?t.levels:t,{name:"each",hash:{},fn:e.program(4,i,0,o,a),inverse:e.noop,data:i}))?u:""},4:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{};return'

    '+e.escapeExpression((s=null!=(s=n.name||(null!=t?t.name:t))?s:n.helperMissing,"function"==typeof s?s.call(l,{name:"name",hash:{},data:i}):s))+"

    \n"+(null!=(u=n.each.call(l,null!=t?t.results:t,{name:"each",hash:{},fn:e.program(5,i,0,o,a),inverse:e.noop,data:i}))?u:"")},5:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{},c=n.helperMissing,f="function";return'
    \n

    \n '+(null!=(s=null!=(s=n.match||(null!=t?t.match:t))?s:c,u=typeof s===f?s.call(l,{name:"match",hash:{},data:i}):s)?u:"")+'\n

    \n
      \n'+(null!=(u=n.each.call(l,null!=t?t.functions:t,{name:"each",hash:{},fn:e.program(6,i,0,o,a),inverse:e.noop,data:i}))?u:"")+'
    \n
      \n'+(null!=(u=n.each.call(l,null!=t?t.macros:t,{name:"each",hash:{},fn:e.program(6,i,0,o,a),inverse:e.noop,data:i}))?u:"")+'
    \n
      \n'+(null!=(u=n.each.call(l,null!=t?t.callbacks:t,{name:"each",hash:{},fn:e.program(8,i,0,o,a),inverse:e.noop,data:i}))?u:"")+'
    \n
      \n'+(null!=(u=n.each.call(l,null!=t?t.types:t,{name:"each",hash:{},fn:e.program(10,i,0,o,a),inverse:e.noop,data:i}))?u:"")+"
    \n
    \n"},6:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{},c=n.helperMissing,f="function";return'
  • '+(null!=(s=null!=(s=n.match||(null!=t?t.match:t))?s:c,u=typeof s===f?s.call(l,{name:"match",hash:{},data:i}):s)?u:"")+"
    • \n"},8:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{},c=n.helperMissing,f="function";return'
  • '+(null!=(s=null!=(s=n.match||(null!=t?t.match:t))?s:c,u=typeof s===f?s.call(l,{name:"match",hash:{},data:i}):s)?u:"")+" (callback)
    • \n"},10:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{},c=n.helperMissing,f="function";return'
  • '+(null!=(s=null!=(s=n.match||(null!=t?t.match:t))?s:c,u=typeof s===f?s.call(l,{name:"match",hash:{},data:i}):s)?u:"")+" (type)
    • \n"},compiler:[7,">= 4.0.0"],main:function(e,t,n,r,i,o,a){var u,s,l=null!=t?t:e.nullContext||{};return"

    Search Results for "+e.escapeExpression((s=null!=(s=n.value||(null!=t?t.value:t))?s:n.helperMissing,"function"==typeof s?s.call(l,{name:"value",hash:{},data:i}):s))+"

    \n\n"+(null!=(u=n["if"].call(l,null!=t?t.empty:t,{name:"if",hash:{},fn:e.program(1,i,0,o,a),inverse:e.program(3,i,0,o,a),data:i}))?u:"")},useData:!0,useDepths:!0})},function(e,t,n){e.exports=n(63)["default"]},function(e,t,n){"use strict";function r(e){return e&&e.__esModule?e:{"default":e}}function i(e){if(e&&e.__esModule)return e;var t={};if(null!=e)for(var n in e)Object.prototype.hasOwnProperty.call(e,n)&&(t[n]=e[n]);return t["default"]=e,t}function o(){var e=new u.HandlebarsEnvironment;return d.extend(e,u),e.SafeString=l["default"],e.Exception=f["default"],e.Utils=d,e.escapeExpression=d.escapeExpression,e.VM=g,e.template=function(t){return g.template(t,e)},e}t.__esModule=!0;var a=n(64),u=i(a),s=n(78),l=r(s),c=n(66),f=r(c),p=n(65),d=i(p),h=n(79),g=i(h),v=n(80),m=r(v),y=o();y.create=o,m["default"](y),y["default"]=y,t["default"]=y,e.exports=t["default"]},function(e,t,n){"use strict";function r(e){return e&&e.__esModule?e:{"default":e}}function i(e,t,n){this.helpers=e||{},this.partials=t||{},this.decorators=n||{},s.registerDefaultHelpers(this),l.registerDefaultDecorators(this)}t.__esModule=!0,t.HandlebarsEnvironment=i;var o=n(65),a=n(66),u=r(a),s=n(67),l=n(75),c=n(77),f=r(c),p="4.0.10";t.VERSION=p;var d=7;t.COMPILER_REVISION=d;var h={1:"<= 1.0.rc.2",2:"== 1.0.0-rc.3",3:"== 1.0.0-rc.4",4:"== 1.x.x",5:"== 2.0.0-alpha.x",6:">= 2.0.0-beta.1",7:">= 4.0.0"};t.REVISION_CHANGES=h;var g="[object Object]";i.prototype={constructor:i,logger:f["default"],log:f["default"].log,registerHelper:function(e,t){if(o.toString.call(e)===g){if(t)throw new u["default"]("Arg not supported with multiple helpers");o.extend(this.helpers,e)}else this.helpers[e]=t},unregisterHelper:function(e){delete this.helpers[e]},registerPartial:function(e,t){if(o.toString.call(e)===g)o.extend(this.partials,e);else{if("undefined"==typeof t)throw new u["default"]('Attempting to register a partial called "'+e+'" as undefined');this.partials[e]=t}},unregisterPartial:function(e){delete this.partials[e]},registerDecorator:function(e,t){if(o.toString.call(e)===g){if(t)throw new u["default"]("Arg not supported with multiple decorators");o.extend(this.decorators,e)}else this.decorators[e]=t},unregisterDecorator:function(e){delete this.decorators[e]}};var v=f["default"].log;t.log=v,t.createFrame=o.createFrame,t.logger=f["default"]},function(e,t){"use strict";function n(e){return c[e]}function r(e){for(var t=1;t":">",'"':""","'":"'","`":"`","=":"="},f=/[&<>"'`=]/g,p=/[&<>"'`=]/,d=Object.prototype.toString;t.toString=d;var h=function(e){return"function"==typeof e};h(/x/)&&(t.isFunction=h=function(e){return"function"==typeof e&&"[object Function]"===d.call(e)}),t.isFunction=h;var g=Array.isArray||function(e){return!(!e||"object"!=typeof e)&&"[object Array]"===d.call(e)};t.isArray=g},function(e,t){"use strict";function n(e,t){var i=t&&t.loc,o=void 0,a=void 0;i&&(o=i.start.line,a=i.start.column,e+=" - "+o+":"+a);for(var u=Error.prototype.constructor.call(this,e),s=0;s0?(n.ids&&(n.ids=[n.name]),e.helpers.each(t,n)):i(this);if(n.data&&n.ids){var a=r.createFrame(n.data);a.contextPath=r.appendContextPath(n.data.contextPath,n.name),n={data:a}}return o(t,n)})},e.exports=t["default"]},function(e,t,n){"use strict";function r(e){return e&&e.__esModule?e:{"default":e}}t.__esModule=!0;var i=n(65),o=n(66),a=r(o);t["default"]=function(e){e.registerHelper("each",function(e,t){function n(t,n,o){l&&(l.key=t,l.index=n,l.first=0===n,l.last=!!o,c&&(l.contextPath=c+t)),s+=r(e[t],{data:l,blockParams:i.blockParams([e[t],t],[c+t,null])})}if(!t)throw new a["default"]("Must pass iterator to #each");var r=t.fn,o=t.inverse,u=0,s="",l=void 0,c=void 0;if(t.data&&t.ids&&(c=i.appendContextPath(t.data.contextPath,t.ids[0])+"."),i.isFunction(e)&&(e=e.call(this)),t.data&&(l=i.createFrame(t.data)),e&&"object"==typeof e)if(i.isArray(e))for(var f=e.length;u=0?t:parseInt(e,10)}return e},log:function(e){if(e=i.lookupLevel(e),"undefined"!=typeof console&&i.lookupLevel(i.level)<=e){var t=i.methodMap[e];console[t]||(t="log");for(var n=arguments.length,r=Array(n>1?n-1:0),o=1;o\n\n\n '+p(f(null!=(l=u[0][0])?l.title:l,t))+'\n\n
      \n
    • \n Top\n
      • \n\n'+(null!=(l=r(n(84)).call(c,null!=(l=u[0][0])?l.headers:l,{name:"isArray",hash:{},fn:e.program(6,a,0,u,s),inverse:e.program(9,a,0,u,s),data:a,blockParams:u}))?l:"")+"
    \n \n"},2:function(e,t,n,r,i,o){var a;return'
  • '+e.escapeExpression(e.lambda(null!=(a=o[1][0])?a.group:a,t))+"
    • \n"},4:function(e,t,n,r,i){return"clicked open"},6:function(e,t,n,r,i,o){var a;return null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[1][0])?a.headers:a,{name:"each",hash:{},fn:e.program(7,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:""},7:function(e,t,n,r,i,o){var a,u=e.lambda,s=e.escapeExpression;return'
  • \n '+s(u(null!=t?t.id:t,t))+"\n
    • \n"},9:function(e,t,i,o,a,u){var s,l=null!=t?t:e.nullContext||{};return(null!=(s=r(n(85)).call(l,u[1][0],{name:"showSummary",hash:{},fn:e.program(10,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")+(null!=(s=i["if"].call(l,null!=(s=u[1][0])?s.types:s,{name:"if",hash:{},fn:e.program(12,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")+(null!=(s=i["if"].call(l,null!=(s=u[1][0])?s.functions:s,{name:"if",hash:{},fn:e.program(15,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")+(null!=(s=i["if"].call(l,null!=(s=u[1][0])?s.macros:s,{name:"if",hash:{},fn:e.program(17,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")+(null!=(s=i["if"].call(l,null!=(s=u[1][0])?s.callbacks:s,{name:"if",hash:{},fn:e.program(20,a,0,u),inverse:e.noop,data:a,blockParams:u}))?s:"")},10:function(e,t,n,r,i,o){var a;return'
  • \n Summary\n
    • \n'},12:function(e,t,n,r,i,o){var a;return'
  • \n Types\n
      \n'+(null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[2][0])?a.types:a,{name:"each",hash:{},fn:e.program(13,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:"")+"
    \n
    • \n"},13:function(e,t,n,r,i,o){var a,u=e.lambda,s=e.escapeExpression;return'
  • \n '+s(u(null!=t?t.id:t,t))+"\n
    • \n"},15:function(e,t,n,r,i,o){var a;return'
  • \n Functions\n
      \n'+(null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[2][0])?a.functions:a,{name:"each",hash:{},fn:e.program(13,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:"")+"
    \n
    • \n"},17:function(e,t,n,r,i,o){var a;return'
  • \n Macros\n
      \n'+(null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[2][0])?a.macros:a,{name:"each",hash:{},fn:e.program(18,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:"")+"
    \n
    • \n"},18:function(e,t,n,r,i,o){var a,u=e.lambda,s=e.escapeExpression;return'
  • \n '+s(u(null!=t?t.id:t,t))+"\n
    • \n"},20:function(e,t,n,r,i,o){var a;return'
  • \n Callbacks\n
      \n'+(null!=(a=n.each.call(null!=t?t:e.nullContext||{},null!=(a=o[2][0])?a.callbacks:a,{name:"each",hash:{},fn:e.program(18,i,0,o),inverse:e.noop,data:i,blockParams:o}))?a:"")+"
    \n
    • \n"},compiler:[7,">= 4.0.0"],main:function(e,t,n,r,i,o,a){var u;return'
      \n'+(null!=(u=n.each.call(null!=t?t:e.nullContext||{},null!=t?t.nodes:t,{name:"each",hash:{},fn:e.program(1,i,2,o,a),inverse:e.noop,data:i,blockParams:o}))?u:"")+"
    \n"},useData:!0,useDepths:!0,useBlockParams:!0})},function(e,t){"use strict";Object.defineProperty(t,"__esModule",{value:!0}),t["default"]=function(e,t,n){var r=t||"";if(e.group!==r)return e.group=r,n.fn(this)},e.exports=t["default"]},function(e,t){"use strict";Object.defineProperty(t,"__esModule",{value:!0}),t["default"]=function(e,t){var n=window.location.pathname.split("/");return e+=".html",e===n[n.length-1]?t.fn(this):t.inverse(this)},e.exports=t["default"]},function(e,t){"use strict";Object.defineProperty(t,"__esModule",{value:!0}),t["default"]=function(e,t){return Array.isArray(e)?t.fn(this):t.inverse(this)},e.exports=t["default"]},function(e,t){"use strict";Object.defineProperty(t,"__esModule",{value:!0}),t["default"]=function(e,t){if(e.types||e.functions||e.macros||e.callbacks)return t.fn(this)},e.exports=t["default"]},function(e,t,n){"use strict";function r(){p.addClass(y).removeClass(g).removeClass(v),_=setTimeout(function(){return p.addClass(m).removeClass(y)},h)}function i(){p.addClass(v).removeClass(m).removeClass(y),_=setTimeout(function(){return p.addClass(g).removeClass(v)},h)}function o(){var e=p.attr("class")||"";clearTimeout(_),e.includes(m)||e.includes(y)?i():r()}function a(){p.removeClass(b),p.addClass(window.innerWidth>d?g:m)}function u(){a();var e=window.innerWidth;(0,c["default"])(window).resize((0,f.throttle)(function(){e!==window.innerWidth&&(e=window.innerWidth,a())},100)),(0,c["default"])(".sidebar-toggle").click(function(){o()})}var s=n(1)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.initialize=u;var l=n(2),c=s(l),f=n(87),p=(0,c["default"])("body"),d=768,h=300,g="sidebar-opened",v="sidebar-opening",m="sidebar-closed",y="sidebar-closing",b=[g,v,m,y].join(" "),_=void 0;t.breakpoint=d,t.closeSidebar=r},function(e,t,n){var r;(function(e,i){(function(){function o(e,t){if(e!==t){var n=null===e,r=e===T,i=e===e,o=null===t,a=t===T,u=t===t;if(e>t&&!o||!i||n&&!a&&u||r&&u)return 1;if(e-1;);return n}function f(e,t){for(var n=e.length;n--&&t.indexOf(e.charAt(n))>-1;);return n}function p(e,t){return o(e.criteria,t.criteria)||e.index-t.index}function d(e,t,n){for(var r=-1,i=e.criteria,a=t.criteria,u=i.length,s=n.length;++r=s)return l;var c=n[r];return l*("asc"===c||c===!0?1:-1)}}return e.index-t.index}function h(e){return Ke[e]}function g(e){return Ge[e]}function v(e,t,n){return t?e=Ze[e]:n&&(e=Qe[e]),"\\"+e}function m(e){return"\\"+Qe[e]}function y(e,t,n){for(var r=e.length,i=t+(n?0:-1);n?i--:++i=9&&e<=13||32==e||160==e||5760==e||6158==e||e>=8192&&(e<=8202||8232==e||8233==e||8239==e||8287==e||12288==e||65279==e)}function x(e,t){for(var n=-1,r=e.length,i=-1,o=[];++n=z?gn(t):null,l=t.length;s&&(o=Qe,a=!1,t=s);e:for(;++ii?0:i+n),r=r===T||r>i?i:+r||0,r<0&&(r+=i),i=n>r?0:r>>>0,n>>>=0;ni?0:i+t),n=n===T||n>i?i:+n||0,n<0&&(n+=i),i=t>n?0:n-t>>>0,t>>>=0;for(var o=$o(i);++r=z,s=a?gn():null,l=[];s?(r=Qe,o=!1):(a=!1,s=t?[]:l);e:for(;++n>>1,a=e[o];(n?a<=t:a2?n[i-2]:T,a=i>2?n[2]:T,u=i>1?n[i-1]:T;for("function"==typeof o?(o=an(o,u,5),i-=2):(o="function"==typeof u?u:T,i-=o?1:0),a&&Jn(n[0],n[1],a)&&(o=i<3?T:o,i=1);++r-1?n[o]:T}return At(n,r,e)}}function wn(e){return function(t,n,r){return t&&t.length?(n=$n(n,r,3),a(t,n,e)):-1}}function En(e){return function(t,n,r){return n=$n(n,r,3),At(t,n,e,!0)}}function kn(e){return function(){for(var t,n=arguments.length,i=e?n:-1,o=0,a=$o(n);e?i--:++i=z)return t.plant(r).value();for(var i=0,o=n?a[i].apply(this,e):r;++i=t||!_a(t))return"";var i=t-r;return n=null==n?" ":n+"",mo(n,va(i/n.length)).slice(0,i)}function Rn(e,t,n,r){function i(){for(var t=-1,u=arguments.length,s=-1,l=r.length,c=$o(l+u);++ss))return!1;for(;++u-1&&e%1==0&&e-1&&e%1==0&&e<=Ma}function rr(e){return e===e&&!Ri(e)}function ir(e,t){var n=e[1],r=t[1],i=n|r,o=i-1;)da.call(t,o,1);return t}function Sr(e,t,n){var r=[];if(!e||!e.length)return r;var i=-1,o=[],a=e.length;for(t=$n(t,n,3);++i-1:!!i&&Fn(e,t,n)>-1}function ei(e,t,n){var r=Su(e)?st:qt;return t=$n(t,n,3),r(e,t)}function ti(e,t){return ei(e,Ro(t))}function ni(e,t,n){var r=Su(e)?ut:Tt;return t=$n(t,n,3),r(e,function(e,n,r){return!t(e,n,r)})}function ri(e,t,n){if(n?Jn(e,t,n):null==t){e=cr(e);var r=e.length;return r>0?e[Kt(0,r-1)]:T}var i=-1,o=Gi(e),r=o.length,a=r-1;for(t=Ea(t<0?0:+t||0,r);++i0&&(n=t.apply(this,arguments)),e<=1&&(t=T),n}}function di(e,t,n){function r(){d&&ua(d),l&&ua(l),g=0,l=d=h=T}function i(t,n){n&&ua(n),l=d=h=T,t&&(g=gu(),c=e.apply(p,s),d||l||(s=p=T))}function o(){var e=t-(gu()-f);e<=0||e>t?i(h,l):d=pa(o,e)}function a(){i(m,d)}function u(){if(s=arguments,f=gu(),p=this,h=m&&(d||!y),v===!1)var n=y&&!d;else{l||y||(g=f);var r=v-(f-g),i=r<=0||r>v;i?(l&&(l=ua(l)),g=f,c=e.apply(p,s)):l||(l=pa(a,r))}return i&&d?d=ua(d):d||t===v||(d=pa(o,t)),n&&(i=!0,c=e.apply(p,s)),!i||d||l||(s=p=T),c}var s,l,c,f,p,d,h,g=0,v=!1,m=!0;if("function"!=typeof e)throw new Zo(U);if(t=t<0?0:+t||0,n===!0){var y=!0;m=!1}else Ri(n)&&(y=!!n.leading,v="maxWait"in n&&wa(+n.maxWait||0,t),m="trailing"in n?!!n.trailing:m);return u.cancel=r,u}function hi(e,t){if("function"!=typeof e||t&&"function"!=typeof t)throw new Zo(U);var n=function(){var r=arguments,i=t?t.apply(this,r):r[0],o=n.cache;if(o.has(i))return o.get(i);var a=e.apply(this,r);return n.cache=o.set(i,a),a};return n.cache=new hi.Cache,n}function gi(e){if("function"!=typeof e)throw new Zo(U);return function(){return!e.apply(this,arguments)}}function vi(e){return pi(2,e)}function mi(e,t){if("function"!=typeof e)throw new Zo(U);return t=wa(t===T?e.length-1:+t||0,0),function(){for(var n=arguments,r=-1,i=wa(n.length-t,0),o=$o(i);++rt}function ki(e,t){return e>=t}function Ci(e){return b(e)&&Qn(e)&&ta.call(e,"callee")&&!ca.call(e,"callee")}function Ni(e){return e===!0||e===!1||b(e)&&ra.call(e)==X}function Ti(e){return b(e)&&ra.call(e)==Z}function Ai(e){return!!e&&1===e.nodeType&&b(e)&&!Bi(e)}function Si(e){return null==e||(Qn(e)&&(Su(e)||zi(e)||Ci(e)||b(e)&&Di(e.splice))?!e.length:!Bu(e).length)}function ji(e,t,n,r){n="function"==typeof n?an(n,r,3):T;var i=n?n(e,t):T;return i===T?Pt(e,t,n):!!i}function Oi(e){return b(e)&&"string"==typeof e.message&&ra.call(e)==Q}function Mi(e){return"number"==typeof e&&_a(e)}function Di(e){return Ri(e)&&ra.call(e)==Y}function Ri(e){var t=typeof e;return!!e&&("object"==t||"function"==t)}function Pi(e,t,n,r){return n="function"==typeof n?an(n,r,3):T,It(e,Wn(t),n)}function Li(e){return Hi(e)&&e!=+e}function Ii(e){return null!=e&&(Di(e)?oa.test(ea.call(e)):b(e)&&Le.test(e))}function qi(e){return null===e}function Hi(e){return"number"==typeof e||b(e)&&ra.call(e)==ee}function Bi(e){var t;if(!b(e)||ra.call(e)!=te||Ci(e)||!ta.call(e,"constructor")&&(t=e.constructor,"function"==typeof t&&!(t instanceof t)))return!1;var n;return jt(e,function(e,t){n=t}),n===T||ta.call(e,n)}function $i(e){return Ri(e)&&ra.call(e)==ne}function zi(e){return"string"==typeof e||b(e)&&ra.call(e)==ie}function Fi(e){return b(e)&&nr(e.length)&&!!We[ra.call(e)]}function Wi(e){return e===T}function Ui(e,t){return e0;++r=Ea(t,n)&&e=0&&e.indexOf(t,n)==n}function po(e){return e=l(e),e&&we.test(e)?e.replace(_e,g):e}function ho(e){return e=l(e),e&&je.test(e)?e.replace(Se,v):e||"(?:)"}function go(e,t,n){e=l(e),t=+t;var r=e.length;if(r>=t||!_a(t))return e;var i=(t-r)/2,o=ya(i),a=va(i);return n=Dn("",a,n),n.slice(0,o)+e+n}function vo(e,t,n){return(n?Jn(e,t,n):null==t)?t=0:t&&(t=+t),e=_o(e),Ca(e,t||(Pe.test(e)?16:10))}function mo(e,t){var n="";if(e=l(e),t=+t,t<1||!e||!_a(t))return n;do t%2&&(n+=e),t=ya(t/2),e+=e;while(t);return n}function yo(e,t,n){return e=l(e),n=null==n?0:Ea(n<0?0:+n||0,e.length),e.lastIndexOf(t,n)==n}function bo(e,n,r){var i=t.templateSettings;r&&Jn(e,n,r)&&(n=r=T),e=l(e),n=vt(mt({},r||n),i,gt);var o,a,u=vt(mt({},n.imports),i.imports,gt),s=Bu(u),c=en(u,s),f=0,p=n.interpolate||He,d="__p += '",h=Vo((n.escape||He).source+"|"+p.source+"|"+(p===Ce?De:He).source+"|"+(n.evaluate||He).source+"|$","g"),g="//# sourceURL="+("sourceURL"in n?n.sourceURL:"lodash.templateSources["+ ++Fe+"]")+"\n";e.replace(h,function(t,n,r,i,u,s){return r||(r=i),d+=e.slice(f,s).replace(Be,m),n&&(o=!0,d+="' +\n__e("+n+") +\n'"),u&&(a=!0,d+="';\n"+u+";\n__p += '"),r&&(d+="' +\n((__t = ("+r+")) == null ? '' : __t) +\n'"),f=s+t.length,t}),d+="';\n";var v=n.variable;v||(d="with (obj) {\n"+d+"\n}\n"),d=(a?d.replace(ve,""):d).replace(me,"$1").replace(ye,"$1;"),d="function("+(v||"obj")+") {\n"+(v?"":"obj || (obj = {});\n")+"var __t, __p = ''"+(o?", __e = _.escape":"")+(a?", __j = Array.prototype.join;\nfunction print() { __p += __j.call(arguments, '') }\n":";\n")+d+"return __p\n}";var y=Qu(function(){return Wo(s,g+"return "+d).apply(T,c)});if(y.source=d,Oi(y))throw y;return y}function _o(e,t,n){var r=e;return(e=l(e))?(n?Jn(r,t,n):null==t)?e.slice(E(e),k(e)+1):(t+="",e.slice(c(e,t),f(e,t)+1)):e}function xo(e,t,n){var r=e;return e=l(e),e?(n?Jn(r,t,n):null==t)?e.slice(E(e)):e.slice(c(e,t+"")):e}function wo(e,t,n){var r=e;return e=l(e),e?(n?Jn(r,t,n):null==t)?e.slice(0,k(e)+1):e.slice(0,f(e,t+"")+1):e}function Eo(e,t,n){n&&Jn(e,t,n)&&(t=T);var r=q,i=H;if(null!=t)if(Ri(t)){var o="separator"in t?t.separator:o;r="length"in t?+t.length||0:r,i="omission"in t?l(t.omission):i}else r=+t||0;if(e=l(e),r>=e.length)return e;var a=r-i.length;if(a<1)return i;var u=e.slice(0,a);if(null==o)return u+i;if($i(o)){if(e.slice(a).search(o)){var s,c,f=e.slice(0,a);for(o.global||(o=Vo(o.source,(Re.exec(o)||"")+"g")),o.lastIndex=0;s=o.exec(f);)c=s.index;u=u.slice(0,null==c?a:c)}}else if(e.indexOf(o,a)!=a){var p=u.lastIndexOf(o);p>-1&&(u=u.slice(0,p))}return u+i}function ko(e){return e=l(e),e&&xe.test(e)?e.replace(be,C):e}function Co(e,t,n){return n&&Jn(e,t,n)&&(t=T),e=l(e),e.match(t||$e)||[]}function No(e,t,n){return n&&Jn(e,t,n)&&(t=T),b(e)?So(e):_t(e,t)}function To(e){return function(){return e}}function Ao(e){return e}function So(e){return Ht(xt(e,!0))}function jo(e,t){return Bt(e,xt(t,!0))}function Oo(e,t,n){if(null==n){var r=Ri(t),i=r?Bu(t):T,o=i&&i.length?Dt(t,i):T;(o?o.length:r)||(o=!1,n=t,t=e,e=this)}o||(o=Dt(t,Bu(t)));var a=!0,u=-1,s=Di(e),l=o.length;n===!1?a=!1:Ri(n)&&"chain"in n&&(a=n.chain);for(;++u>>1,Ma=9007199254740991,Da=ga&&new ga,Ra={};t.support={};t.templateSettings={escape:Ee,evaluate:ke,interpolate:Ce,variable:"",imports:{_:t}};var Pa=function(){function e(){}return function(t){if(Ri(t)){e.prototype=t;var n=new e;e.prototype=T}return n||{}}}(),La=pn(Ot),Ia=pn(Mt,!0),qa=dn(),Ha=dn(!0),Ba=Da?function(e,t){return Da.set(e,t),e}:Ao,$a=Da?function(e){return Da.get(e)}:Do,za=Ft("length"),Fa=function(){var e=0,t=0;return function(n,r){var i=gu(),o=$-(i-t);if(t=i,o>0){if(++e>=B)return n}else e=0;return Ba(n,r)}}(),Wa=mi(function(e,t){return b(e)&&Qn(e)?Et(e,St(t,!1,!0)):[]}),Ua=wn(),Ka=wn(!0),Ga=mi(function(e){for(var t=e.length,n=t,r=$o(f),i=Fn(),o=i==u,a=[];n--;){var s=e[n]=Qn(s=e[n])?s:[];r[n]=o&&s.length>=120?gn(n&&s):null}var l=e[0],c=-1,f=l?l.length:0,p=r[0];e:for(;++c2?e[t-2]:T,r=t>1?e[t-1]:T;return t>2&&"function"==typeof n?t-=2:(n=t>1&&"function"==typeof r?(--t,r):T,r=T),e.length=t,qr(e,n,r)}),tu=mi(function(e){return e=St(e),this.thru(function(t){return Je(Su(t)?t:[fr(t)],e)})}),nu=mi(function(e,t){return yt(e,St(t))}),ru=cn(function(e,t,n){ta.call(e,n)?++e[n]:e[n]=1}),iu=xn(La),ou=xn(Ia,!0),au=Cn(tt,La),uu=Cn(nt,Ia),su=cn(function(e,t,n){ta.call(e,n)?e[n].push(t):e[n]=[t]}),lu=cn(function(e,t,n){e[n]=t}),cu=mi(function(e,t,n){var r=-1,i="function"==typeof t,o=er(t),a=Qn(e)?$o(e.length):[];return La(e,function(e){var u=i?t:o&&null!=e?e[t]:T;a[++r]=u?u.apply(e,n):Zn(e,t,n)}),a}),fu=cn(function(e,t,n){e[n?0:1].push(t)},function(){return[[],[]]}),pu=On(ct,La),du=On(ft,Ia),hu=mi(function(e,t){if(null==e)return[];var n=t[2];return n&&Jn(t[0],t[1],n)&&(t.length=1),Qt(e,St(t),[])}),gu=ka||function(){return(new zo).getTime()},vu=mi(function(e,t,n){var r=S;if(n.length){var i=x(n,vu.placeholder);r|=R}return In(e,r,t,n,i)}),mu=mi(function(e,t){t=t.length?St(t):Zi(e);for(var n=-1,r=t.length;++n0||t<0)?new i(n):(e<0?n=n.takeRight(-e):e&&(n=n.drop(e)),t!==T&&(t=+t||0,n=t<0?n.dropRight(-t):n.take(t-e)),n)},i.prototype.takeRightWhile=function(e,t){return this.reverse().takeWhile(e,t).reverse()},i.prototype.toArray=function(){return this.take(Aa)},Ot(i.prototype,function(e,n){var o=/^(?:filter|map|reject)|While$/.test(n),a=/^(?:first|last)$/.test(n),u=t[a?"take"+("last"==n?"Right":""):n];u&&(t.prototype[n]=function(){var t=a?[1]:arguments,n=this.__chain__,s=this.__wrapped__,l=!!this.__actions__.length,c=s instanceof i,f=t[0],p=c||Su(s);p&&o&&"function"==typeof f&&1!=f.length&&(c=p=!1);var d=function(e){return a&&n?u(e,1)[0]:u.apply(T,lt([e],t))},h={func:Fr,args:[d],thisArg:T},g=c&&!l;if(a&&!n)return g?(s=s.clone(),s.__actions__.push(h),e.call(s)):u.call(T,this.value())[0];if(!a&&p){s=g?s:new i(this);var v=e.apply(s,t);return v.__actions__.push(h),new r(v,n)}return this.thru(d)})}),tt(["join","pop","push","replace","shift","sort","splice","split","unshift"],function(e){var n=(/^(?:replace|split)$/.test(e)?Jo:Qo)[e],r=/^(?:push|sort|unshift)$/.test(e)?"tap":"thru",i=/^(?:join|pop|replace|shift)$/.test(e);t.prototype[e]=function(){var e=arguments;return i&&!this.__chain__?n.apply(this.value(),e):this[r](function(t){return n.apply(t,e)})}}),Ot(i.prototype,function(e,n){var r=t[n];if(r){var i=r.name,o=Ra[i]||(Ra[i]=[]);o.push({name:n,func:r})}}),Ra[Mn(T,j).name]=[{name:"wrapper",func:T}],i.prototype.clone=_,i.prototype.reverse=J,i.prototype.value=re,t.prototype.chain=Wr,t.prototype.commit=Ur,t.prototype.concat=tu,t.prototype.plant=Kr,t.prototype.reverse=Gr,t.prototype.toString=Vr,t.prototype.run=t.prototype.toJSON=t.prototype.valueOf=t.prototype.value=Xr,t.prototype.collect=t.prototype.map,t.prototype.head=t.prototype.first,t.prototype.select=t.prototype.filter,t.prototype.tail=t.prototype.rest,t}var T,A="3.10.1",S=1,j=2,O=4,M=8,D=16,R=32,P=64,L=128,I=256,q=30,H="...",B=150,$=16,z=200,F=1,W=2,U="Expected a function",K="__lodash_placeholder__",G="[object Arguments]",V="[object Array]",X="[object Boolean]",Z="[object Date]",Q="[object Error]",Y="[object Function]",J="[object Map]",ee="[object Number]",te="[object Object]",ne="[object RegExp]",re="[object Set]",ie="[object String]",oe="[object WeakMap]",ae="[object ArrayBuffer]",ue="[object Float32Array]",se="[object Float64Array]",le="[object Int8Array]",ce="[object Int16Array]",fe="[object Int32Array]",pe="[object Uint8Array]",de="[object Uint8ClampedArray]",he="[object Uint16Array]",ge="[object Uint32Array]",ve=/\b__p \+= '';/g,me=/\b(__p \+=) '' \+/g,ye=/(__e\(.*?\)|\b__t\)) \+\n'';/g,be=/&(?:amp|lt|gt|quot|#39|#96);/g,_e=/[&<>"'`]/g,xe=RegExp(be.source),we=RegExp(_e.source),Ee=/<%-([\s\S]+?)%>/g,ke=/<%([\s\S]+?)%>/g,Ce=/<%=([\s\S]+?)%>/g,Ne=/\.|\[(?:[^[\]]*|(["'])(?:(?!\1)[^\n\\]|\\.)*?\1)\]/,Te=/^\w*$/,Ae=/[^.[\]]+|\[(?:(-?\d+(?:\.\d+)?)|(["'])((?:(?!\2)[^\n\\]|\\.)*?)\2)\]/g,Se=/^[:!,]|[\\^$.*+?()[\]{}|\/]|(^[0-9a-fA-Fnrtuvx])|([\n\r\u2028\u2029])/g,je=RegExp(Se.source),Oe=/[\u0300-\u036f\ufe20-\ufe23]/g,Me=/\\(\\)?/g,De=/\$\{([^\\}]*(?:\\.[^\\}]*)*)\}/g,Re=/\w*$/,Pe=/^0[xX]/,Le=/^\[object .+?Constructor\]$/,Ie=/^\d+$/,qe=/[\xc0-\xd6\xd8-\xde\xdf-\xf6\xf8-\xff]/g,He=/($^)/,Be=/['\n\r\u2028\u2029\\]/g,$e=function(){var e="[A-Z\\xc0-\\xd6\\xd8-\\xde]",t="[a-z\\xdf-\\xf6\\xf8-\\xff]+";return RegExp(e+"+(?="+e+t+")|"+e+"?"+t+"|"+e+"+|[0-9]+","g")}(),ze=["Array","ArrayBuffer","Date","Error","Float32Array","Float64Array","Function","Int8Array","Int16Array","Int32Array","Math","Number","Object","RegExp","Set","String","_","clearTimeout","isFinite","parseFloat","parseInt","setTimeout","TypeError","Uint8Array","Uint8ClampedArray","Uint16Array","Uint32Array","WeakMap"],Fe=-1,We={};We[ue]=We[se]=We[le]=We[ce]=We[fe]=We[pe]=We[de]=We[he]=We[ge]=!0,We[G]=We[V]=We[ae]=We[X]=We[Z]=We[Q]=We[Y]=We[J]=We[ee]=We[te]=We[ne]=We[re]=We[ie]=We[oe]=!1;var Ue={};Ue[G]=Ue[V]=Ue[ae]=Ue[X]=Ue[Z]=Ue[ue]=Ue[se]=Ue[le]=Ue[ce]=Ue[fe]=Ue[ee]=Ue[te]=Ue[ne]=Ue[ie]=Ue[pe]=Ue[de]=Ue[he]=Ue[ge]=!0,Ue[Q]=Ue[Y]=Ue[J]=Ue[re]=Ue[oe]=!1;var Ke={"À":"A","Á":"A","Â":"A","Ã":"A","Ä":"A","Å":"A","à":"a","á":"a","â":"a","ã":"a","ä":"a","å":"a","Ç":"C","ç":"c","Ð":"D","ð":"d","È":"E","É":"E","Ê":"E","Ë":"E","è":"e","é":"e","ê":"e","ë":"e","Ì":"I","Í":"I","Î":"I","Ï":"I","ì":"i","í":"i","î":"i","ï":"i","Ñ":"N","ñ":"n","Ò":"O","Ó":"O","Ô":"O","Õ":"O","Ö":"O","Ø":"O","ò":"o","ó":"o","ô":"o","õ":"o","ö":"o","ø":"o","Ù":"U","Ú":"U","Û":"U","Ü":"U","ù":"u","ú":"u","û":"u","ü":"u","Ý":"Y","ý":"y","ÿ":"y","Æ":"Ae","æ":"ae","Þ":"Th","þ":"th","ß":"ss"},Ge={"&":"&","<":"<",">":">",'"':""","'":"'","`":"`"},Ve={"&":"&","<":"<",">":">",""":'"',"'":"'","`":"`"},Xe={"function":!0,object:!0},Ze={0:"x30",1:"x31",2:"x32",3:"x33",4:"x34",5:"x35",6:"x36",7:"x37",8:"x38",9:"x39",A:"x41",B:"x42",C:"x43",D:"x44",E:"x45",F:"x46",a:"x61",b:"x62",c:"x63",d:"x64",e:"x65",f:"x66",n:"x6e",r:"x72",t:"x74",u:"x75",v:"x76",x:"x78"},Qe={"\\":"\\","'":"'","\n":"n","\r":"r","\u2028":"u2028","\u2029":"u2029"},Ye=Xe[typeof t]&&t&&!t.nodeType&&t,Je=Xe[typeof e]&&e&&!e.nodeType&&e,et=Ye&&Je&&"object"==typeof i&&i&&i.Object&&i,tt=Xe[typeof self]&&self&&self.Object&&self,nt=Xe[typeof window]&&window&&window.Object&&window,rt=(Je&&Je.exports===Ye&&Ye,et||nt!==(this&&this.window)&&nt||tt||this),it=N();rt._=it,r=function(){return it}.call(t,n,t,e),!(r!==T&&(e.exports=r))}).call(this)}).call(t,n(88)(e),function(){return this}())},function(e,t){e.exports=function(e){return e.webpackPolyfill||(e.deprecate=function(){},e.paths=[],e.children=[],e.webpackPolyfill=1),e}},function(e,t,n){"use strict";function r(){f.addClass(p);try{localStorage.setItem(p,!0)}catch(e){}}function i(){f.removeClass(p);try{localStorage.removeItem(p)}catch(e){}}function o(){try{localStorage.getItem(p)&&r()}catch(e){}}function a(){f.hasClass(p)?i():r()}function u(){o(),d.click(function(){a()})}var s=n(1)["default"];Object.defineProperty(t,"__esModule",{value:!0}),t.initialize=u;var l=n(2),c=s(l),f=(0,c["default"])("body"),p="night-mode",d=(0,c["default"])(".night-mode-toggle")}]); -//# sourceMappingURL=app-9bd040e5e5.js.map diff --git a/doc/dist/sidebar_items-3a30d3745e.js b/doc/dist/sidebar_items-3a30d3745e.js deleted file mode 100644 index 2957579..0000000 --- a/doc/dist/sidebar_items-3a30d3745e.js +++ /dev/null @@ -1 +0,0 @@ -sidebarNodes={"extras":[{"id":"api-reference","title":"API Reference","group":"","headers":[{"id":"Modules","anchor":"modules"}]}],"exceptions":[],"modules":[{"id":"Tensorflex","title":"Tensorflex","group":"","functions":[{"id":"append_to_matrix/2","anchor":"append_to_matrix/2"},{"id":"create_matrix/3","anchor":"create_matrix/3"},{"id":"float32_tensor/1","anchor":"float32_tensor/1"},{"id":"float32_tensor/2","anchor":"float32_tensor/2"},{"id":"float32_tensor_alloc/1","anchor":"float32_tensor_alloc/1"},{"id":"float64_tensor/1","anchor":"float64_tensor/1"},{"id":"float64_tensor/2","anchor":"float64_tensor/2"},{"id":"float64_tensor_alloc/1","anchor":"float64_tensor_alloc/1"},{"id":"get_graph_ops/1","anchor":"get_graph_ops/1"},{"id":"int32_tensor/1","anchor":"int32_tensor/1"},{"id":"int32_tensor/2","anchor":"int32_tensor/2"},{"id":"int32_tensor_alloc/1","anchor":"int32_tensor_alloc/1"},{"id":"load_csv_as_matrix/2","anchor":"load_csv_as_matrix/2"},{"id":"load_image_as_tensor/1","anchor":"load_image_as_tensor/1"},{"id":"matrix_pos/3","anchor":"matrix_pos/3"},{"id":"matrix_to_lists/1","anchor":"matrix_to_lists/1"},{"id":"read_graph/1","anchor":"read_graph/1"},{"id":"run_session/5","anchor":"run_session/5"},{"id":"size_of_matrix/1","anchor":"size_of_matrix/1"},{"id":"string_tensor/1","anchor":"string_tensor/1"},{"id":"tensor_datatype/1","anchor":"tensor_datatype/1"}]}],"tasks":[]} \ No newline at end of file diff --git a/doc/fonts/icomoon.eot b/doc/fonts/icomoon.eot deleted file mode 100644 index 2786a1e5b0e9ee7a593e544f2bc840cf821af61a..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 3096 zcma)8O>7&-6@Ig`JIh@zx%~f=#Qz`HvSgX0$ka`2ra&DylHtIK5~K)FF|s1LF-2LH zfCwlGqp%&IhsHf5Xwpjqni=_g{O(1V51EhdrqCmzJ5ty{VeJ^N)9f%zb zyZ+nz*YCq<9P;&*?$veB5aeHg`c^J)zi{KHe?AZSZ6f`-)#dKRzDy;%fU^X}jVEJ8bxAkH2!~7@go_4P=_|IVAcsRgmA7_J) zbA#L;v!7-Ep7hyT@Kg90jqP;w8b%DB>=8XSQ~R5a`9Ytr#}tYWcG*AK9r`xiz%c8{ zg#5*0y0Dqh@hcT;Rs32FN7bmeTc=s8-I{GWaJ5vFmXf%5Lw8+x{aiKN#AB{!*FmCD zZ@2+G_tn{^=eXT!uou(GR3vJcrWxrdJd#KxCo-{6$Owdm?hnUPp-{-`Sh^6pWq}We z?7&n!$#t%DBB@mRwg1~XnoMSfv5cq}i`_BI$WQ|;orbrqakU1q&)q=p@au9kxheDp3owTfp?n?OiHhW6NFlRiynDn?{FSlS3EA zo=5+uyyzr{4I<}kK%>&NQ9-W%6e-B@HdUxiRHapZJ7rlwP4bCa5%Ur$82=$)yb4>i_G z>3|)t;Qld*20IF|9eg2@$h$$`p=a=M*`%BFGpuqEdlH8nJF%)4G0nI>FzbTXTaU0N zZVNHNVu`3rU?xGHb%xh!9%ohT>sXFy9z9T0n$=0y_qSNRRe>Lbku2hrG<3pLz#VQV zTu(*4vd`i&b99Vm=jLW*E!FEKSxvv+9I0<|A6I=`*C%vbHhPv#3rQz{IdFFGuHT$A zEX$ZQ{p>X~p1AkJtgdsPivI*wJuFyW)j2Q`hA*8N8e5o~ZO+d(XXke6rE+bWFNUwu z486+(A)>$4Aazh^)EIr-i9+Fwro=?`-FF2zYm;UwVQv;Uk!ALso*iaaf( zq#Q4j9Tv;=Fs;D!c?7OOs-hrCs&}`w81sPVxvp#ovcUI>WBEQ&4In{i zTwVdc;_)2(BOb4U|AEJAcpu@F=JXZ7FL=C3O?vCnQoq;luRZvpeR_H0>ZSf#{>1dm z;m_DZdmRd1SYFw>+}%LXk@xYTmS_5FoB5UHwdIZO=JLh-^V|8eOZ{`mZJN$FPv}jI kVg+v2P&e?*-UPh}?0MQod)9f9pYxvN)0l7hQv90oU%@nu1ONa4 diff --git a/doc/fonts/icomoon.svg b/doc/fonts/icomoon.svg deleted file mode 100644 index 0998daf..0000000 --- a/doc/fonts/icomoon.svg +++ /dev/null @@ -1,18 +0,0 @@ - - - -Generated by IcoMoon - - - - - - - - - - - - - - \ No newline at end of file diff --git a/doc/fonts/icomoon.ttf b/doc/fonts/icomoon.ttf deleted file mode 100644 index 538fce7b3e34530ac4af0c2d8b5c3447b3020a87..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 2932 zcma)8TWnm#8UE+YIdk@$J?pdgo9(?{(l~aU-SsX_Az3OlQR+lNDFziG)r+&(DOs=M z*b4Sj>dJ(E3#7j`U{pOsr@wQmd z*`4{X^Z)~Br`hA+ob=!~l%lIrZ=^AV{ok!kbUV*>20N5v=n%@i7`B{)NN=)FK;}a3Xibnx zE8xem8=M4qxgBk%`B~z#+~3)4?RI|f@4>*au*avnd^Y$ZJIH>R`B&!4gwNK-{s9}K zzMYBQz=*+;Jt9|#T0eHw_vU;*Podaghy9b?q@U6?EVGt~%U>+23!4cYzha?U!LM0& zMD<#$d7L#{&AEodt`rMWQv#P?*Ig6dJXZ`ev8Zd=H4v}Y>aGLNd}XfT8E!S}?1@w& z8IBmHX@=Vh569z)$#gUrGy)-^`$Ms0Fc`GjmM(;DS-?XHpS_ zBogVNFC*glVmA#lJd}VJ^cevo7>opzu-}5q_RuE?U0BnCsqkwWFIuJ@j9N+vS|R`~ zI~1~mRzRQyriIGda!E!coJgip$wZ{h)S%Btw08yvX`U!Dat6A46ve_}c^IR6)xsIn zW7vH88{J$fxO+I;)lxB?wr#%|+k&EDh^Ye8>69G|+NpG@Sj|j~N23;6G&(+! z$&OD9sfqC{yJ_3$RH;%PlreMhI>W>@6Wg|X|wAEXzZ;E*74&jvI~T^S|h z`A?IA7;jLST14dvPP9;+VXb2fV{we>B9Xy{C62OstqE?cQK{-0vp_NZg=BuhuA5FIlCmzw&W7E<6OmX3Wk^WR=S}K)W zIiv=JCiLFvLq9rjJQh_`$?^Fq&CjpKBV6=O)XoO$tHo5nj+Jr$m_&nZh1fP;hy>!U z)6eK(>e2>Xr_`(9GO_r3)p5{AJ}!lYt8#u1Gj~kWYKuU1yGYro^h7fY#e44 za(v3tWZ51N)Uuo6eZ%nZ^$ zBr;#SV94B(YKS50jD~4Fmw}d)W-v#HuZD=r#(6QkO0)D1J40!_5vV_0CfMg;)nUVA z*~Me;@M~P(Nls#aPNTvBj{oE}k^<#;!csQx?0AsplHxRepS{cEJQ8d4R7J;A^Zt4#R=Vo?>F z^GgWMR#X`>zu$=w8pFshVYbTIk=^H1g)xnRVwW)6L2nJUM%Tyjtjl*)_Fvx9`IUX| zH{{22iuTFTk3QIgsDvac0pC5SWw&9D^Zh`ItQJy|julVn0Bfi!Dd_@q%HbxW|khpO|wpyp#TBw96|2Q z?mKH==DF_j=EcrBoOTiDCS9Zst?#ROqQAP4TkfuQ*E<{C^SQ^ia;KL1XAs*Aop5|_ m!22@HtU3{OaQ}7bJa~`M7RFP~0?xP#K#cAeqqFh*5B~v!Lw&9Q diff --git a/doc/fonts/icomoon.woff b/doc/fonts/icomoon.woff deleted file mode 100644 index a27a656044f41d8a567a5ffa9ed74c75ceae77f8..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 3008 zcma)8U2Ggz6+U{O-{;fWW>9 zxgv=_(4uRDt>=Ng0r`}q$h-8%Yr_?3_b%kDBs*OOgjf04wYnd+qm9?-l;U?i=@uqkoS^(8JFG zP6s#}{XRd+|1$UY+#gaCt}*%Nz^m%;p^6cMCs#%FRT@#}Q!3TNd_Nx3C^6b&e_?m& zF}j6eHd0CXOTE=fz-_;}OwT?GAnHQXfP zzGvS-ve{_*0Rs26g;wCW({8foGO2VlX4$qK?P@%lOr~bCanG~D5n+ZRiL~c=PS-Jo zFdYYc#B;-Qi4-@v){UmqnOFXA>sTt49mld_K`eIHvZG@SFmJ*NTb>sSYtfK{klk@i z2&Qo61k(}MwqAB@*NZz^1XdypD>o8xJtr(s!gE4rT{$Ex7EPrynRF`FWx6+k<8-(N z$7zWuc4Q4qe<{kPlX5YR&Q%9%&`e<9qB-bop>k|bgfq^ z8Mr?#`@;I6p0GccagOe=Z93-futeNthXeD+wa_}p>S`A>8+j%-!OIpF_{A5o$cTFT zwy-`dvuyzoDnBi0DvgvWShTnYXqFb(F1t(-D$9E*R>btm?OiTnW6NFlQKa(`n`W22 zAcroj{haWPnI zT=JUhn7FriBv=)|j8|6x) zQ7JbVDL=}^{vF3YZab=C6^??QIc)!k$o<1_#^8=LLk@XWTekC34pvf|LmwgjHfDSn z>&5UYoub#-MftvVaQ<+aV4laP4ihGiT{7`Hzr}UVCpq8PP>l}Y9!CWXijC_W4b(7` zZw{yM=mRayz5&1UbRNFtaSl8OKF6mW+sG79)oYy?E-8e=o<3qOkK;iiaq}lCAOhpB zLduL@`wF(U9?><8J;wk?b*B9cx#*h9`E?}cYPw9BKdN$s#W?fp=&f^hdjBO|W6WTX z*md+)?7gwo(eZIS>+&6y^o39MzT|%9eG1_@MfuH{U;W~+#YsrxB*v?6S#EX z(5v!XqNr|)JFF(@%^(LxCSch^TAFk@;5qnq!0X_*1Kz+HPXxRGz7+5_*8E%Y`$V^Z zOhV)H8u-dJ619In$Ttbqdj1qWjhkkZZomKv_&JTecj}?N9tu9yU){de+eFYl65Xb2)T7OZS}qRPw+gHM z_5NmWtAC~N>`vkQ%J3p`o2Nw;dkfK5;bvWB)WiMPrz^mor5)7g)c`K~13-@QynJl= E4 - - - - tensorflex v0.1.0 – Documentation - - - - - - diff --git a/doc/search.html b/doc/search.html deleted file mode 100644 index f3bb0da..0000000 --- a/doc/search.html +++ /dev/null @@ -1,96 +0,0 @@ - - - - - - - - Search – tensorflex v0.1.0 - - - - - - - - - - - -
    - - -
    - - - -
    -

    - tensorflex -

    -

    - v0.1.0 -

    -
    - -     - -
    - - -
    - - -
    - -
    - -
    -
    -
    - - - -
    -
    -
    -
    - - - - - - - - diff --git a/lib/tensorflex.ex b/lib/tensorflex.ex index b5a0302..5eddb39 100644 --- a/lib/tensorflex.ex +++ b/lib/tensorflex.ex @@ -1,15 +1,30 @@ defmodule Tensorflex do @moduledoc """ - A simple and fast library for running Tensorflow graph models in Elixir. Tensorflex is written around the [Tensorflow C API](https://www.tensorflow.org/install/install_c), and allows Elixir developers to leverage Machine Learning and Deep Learning solutions in their projects. + A simple and fast library for running Tensorflow graph models in Elixir. +Tensorflex is written around the [Tensorflow C +API](https://www.tensorflow.org/install/install_c), and allows Elixir +developers to leverage Machine Learning and Deep Learning solutions in their +projects. __NOTE__: - - Make sure that the C API version and Python API version (assuming you are using the Python API for first training your models) are the latest. As of July 2018, the latest version is `r1.9`. + - Make sure that the C API version and Python API version (assuming you are + using the Python API for first training your models) are the latest. As of +July 2018, the latest version is `r1.9`. - - Since Tensorflex provides Inference capability for pre-trained graph models, it is assumed you have adequate knowledge of the pre-trained models you are using (such as the input data type/dimensions, input and output operation names, etc.). Some basic understanding of the [Tensorflow Python API](https://www.tensorflow.org/api_docs/python/) can come in very handy. + - Since Tensorflex provides Inference capability for pre-trained graph + models, it is assumed you have adequate knowledge of the pre-trained models +you are using (such as the input data type/dimensions, input and output +operation names, etc.). Some basic understanding of the [Tensorflow Python +API](https://www.tensorflow.org/api_docs/python/) can come in very handy. - - Tensorflex consists of multiple NIFs, so exercise caution while using it-- providing incorrect operation names for running sessions, incorrect dimensions of tensors than the actual pre-trained graph requires, providing different tensor datatypes than the ones required by the graph can all lead to failure. While these are not easy errors to make, do ensure that you test your solution well before deployment. - """ + - Tensorflex consists of multiple NIFs, so exercise caution while using it-- + providing incorrect operation names for running sessions, incorrect +dimensions of tensors than the actual pre-trained graph requires, providing +different tensor datatypes than the ones required by the graph can all lead to +failure. While these are not easy errors to make, do ensure that you test your +solution well before deployment. +""" alias Tensorflex.{NIFs, Graph, Tensor, Matrix} @@ -25,14 +40,18 @@ defmodule Tensorflex do Returns a tuple `{:ok, %Graph}`. - `%Graph` is an internal Tensorflex struct which holds the name of the graph file and the binary definition data that is read in via the `.pb` file. + `%Graph` is an internal Tensorflex struct which holds the name of the graph +file and the binary definition data that is read in via the `.pb` file. ## Examples: _Reading in a graph_ - As an example, we can try reading in the [Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) convolutional neural network based image classification graph model by Google. The graph file is named `classify_image_graph_def.pb`: - ```elixir + As an example, we can try reading in the +[Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) +convolutional neural network based image classification graph model by Google. +The graph file is named `classify_image_graph_def.pb`: +```elixir iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb" 2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization(). {:ok, @@ -41,8 +60,9 @@ defmodule Tensorflex do name: "classify_image_graph_def.pb" }} ``` - Generally to check that the loaded graph model is correct and contains computational operations, the `get_graph_ops/1` function is useful: - ```elixir + Generally to check that the loaded graph model is correct and contains +computational operations, the `get_graph_ops/1` function is useful: +```elixir iex(2)> Tensorflex.get_graph_ops graph ["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims", "ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul", @@ -95,11 +115,13 @@ defmodule Tensorflex do Reads in a Tensorflex ```%Graph``` struct obtained from `read_graph/1`. - Returns a list of all the operation names (as strings) that populate the graph model. + Returns a list of all the operation names (as strings) that populate the +graph model. ## Examples - - _Google Inception CNN Model_ ([source](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz)) + - _Google Inception CNN Model_ + ([source](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz)) ```elixir iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb" @@ -129,7 +151,8 @@ defmodule Tensorflex do "conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...] ``` - - _Iris Dataset MLP Model_ ([source](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/)) + - _Iris Dataset MLP Model_ + ([source](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/)) ```elixir iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_iris.pb" @@ -144,7 +167,8 @@ defmodule Tensorflex do ``` - - _Toy Computational Graph Model_ ([source](https://github.com/anshuman23/tensorflex/tree/master/examples/toy-example)) + - _Toy Computational Graph Model_ + ([source](https://github.com/anshuman23/tensorflex/tree/master/examples/toy-example)) ```elixir iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_toy.pb" @@ -158,7 +182,8 @@ defmodule Tensorflex do ["input", "weights", "weights/read", "biases", "biases/read", "MatMul", "add", "output"] ``` - - _RNN LSTM Sentiment Analysis Model_ ([source](https://github.com/anshuman23/tensorflex/pull/25)) + - _RNN LSTM Sentiment Analysis Model_ + ([source](https://github.com/anshuman23/tensorflex/pull/25)) ```elixir iex(1)> {:ok, graph} = Tensorflex.read_graph "frozen_model_lstm.pb" @@ -198,7 +223,9 @@ defmodule Tensorflex do @doc """ Creates a 2-D Tensorflex matrix from custom input specifications. - Takes three input arguments: number of rows in matrix (`nrows`), number of columns in matrix (`ncols`), and a list of lists of the data that will form the matrix (`datalist`). + Takes three input arguments: number of rows in matrix (`nrows`), number of +columns in matrix (`ncols`), and a list of lists of the data that will form the +matrix (`datalist`). Returns a `%Matrix` Tensorflex struct type. @@ -214,7 +241,9 @@ defmodule Tensorflex do } ``` - All `%Matrix` Tensorflex matrices can be passed in to the other matrix inspection and manipulation functions-- `matrix_pos/3`,`size_of_matrix/1`, `matrix_to_lists/1`, and `append_to_matrix/2`: + All `%Matrix` Tensorflex matrices can be passed in to the other matrix +inspection and manipulation functions-- `matrix_pos/3`,`size_of_matrix/1`, +`matrix_to_lists/1`, and `append_to_matrix/2`: ```elixir iex(1)> mat = Tensorflex.create_matrix(4,4,[[123,431,23,1],[1,2,3,4],[5,6,7,8],[768,564,44,5]]) @@ -287,7 +316,9 @@ defmodule Tensorflex do Used for accessing an element of a Tensorflex matrix. - Takes in three input arguments: a Tensorflex `%Matrix` struct matrix, and the row (`row`) and column (`col`) values of the required element in the matrix. Both `row` and `col` here are __NOT__ zero indexed. + Takes in three input arguments: a Tensorflex `%Matrix` struct matrix, and the +row (`row`) and column (`col`) values of the required element in the matrix. +Both `row` and `col` here are __NOT__ zero indexed. Returns the value as float. @@ -319,7 +350,8 @@ defmodule Tensorflex do Takes a Tensorflex `%Matrix` struct matrix as input. -Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of the matrix and `ncols` represents the number of columns of the matrix. +Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of +the matrix and `ncols` represents the number of columns of the matrix. ## Examples @@ -343,7 +375,9 @@ Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of @doc """ Appends a single row to the back of a Tensorflex matrix. - Takes a Tensorflex `%Matrix` matrix as input and a single row of data (with the same number of columns as the original matrix) as a list of lists (`datalist`) to append to the original matrix. + Takes a Tensorflex `%Matrix` matrix as input and a single row of data (with +the same number of columns as the original matrix) as a list of lists +(`datalist`) to append to the original matrix. Returns the extended and modified `%Matrix` struct matrix. @@ -404,7 +438,10 @@ Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of Returns a list of lists representing the data stored in the matrix. - __NOTE__: If the matrix contains very high dimensional data, typically obtained from a function like `load_csv_as_matrix/2`, then it is not recommended to convert the matrix back to a list of lists format due to a possibility of memory errors. + __NOTE__: If the matrix contains very high dimensional data, typically +obtained from a function like `load_csv_as_matrix/2`, then it is not +recommended to convert the matrix back to a list of lists format due to a +possibility of memory errors. ## Examples @@ -419,11 +456,15 @@ Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of end @doc """ - Creates a `TF_DOUBLE` tensor from Tensorflex matrices containing the values and dimensions specified. + Creates a `TF_DOUBLE` tensor from Tensorflex matrices containing the values +and dimensions specified. - Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the tensor should have and another `%Matrix` matrix (`matrix2`) containing the dimensions of the required tensor. + Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the +tensor should have and another `%Matrix` matrix (`matrix2`) containing the +dimensions of the required tensor. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding tensor data and type. ## Examples: @@ -458,11 +499,13 @@ Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of end @doc """ - Creates a `TF_DOUBLE` constant value one-dimensional tensor from the floating point value specified. + Creates a `TF_DOUBLE` constant value one-dimensional tensor from the floating +point value specified. Takes in a float value as input. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding tensor data and type. ## Examples @@ -493,11 +536,15 @@ Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of end @doc """ - Creates a `TF_FLOAT` tensor from Tensorflex matrices containing the values and dimensions specified. + Creates a `TF_FLOAT` tensor from Tensorflex matrices containing the values +and dimensions specified. - Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the tensor should have and another `%Matrix` matrix (`matrix2`) containing the dimensions of the required tensor. + Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the +tensor should have and another `%Matrix` matrix (`matrix2`) containing the +dimensions of the required tensor. -Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. +Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding tensor data and type. ## Examples: @@ -532,11 +579,13 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl end @doc """ - Creates a `TF_FLOAT` constant value one-dimensional tensor from the floating point value specified. + Creates a `TF_FLOAT` constant value one-dimensional tensor from the floating +point value specified. Takes in a float value as input. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding tensor data and type. ## Examples @@ -567,13 +616,19 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl end @doc """ - Creates a `TF_INT32` tensor from Tensorflex matrices containing the values and dimensions specified. + Creates a `TF_INT32` tensor from Tensorflex matrices containing the values +and dimensions specified. - Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the tensor should have and another `%Matrix` matrix (`matrix2`) containing the dimensions of the required tensor. + Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the +tensor should have and another `%Matrix` matrix (`matrix2`) containing the +dimensions of the required tensor. -Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. +Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding tensor data and type. - __NOTE__: In case floating point values are passed in the values matrix (`matrix1`) as arguments for this function, the tensor will still be created and all the float values will be typecast to integers. + __NOTE__: In case floating point values are passed in the values matrix +(`matrix1`) as arguments for this function, the tensor will still be created +and all the float values will be typecast to integers. ## Examples: @@ -608,11 +663,13 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl end @doc """ - Creates a `TF_INT32` constant value one-dimensional tensor from the integer value specified. + Creates a `TF_INT32` constant value one-dimensional tensor from the integer +value specified. Takes in an integer value as input. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding tensor data and type. ## Examples @@ -643,11 +700,13 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl end @doc """ - Creates a `TF_STRING` constant value string tensor from the string value specified. + Creates a `TF_STRING` constant value string tensor from the string value +specified. Takes in a string value as input. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding tensor data and type. ## Examples @@ -680,15 +739,22 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl @doc """ Allocates a `TF_INT32` tensor of specified dimensions. - This function is generally used to allocate output tensors that do not hold any value data yet, but _will_ after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the `run_session/5` function to hold the output values generated as predictions. + This function is generally used to allocate output tensors that do not hold +any value data yet, but _will_ after the session is run for Inference. Output +tensors of the required dimensions are allocated and then passed to the +`run_session/5` function to hold the output values generated as predictions. Takes a Tensorflex `%Matrix` struct matrix as input. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding the potential tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding the potential tensor data and +type. ## Examples - As an example, we can allocate an `int32` output tensor that will be a vector of 250 values (`1x250` matrix). Therefore, after the session is run, the output will be an `integer` vector containing 250 values: + As an example, we can allocate an `int32` output tensor that will be a vector +of 250 values (`1x250` matrix). Therefore, after the session is run, the output +will be an `integer` vector containing 250 values: ```elixir iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.int32_tensor_alloc @@ -709,15 +775,22 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl @doc """ Allocates a `TF_FLOAT` tensor of specified dimensions. - This function is generally used to allocate output tensors that do not hold any value data yet, but _will_ after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the `run_session/5` function to hold the output values generated as predictions. + This function is generally used to allocate output tensors that do not hold +any value data yet, but _will_ after the session is run for Inference. Output +tensors of the required dimensions are allocated and then passed to the +`run_session/5` function to hold the output values generated as predictions. Takes a Tensorflex `%Matrix` struct matrix as input. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding the potential tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding the potential tensor data and +type. ## Examples - As an example, we can allocate a `float32` output tensor that will be a vector of 250 values (`1x250` matrix). Therefore, after the session is run, the output will be a `float` vector containing 250 values: + As an example, we can allocate a `float32` output tensor that will be a +vector of 250 values (`1x250` matrix). Therefore, after the session is run, the +output will be a `float` vector containing 250 values: ```elixir iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float32_tensor_alloc @@ -738,15 +811,22 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl @doc """ Allocates a `TF_DOUBLE` tensor of specified dimensions. - This function is generally used to allocate output tensors that do not hold any value data yet, but _will_ after the session is run for Inference. Output tensors of the required dimensions are allocated and then passed to the `run_session/5` function to hold the output values generated as predictions. + This function is generally used to allocate output tensors that do not hold +any value data yet, but _will_ after the session is run for Inference. Output +tensors of the required dimensions are allocated and then passed to the +`run_session/5` function to hold the output values generated as predictions. Takes a Tensorflex `%Matrix` struct matrix as input. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding the potential tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding the potential tensor data and +type. ## Examples - As an example, we can allocate a `float64` output tensor that will be a vector of 250 values (`1x250` matrix). Therefore, after the session is run, the output will be a `double` vector containing 250 values: + As an example, we can allocate a `float64` output tensor that will be a +vector of 250 values (`1x250` matrix). Therefore, after the session is run, the +output will be a `double` vector containing 250 values: ```elixir iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float64_tensor_alloc @@ -769,7 +849,10 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl Takes in a `%Tensor` struct tensor as input. - Returns a tuple `{:ok, datatype}` where `datatype` is an atom representing the list of Tensorflow `TF_DataType` tensor datatypes. Click [here](https://github.com/anshuman23/tensorflex/blob/master/c_src/c_api.h#L98-L122) to view a list of all possible datatypes. + Returns a tuple `{:ok, datatype}` where `datatype` is an atom representing +the list of Tensorflow `TF_DataType` tensor datatypes. Click +[here](https://github.com/anshuman23/tensorflex/blob/master/c_src/c_api.h#L98-L122) +to view a list of all possible datatypes. ## Examples @@ -791,19 +874,38 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl end @doc """ - Loads `JPEG` images into Tensorflex directly as a `TF_UINT8` tensor of dimensions `image height x image width x number of color channels`. - - This function is very useful if you wish to do image classification using Convolutional Neural Networks, or other Deep Learning Models. One of the most widely adopted and robust image classification models is the [Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) model by Google. It makes classifications on images from over a 1000 classes with highly accurate results. The `load_image_as_tensor/1` function is an essential component for the prediction pipeline of the Inception model (and for other similar image classification models) to work in Tensorflex. + Loads `JPEG` images into Tensorflex directly as a `TF_UINT8` tensor of +dimensions `image height x image width x number of color channels`. + + This function is very useful if you wish to do image classification using +Convolutional Neural Networks, or other Deep Learning Models. One of the most +widely adopted and robust image classification models is the +[Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) +model by Google. It makes classifications on images from over a 1000 classes +with highly accurate results. The `load_image_as_tensor/1` function is an +essential component for the prediction pipeline of the Inception model (and for +other similar image classification models) to work in Tensorflex. Reads in the path to a `JPEG` image file (`.jpg` or `.jpeg`). - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorflex struct type that is used for holding the tensor data and type. Here the created Tensor is a `uint8` tensor (`TF_UINT8`). + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal +Tensorflex struct type that is used for holding the tensor data and type. Here +the created Tensor is a `uint8` tensor (`TF_UINT8`). - __NOTE__: For now, only 3 channel RGB `JPEG` color images can be passed as arguments. Support for grayscale images and other image formats such as `PNG` will be added in the future. + __NOTE__: For now, only 3 channel RGB `JPEG` color images can be passed as +arguments. Support for grayscale images and other image formats such as `PNG` +will be added in the future. - ## Examples +## Examples - To exemplify the working of the `load_image_as_tensor/1` function we will cover the entire prediction pipeline for the Inception model. However, this makes use of many other Tensorflex functions such as `run_session/5` and the other tensor functions so it would be advisable to go through them first. Also, the Inception model can be downloaded [here](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz). We will be making use of the `cropped_panda.jpg` image file that comes along with the model to test out the model in Tensorflex. + To exemplify the working of the `load_image_as_tensor/1` function we will +cover the entire prediction pipeline for the Inception model. However, this +makes use of many other Tensorflex functions such as `run_session/5` and the +other tensor functions so it would be advisable to go through them first. Also, +the Inception model can be downloaded +[here](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz). +We will be making use of the `cropped_panda.jpg` image file that comes along +with the model to test out the model in Tensorflex. First the graph is loaded: @@ -826,7 +928,9 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl tensor: #Reference<0.1203951739.122552322.52747> }} ``` - Then we create the output tensor which will hold out output vector values. For the Inception model, the output is received as a `1008x1 float32` tensor, as there are 1008 classes in the model: + Then we create the output tensor which will hold out output vector values. +For the Inception model, the output is received as a `1008x1 float32` tensor, +as there are 1008 classes in the model: ```elixir iex(3)> {:ok, output_tensor} = Tensorflex.create_matrix(1,2,[[1008,1]]) |> Tensorflex.float32_tensor_alloc @@ -862,16 +966,31 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl ] ``` - Finally, we need to find which class has the maximum probability and identify it's label. Since `results` is a List of Lists, it's better to read in the flattened list. Then we need to find the index of the element in the new list which as the maximum value. Therefore: - ```elixir + Finally, we need to find which class has the maximum probability and identify +it's label. Since `results` is a List of Lists, it's better to read in the +flattened list. Then we need to find the index of the element in the new list +which as the maximum value. Therefore: +```elixir iex(5)> max_prob = List.flatten(results) |> Enum.max 0.8849328756332397 iex(6)> Enum.find_index(results |> List.flatten, fn(x) -> x == max_prob end) 169 ``` - We can thus see that the class with the maximum probability predicted (`0.8849328756332397`) for the image is `169`. We will now find what the `169` label corresponds to. For this we can look back into the unzipped Inception folder, where there is a file called `imagenet_2012_challenge_label_map_proto.pbtxt`. On opening this file, we can find the string class identifier for the `169` class index. This is `n02510455` and is present on Line 1556 in the file. Finally, we need to match this string identifier to a set of identification labels by referring to the file `imagenet_synset_to_human_label_map.txt` file. Here we can see that corresponding to the string class `n02510455` the human labels are `giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca` (Line 3691 in the file). Thus, we have correctly identified the animal in the image as a panda using Tensorflex. - """ + We can thus see that the class with the maximum probability predicted +(`0.8849328756332397`) for the image is `169`. We will now find what the `169` +label corresponds to. For this we can look back into the unzipped Inception +folder, where there is a file called +`imagenet_2012_challenge_label_map_proto.pbtxt`. On opening this file, we can +find the string class identifier for the `169` class index. This is `n02510455` +and is present on Line 1556 in the file. Finally, we need to match this string +identifier to a set of identification labels by referring to the file +`imagenet_synset_to_human_label_map.txt` file. Here we can see that +corresponding to the string class `n02510455` the human labels are `giant +panda, panda, panda bear, coon bear, Ailuropoda melanoleuca` (Line 3691 in the +file). Thus, we have correctly identified the animal in the image as a panda +using Tensorflex. +""" def load_image_as_tensor(imagepath) do unless File.exists?(imagepath) do @@ -887,18 +1006,34 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl end @doc """ - Loads high-dimensional data from a `CSV` file as a Tensorflex 2-D matrix in a super-fast manner. - - The `load_csv_as_matrix/2` function is very fast-- when compared with the Python based `pandas` library for data science and analysis' function `read_csv` on the `test.csv` file from MNIST Kaggle data ([source](https://www.kaggle.com/c/digit-recognizer/data)), the following execution times were obtained: + Loads high-dimensional data from a `CSV` file as a Tensorflex 2-D matrix in a +super-fast manner. + + The `load_csv_as_matrix/2` function is very fast-- when compared with the +Python based `pandas` library for data science and analysis' function +`read_csv` on the `test.csv` file from MNIST Kaggle data +([source](https://www.kaggle.com/c/digit-recognizer/data)), the following +execution times were obtained: - `read_csv`: `2.549233` seconds - `load_csv_as_matrix/2`: `1.711494` seconds - This function takes in 2 arguments: a path to a valid CSV file (`filepath`) and other optional arguments `opts`. These include whether or not a header needs to be discarded in the CSV, and what the delimiter type is. These are specified by passing in an atom `:true` or `:false` to the `header:` key, and setting a string value for the `delimiter:` key. By default, the header is considered to be present (`:true`) and the delimiter is set to `,`. + This function takes in 2 arguments: a path to a valid CSV file (`filepath`) +and other optional arguments `opts`. These include whether or not a header +needs to be discarded in the CSV, and what the delimiter type is. These are +specified by passing in an atom `:true` or `:false` to the `header:` key, and +setting a string value for the `delimiter:` key. By default, the header is +considered to be present (`:true`) and the delimiter is set to `,`. Returns a `%Matrix` Tensorflex struct type. ## Examples: - We first exemplify the working with the `test.csv` file which belongs to the MNIST Kaggle CSV data ([source](https://www.kaggle.com/c/digit-recognizer/data)), which contains `28000` rows and `784` columns (without the header). It is comma delimited and also contains a header. From the `test.csv` file, we also create a custom file withou the header present which we refer to as `test_without_header.csv` in the examples below: + We first exemplify the working with the `test.csv` file which belongs to the +MNIST Kaggle CSV data +([source](https://www.kaggle.com/c/digit-recognizer/data)), which contains +`28000` rows and `784` columns (without the header). It is comma delimited and +also contains a header. From the `test.csv` file, we also create a custom file +withou the header present which we refer to as `test_without_header.csv` in the +examples below: ```elixir iex(1)> mat = Tensorflex.load_csv_as_matrix("test.csv") @@ -915,8 +1050,10 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl 13.0 ``` - On a visual inspection of the very large `test.csv` file, one can see that the values in these particular positions are correct. Now we show usage for the same file but without header, `test_without_header.csv`: - ```elixir + On a visual inspection of the very large `test.csv` file, one can see that +the values in these particular positions are correct. Now we show usage for the +same file but without header, `test_without_header.csv`: +```elixir iex(1)> no_header = Tensorflex.load_csv_as_matrix("test/test_without_header.csv", header: :false) %Tensorflex.Matrix{ data: #Reference<0.4024686574.590479364.257078>, @@ -931,7 +1068,8 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl 13.0 ``` - Next we see the delimiter functionalities. First, assuming we have two simple `CSV` files, `sample1.csv` and `sample2.csv` + Next we see the delimiter functionalities. First, assuming we have two simple +`CSV` files, `sample1.csv` and `sample2.csv` _sample1.csv_: @@ -1008,22 +1146,41 @@ Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal Tensorfl end @doc """ - Runs a Tensorflow session to generate predictions for a given graph, input data, and required input/output operations. - - This function is the final step of the Inference (prediction) pipeline and generates output for a given set of input data, a pre-trained graph model, and the specified input and output operations of the graph. - - Takes in five arguments: a pre-trained Tensorflow graph `.pb` model read in from the `read_graph/1` function (`graph`), an input tensor with the dimensions and data required for the input operation of the graph to run (`tensor1`), an output tensor allocated with the right dimensions (`tensor2`), the name of the input operation of the graph that needs where the input data is fed (`input_opname`), and the output operation name in the graph where the outputs are obtained (`output_opname`). The input tensor is generally created from the matrices manually or using the `load_csv_as_matrix/2` function, and then passed through to one of the tensor creation functions. For image classification the `load_image_as_tensor/1` can also be used to create the input tensor from an image. The output tensor is created using the tensor allocation functions (generally containing `alloc` at the end of the function name). - - Returns a List of Lists (similar to the `matrix_to_lists/1` function) containing the generated predictions as per the output tensor dimensions. + Runs a Tensorflow session to generate predictions for a given graph, input +data, and required input/output operations. + + This function is the final step of the Inference (prediction) pipeline and +generates output for a given set of input data, a pre-trained graph model, and +the specified input and output operations of the graph. + + Takes in five arguments: a pre-trained Tensorflow graph `.pb` model read in +from the `read_graph/1` function (`graph`), an input tensor with the dimensions +and data required for the input operation of the graph to run (`tensor1`), an +output tensor allocated with the right dimensions (`tensor2`), the name of the +input operation of the graph that needs where the input data is fed +(`input_opname`), and the output operation name in the graph where the outputs +are obtained (`output_opname`). The input tensor is generally created from the +matrices manually or using the `load_csv_as_matrix/2` function, and then passed +through to one of the tensor creation functions. For image classification the +`load_image_as_tensor/1` can also be used to create the input tensor from an +image. The output tensor is created using the tensor allocation functions +(generally containing `alloc` at the end of the function name). + + Returns a List of Lists (similar to the `matrix_to_lists/1` function) +containing the generated predictions as per the output tensor dimensions. ## Examples - - A blog post [here](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/) covers generating predictions and running sessions using an MLP model on the Iris Dataset + - A blog post [here](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/) covers + generating predictions and running sessions using an MLP model on the Iris +Dataset - - Generating predictions from the Inception model by Google is covered in the `load_image_as_tensor/1` function examples. + - Generating predictions from the Inception model by Google is covered in the + `load_image_as_tensor/1` function examples. - - Working with an RNN-LSTM example for sentiment analysis is covered [here](https://github.com/anshuman23/tensorflex/pull/25). - """ + - Working with an RNN-LSTM example for sentiment analysis is covered + [here](https://github.com/anshuman23/tensorflex/pull/25). +""" def run_session(%Graph{def: graphdef, name: filepath}, %Tensor{datatype: input_datatype, tensor: input_ref}, %Tensor{datatype: output_datatype, tensor: output_ref}, input_opname, output_opname) do NIFs.run_session(graphdef, input_ref, output_ref, input_opname, output_opname) From cc42b7ce9ee371bed3a8a5de54fb35c7f4c16786 Mon Sep 17 00:00:00 2001 From: Anshuman Chhabra Date: Sun, 29 Jul 2018 23:50:24 +0530 Subject: [PATCH 4/5] Fixed IEx output indentation --- lib/tensorflex.ex | 335 +++++++++++++++++++++++----------------------- 1 file changed, 168 insertions(+), 167 deletions(-) diff --git a/lib/tensorflex.ex b/lib/tensorflex.ex index 5eddb39..52a69d7 100644 --- a/lib/tensorflex.ex +++ b/lib/tensorflex.ex @@ -56,8 +56,8 @@ The graph file is named `classify_image_graph_def.pb`: 2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization(). {:ok, %Tensorflex.Graph{ - def: #Reference<0.3018278404.759824385.5268>, - name: "classify_image_graph_def.pb" + def: #Reference<0.3018278404.759824385.5268>, + name: "classify_image_graph_def.pb" }} ``` Generally to check that the loaded graph model is correct and contains @@ -65,21 +65,21 @@ computational operations, the `get_graph_ops/1` function is useful: ```elixir iex(2)> Tensorflex.get_graph_ops graph ["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims", - "ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul", - "conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta", - "conv/batchnorm/gamma", "conv/batchnorm/moving_mean", - "conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics", - "conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D", - "conv_1/batchnorm/beta", "conv_1/batchnorm/gamma", - "conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance", - "conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency", - "conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta", - "conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean", - "conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics", - "conv_2/control_dependency", "conv_2", "pool/CheckNumerics", - "pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D", - "conv_3/batchnorm/beta", "conv_3/batchnorm/gamma", - "conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...] + "ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul", + "conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta", + "conv/batchnorm/gamma", "conv/batchnorm/moving_mean", + "conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics", + "conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D", + "conv_1/batchnorm/beta", "conv_1/batchnorm/gamma", + "conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance", + "conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency", + "conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta", + "conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean", + "conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics", + "conv_2/control_dependency", "conv_2", "pool/CheckNumerics", + "pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D", + "conv_3/batchnorm/beta", "conv_3/batchnorm/gamma", + "conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...] ``` @@ -128,27 +128,27 @@ graph model. 2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization(). {:ok, %Tensorflex.Graph{ - def: #Reference<0.3018278404.759824385.5268>, - name: "classify_image_graph_def.pb" + def: #Reference<0.3018278404.759824385.5268>, + name: "classify_image_graph_def.pb" }} iex(2)> Tensorflex.get_graph_ops graph ["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims", - "ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul", - "conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta", - "conv/batchnorm/gamma", "conv/batchnorm/moving_mean", - "conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics", - "conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D", - "conv_1/batchnorm/beta", "conv_1/batchnorm/gamma", - "conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance", - "conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency", - "conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta", - "conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean", - "conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics", - "conv_2/control_dependency", "conv_2", "pool/CheckNumerics", - "pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D", - "conv_3/batchnorm/beta", "conv_3/batchnorm/gamma", - "conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...] + "ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul", + "conv/conv2d_params", "conv/Conv2D", "conv/batchnorm/beta", + "conv/batchnorm/gamma", "conv/batchnorm/moving_mean", + "conv/batchnorm/moving_variance", "conv/batchnorm", "conv/CheckNumerics", + "conv/control_dependency", "conv", "conv_1/conv2d_params", "conv_1/Conv2D", + "conv_1/batchnorm/beta", "conv_1/batchnorm/gamma", + "conv_1/batchnorm/moving_mean", "conv_1/batchnorm/moving_variance", + "conv_1/batchnorm", "conv_1/CheckNumerics", "conv_1/control_dependency", + "conv_1", "conv_2/conv2d_params", "conv_2/Conv2D", "conv_2/batchnorm/beta", + "conv_2/batchnorm/gamma", "conv_2/batchnorm/moving_mean", + "conv_2/batchnorm/moving_variance", "conv_2/batchnorm", "conv_2/CheckNumerics", + "conv_2/control_dependency", "conv_2", "pool/CheckNumerics", + "pool/control_dependency", "pool", "conv_3/conv2d_params", "conv_3/Conv2D", + "conv_3/batchnorm/beta", "conv_3/batchnorm/gamma", + "conv_3/batchnorm/moving_mean", "conv_3/batchnorm/moving_variance", ...] ``` - _Iris Dataset MLP Model_ @@ -158,8 +158,8 @@ graph model. iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_iris.pb" {:ok, %Tensorflex.Graph{ - def: #Reference<0.4109712726.1847984130.24506>, - name: "graphdef_iris.pb" + def: #Reference<0.4109712726.1847984130.24506>, + name: "graphdef_iris.pb" }} iex(2)> Tensorflex.get_graph_ops graph @@ -174,8 +174,8 @@ graph model. iex(1)> {:ok, graph} = Tensorflex.read_graph "graphdef_toy.pb" {:ok, %Tensorflex.Graph{ - def: #Reference<0.1274892327.1580335105.235135>, - name: "graphdef_toy.pb" + def: #Reference<0.1274892327.1580335105.235135>, + name: "graphdef_toy.pb" }} iex(2)> Tensorflex.get_graph_ops graph @@ -189,30 +189,30 @@ graph model. iex(1)> {:ok, graph} = Tensorflex.read_graph "frozen_model_lstm.pb" {:ok, %Tensorflex.Graph{ - def: #Reference<0.713975820.1050542081.11558>, - name: "frozen_model_lstm.pb" + def: #Reference<0.713975820.1050542081.11558>, + name: "frozen_model_lstm.pb" }} iex(2)> Tensorflex.get_graph_ops graph ["Placeholder_1", "embedding_lookup/params_0", "embedding_lookup", - "transpose/perm", "transpose", "rnn/Shape", "rnn/strided_slice/stack", - "rnn/strided_slice/stack_1", "rnn/strided_slice/stack_2", "rnn/strided_slice", - "rnn/stack/1", "rnn/stack", "rnn/zeros/Const", "rnn/zeros", "rnn/stack_1/1", - "rnn/stack_1", "rnn/zeros_1/Const", "rnn/zeros_1", "rnn/Shape_1", - "rnn/strided_slice_2/stack", "rnn/strided_slice_2/stack_1", - "rnn/strided_slice_2/stack_2", "rnn/strided_slice_2", "rnn/time", - "rnn/TensorArray", "rnn/TensorArray_1", "rnn/TensorArrayUnstack/Shape", - "rnn/TensorArrayUnstack/strided_slice/stack", - "rnn/TensorArrayUnstack/strided_slice/stack_1", - "rnn/TensorArrayUnstack/strided_slice/stack_2", - "rnn/TensorArrayUnstack/strided_slice", "rnn/TensorArrayUnstack/range/start", - "rnn/TensorArrayUnstack/range/delta", "rnn/TensorArrayUnstack/range", - "rnn/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3", - "rnn/while/Enter", "rnn/while/Enter_1", "rnn/while/Enter_2", - "rnn/while/Enter_3", "rnn/while/Merge", "rnn/while/Merge_1", - "rnn/while/Merge_2", "rnn/while/Merge_3", "rnn/while/Less/Enter", - "rnn/while/Less", "rnn/while/LoopCond", "rnn/while/Switch", - "rnn/while/Switch_1", "rnn/while/Switch_2", "rnn/while/Switch_3", ...] + "transpose/perm", "transpose", "rnn/Shape", "rnn/strided_slice/stack", + "rnn/strided_slice/stack_1", "rnn/strided_slice/stack_2", "rnn/strided_slice", + "rnn/stack/1", "rnn/stack", "rnn/zeros/Const", "rnn/zeros", "rnn/stack_1/1", + "rnn/stack_1", "rnn/zeros_1/Const", "rnn/zeros_1", "rnn/Shape_1", + "rnn/strided_slice_2/stack", "rnn/strided_slice_2/stack_1", + "rnn/strided_slice_2/stack_2", "rnn/strided_slice_2", "rnn/time", + "rnn/TensorArray", "rnn/TensorArray_1", "rnn/TensorArrayUnstack/Shape", + "rnn/TensorArrayUnstack/strided_slice/stack", + "rnn/TensorArrayUnstack/strided_slice/stack_1", + "rnn/TensorArrayUnstack/strided_slice/stack_2", + "rnn/TensorArrayUnstack/strided_slice", "rnn/TensorArrayUnstack/range/start", + "rnn/TensorArrayUnstack/range/delta", "rnn/TensorArrayUnstack/range", + "rnn/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3", + "rnn/while/Enter", "rnn/while/Enter_1", "rnn/while/Enter_2", + "rnn/while/Enter_3", "rnn/while/Merge", "rnn/while/Merge_1", + "rnn/while/Merge_2", "rnn/while/Merge_3", "rnn/while/Less/Enter", + "rnn/while/Less", "rnn/while/LoopCond", "rnn/while/Switch", + "rnn/while/Switch_1", "rnn/while/Switch_2", "rnn/while/Switch_3", ...] ``` """ @@ -234,10 +234,11 @@ matrix (`datalist`). _Creating a new matrix_ ```elixir - iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]]) %Tensorflex.Matrix{ - data: #Reference<0.759278808.823525378.128525>, - ncols: 3, - nrows: 2 + iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]]) + %Tensorflex.Matrix{ + data: #Reference<0.759278808.823525378.128525>, + ncols: 3, + nrows: 2 } ``` @@ -248,25 +249,25 @@ inspection and manipulation functions-- `matrix_pos/3`,`size_of_matrix/1`, ```elixir iex(1)> mat = Tensorflex.create_matrix(4,4,[[123,431,23,1],[1,2,3,4],[5,6,7,8],[768,564,44,5]]) %Tensorflex.Matrix{ - data: #Reference<0.878138179.2435973124.131489>, - ncols: 4, - nrows: 4 + data: #Reference<0.878138179.2435973124.131489>, + ncols: 4, + nrows: 4 } iex(2)> mat = Tensorflex.append_to_matrix(mat, [[1,1,1,1]]) %Tensorflex.Matrix{ - data: #Reference<0.878138179.2435973124.131489>, - ncols: 4, - nrows: 5 + data: #Reference<0.878138179.2435973124.131489>, + ncols: 4, + nrows: 5 } iex(3)> Tensorflex.matrix_to_lists mat [ - [123.0, 431.0, 23.0, 1.0], - [1.0, 2.0, 3.0, 4.0], - [5.0, 6.0, 7.0, 8.0], - [768.0, 564.0, 44.0, 5.0], - [1.0, 1.0, 1.0, 1.0] + [123.0, 431.0, 23.0, 1.0], + [1.0, 2.0, 3.0, 4.0], + [5.0, 6.0, 7.0, 8.0], + [768.0, 564.0, 44.0, 5.0], + [1.0, 1.0, 1.0, 1.0] ] iex(4)> Tensorflex.matrix_pos(mat,5,3) @@ -327,9 +328,9 @@ Both `row` and `col` here are __NOT__ zero indexed. ```elixir iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]]) %Tensorflex.Matrix{ - data: #Reference<0.759278808.823525378.128525>, - ncols: 3, - nrows: 2 + data: #Reference<0.759278808.823525378.128525>, + ncols: 3, + nrows: 2 } iex(2)> Tensorflex.matrix_pos(mat,2,1) @@ -358,9 +359,9 @@ the matrix and `ncols` represents the number of columns of the matrix. ```elixir iex(1)> mat = Tensorflex.create_matrix(2,3,[[2.2,1.3,44.5],[5.5,6.1,3.333]]) %Tensorflex.Matrix{ - data: #Reference<0.759278808.823525378.128525>, - ncols: 3, - nrows: 2 + data: #Reference<0.759278808.823525378.128525>, + ncols: 3, + nrows: 2 } iex(2)> Tensorflex.size_of_matrix mat @@ -386,23 +387,23 @@ the same number of columns as the original matrix) as a list of lists ```elixir iex(1)> m = Tensorflex.create_matrix(2,3,[[23,23,23],[32,32,32]]) %Tensorflex.Matrix{ - data: #Reference<0.153563642.2042232833.193025>, - ncols: 3, - nrows: 2 + data: #Reference<0.153563642.2042232833.193025>, + ncols: 3, + nrows: 2 } iex(2)> m = Tensorflex.append_to_matrix(m,[[2,2,2]]) %Tensorflex.Matrix{ - data: #Reference<0.153563642.2042232833.193025>, - ncols: 3, - nrows: 3 + data: #Reference<0.153563642.2042232833.193025>, + ncols: 3, + nrows: 3 } iex(3)> m = Tensorflex.append_to_matrix(m,[[3,3,3]]) %Tensorflex.Matrix{ - data: #Reference<0.153563642.2042232833.193025>, - ncols: 3, - nrows: 4 + data: #Reference<0.153563642.2042232833.193025>, + ncols: 3, + nrows: 4 } iex(4)> m |> Tensorflex.matrix_to_lists @@ -471,23 +472,23 @@ Tensorflex struct type that is used for holding tensor data and type. ```elixir iex(1)> vals = Tensorflex.create_matrix(2,3,[[12.0,45.2,2.11],[36.7,8.09,9.81]]) %Tensorflex.Matrix{ - data: #Reference<0.1251941183.3671982081.254268>, - ncols: 3, - nrows: 2 + data: #Reference<0.1251941183.3671982081.254268>, + ncols: 3, + nrows: 2 } iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]]) %Tensorflex.Matrix{ - data: #Reference<0.1251941183.3671982081.254723>, - ncols: 2, - nrows: 1 + data: #Reference<0.1251941183.3671982081.254723>, + ncols: 2, + nrows: 1 } iex(3)> {:ok, tensor} = Tensorflex.float64_tensor vals,dims {:ok, %Tensorflex.Tensor{ - datatype: :tf_double, - tensor: #Reference<0.1251941183.3671982081.255216> + datatype: :tf_double, + tensor: #Reference<0.1251941183.3671982081.255216> }} ``` @@ -513,8 +514,8 @@ Tensorflex struct type that is used for holding tensor data and type. iex(1)> {:ok, tensor} = Tensorflex.float64_tensor 123.123 {:ok, %Tensorflex.Tensor{ - datatype: :tf_double, - tensor: #Reference<0.2778616536.4219338753.155412> + datatype: :tf_double, + tensor: #Reference<0.2778616536.4219338753.155412> }} ``` @@ -551,23 +552,23 @@ Tensorflex struct type that is used for holding tensor data and type. ```elixir iex(1)> vals = Tensorflex.create_matrix(2,3,[[12.0,45.2,2.11],[36.7,8.09,9.81]]) %Tensorflex.Matrix{ - data: #Reference<0.1251941183.3671982081.254268>, - ncols: 3, - nrows: 2 + data: #Reference<0.1251941183.3671982081.254268>, + ncols: 3, + nrows: 2 } iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]]) %Tensorflex.Matrix{ - data: #Reference<0.1251941183.3671982081.254723>, - ncols: 2, - nrows: 1 + data: #Reference<0.1251941183.3671982081.254723>, + ncols: 2, + nrows: 1 } iex(3)> {:ok, tensor} = Tensorflex.float32_tensor vals,dims {:ok, %Tensorflex.Tensor{ - datatype: :tf_float, - tensor: #Reference<0.1251941183.3671982081.255228> + datatype: :tf_float, + tensor: #Reference<0.1251941183.3671982081.255228> }} ``` @@ -593,8 +594,8 @@ Tensorflex struct type that is used for holding tensor data and type. iex(1)> {:ok, tensor} = Tensorflex.float32_tensor 123.123 {:ok, %Tensorflex.Tensor{ - datatype: :tf_float, - tensor: #Reference<0.2011963375.1804468228.236110> + datatype: :tf_float, + tensor: #Reference<0.2011963375.1804468228.236110> }} ``` @@ -635,23 +636,23 @@ and all the float values will be typecast to integers. ```elixir iex(1)> vals = Tensorflex.create_matrix(2,3,[[123,45,333],[2,2,899]]) %Tensorflex.Matrix{ - data: #Reference<0.1256144000.2868510721.170449>, - ncols: 3, - nrows: 2 + data: #Reference<0.1256144000.2868510721.170449>, + ncols: 3, + nrows: 2 } iex(2)> dims = Tensorflex.create_matrix(1,2,[[2,3]]) %Tensorflex.Matrix{ - data: #Reference<0.1256144000.2868510721.170894>, - ncols: 2, - nrows: 1 + data: #Reference<0.1256144000.2868510721.170894>, + ncols: 2, + nrows: 1 } iex(3)> {:ok, tensor} = Tensorflex.int32_tensor vals,dims {:ok, %Tensorflex.Tensor{ - datatype: :tf_int32, - tensor: #Reference<0.1256144000.2868510721.171357> + datatype: :tf_int32, + tensor: #Reference<0.1256144000.2868510721.171357> }} ``` @@ -677,8 +678,8 @@ Tensorflex struct type that is used for holding tensor data and type. iex(1)> {:ok, tensor} = Tensorflex.int32_tensor 123 {:ok, %Tensorflex.Tensor{ - datatype: :tf_int32, - tensor: #Reference<0.1927663658.3415343105.162588> + datatype: :tf_int32, + tensor: #Reference<0.1927663658.3415343105.162588> }} ``` @@ -714,8 +715,8 @@ Tensorflex struct type that is used for holding tensor data and type. iex(1)> {:ok, tensor} = Tensorflex.string_tensor "123.123" {:ok, %Tensorflex.Tensor{ - datatype: :tf_string, - tensor: #Reference<0.2069282048.194904065.41126> + datatype: :tf_string, + tensor: #Reference<0.2069282048.194904065.41126> }} ``` @@ -760,8 +761,8 @@ will be an `integer` vector containing 250 values: iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.int32_tensor_alloc {:ok, %Tensorflex.Tensor{ - datatype: :tf_int32, - tensor: #Reference<0.961157994.2087059457.18950> + datatype: :tf_int32, + tensor: #Reference<0.961157994.2087059457.18950> }} ``` @@ -796,8 +797,8 @@ output will be a `float` vector containing 250 values: iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float32_tensor_alloc {:ok, %Tensorflex.Tensor{ - datatype: :tf_float, - tensor: #Reference<0.961157994.2087059457.19014> + datatype: :tf_float, + tensor: #Reference<0.961157994.2087059457.19014> }} ``` @@ -832,8 +833,8 @@ output will be a `double` vector containing 250 values: iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float64_tensor_alloc {:ok, %Tensorflex.Tensor{ - datatype: :tf_double, - tensor: #Reference<0.961157994.2087059457.19025> + datatype: :tf_double, + tensor: #Reference<0.961157994.2087059457.19025> }} ``` @@ -860,8 +861,8 @@ to view a list of all possible datatypes. iex(1)> {:ok, tensor} = Tensorflex.string_tensor "example" {:ok, %Tensorflex.Tensor{ - datatype: :tf_string, - tensor: #Reference<0.4132928949.2894987267.194583> + datatype: :tf_string, + tensor: #Reference<0.4132928949.2894987267.194583> }} iex(2)> Tensorflex.tensor_datatype tensor @@ -914,8 +915,8 @@ with the model to test out the model in Tensorflex. 2018-07-25 14:20:29.079139: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization(). {:ok, %Tensorflex.Graph{ - def: #Reference<0.542869014.389152771.105680>, - name: "classify_image_graph_def.pb" + def: #Reference<0.542869014.389152771.105680>, + name: "classify_image_graph_def.pb" }} ``` Then we load the image as a `uint8` tensor: @@ -924,8 +925,8 @@ with the model to test out the model in Tensorflex. iex(2)> {:ok, input_tensor} = Tensorflex.load_image_as_tensor("cropped_panda.jpg") {:ok, %Tensorflex.Tensor{ - datatype: :tf_uint8, - tensor: #Reference<0.1203951739.122552322.52747> + datatype: :tf_uint8, + tensor: #Reference<0.1203951739.122552322.52747> }} ``` Then we create the output tensor which will hold out output vector values. @@ -936,8 +937,8 @@ as there are 1008 classes in the model: iex(3)> {:ok, output_tensor} = Tensorflex.create_matrix(1,2,[[1008,1]]) |> Tensorflex.float32_tensor_alloc {:ok, %Tensorflex.Tensor{ - datatype: :tf_float, - tensor: #Reference<0.1203951739.122552322.52794> + datatype: :tf_float, + tensor: #Reference<0.1203951739.122552322.52794> }} ``` Next, we obtain the results by running the session: @@ -946,23 +947,23 @@ as there are 1008 classes in the model: iex(4)> results = Tensorflex.run_session(graph, input_tensor, output_tensor, "DecodeJpeg", "softmax") 2018-07-25 14:33:40.992813: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA [ - [1.059142014128156e-4, 2.8240500250831246e-4, 8.30648496048525e-5, - 1.2982363114133477e-4, 7.32232874725014e-5, 8.014426566660404e-5, - 6.63459359202534e-5, 0.003170756157487631, 7.931600703159347e-5, - 3.707312498590909e-5, 3.0997329304227605e-5, 1.4232713147066534e-4, - 1.0381334868725389e-4, 1.1057958181481808e-4, 1.4321311027742922e-4, - 1.203602587338537e-4, 1.3130248407833278e-4, 5.850398520124145e-5, - 2.641105093061924e-4, 3.1629020668333396e-5, 3.906813799403608e-5, - 2.8646905775531195e-5, 2.2863158665131778e-4, 1.2222197256051004e-4, - 5.956588938715868e-5, 5.421260357252322e-5, 5.996063555357978e-5, - 4.867801326327026e-4, 1.1005574924638495e-4, 2.3433618480339646e-4, - 1.3062104699201882e-4, 1.317620772169903e-4, 9.388553007738665e-5, - 7.076268957462162e-5, 4.281177825760096e-5, 1.6863139171618968e-4, - 9.093972039408982e-5, 2.611844101920724e-4, 2.7584232157096267e-4, - 5.157176201464608e-5, 2.144951868103817e-4, 1.3628098531626165e-4, - 8.007588621694595e-5, 1.7929042223840952e-4, 2.2831936075817794e-4, - 6.216531619429588e-5, 3.736453436431475e-5, 6.782123091397807e-5, - 1.1538144462974742e-4, ...] + [1.059142014128156e-4, 2.8240500250831246e-4, 8.30648496048525e-5, + 1.2982363114133477e-4, 7.32232874725014e-5, 8.014426566660404e-5, + 6.63459359202534e-5, 0.003170756157487631, 7.931600703159347e-5, + 3.707312498590909e-5, 3.0997329304227605e-5, 1.4232713147066534e-4, + 1.0381334868725389e-4, 1.1057958181481808e-4, 1.4321311027742922e-4, + 1.203602587338537e-4, 1.3130248407833278e-4, 5.850398520124145e-5, + 2.641105093061924e-4, 3.1629020668333396e-5, 3.906813799403608e-5, + 2.8646905775531195e-5, 2.2863158665131778e-4, 1.2222197256051004e-4, + 5.956588938715868e-5, 5.421260357252322e-5, 5.996063555357978e-5, + 4.867801326327026e-4, 1.1005574924638495e-4, 2.3433618480339646e-4, + 1.3062104699201882e-4, 1.317620772169903e-4, 9.388553007738665e-5, + 7.076268957462162e-5, 4.281177825760096e-5, 1.6863139171618968e-4, + 9.093972039408982e-5, 2.611844101920724e-4, 2.7584232157096267e-4, + 5.157176201464608e-5, 2.144951868103817e-4, 1.3628098531626165e-4, + 8.007588621694595e-5, 1.7929042223840952e-4, 2.2831936075817794e-4, + 6.216531619429588e-5, 3.736453436431475e-5, 6.782123091397807e-5, + 1.1538144462974742e-4, ...] ] ``` @@ -1038,9 +1039,9 @@ examples below: ```elixir iex(1)> mat = Tensorflex.load_csv_as_matrix("test.csv") %Tensorflex.Matrix{ - data: #Reference<0.4024686574.590479361.258459>, - ncols: 784, - nrows: 28000 + data: #Reference<0.4024686574.590479361.258459>, + ncols: 784, + nrows: 28000 } iex(2)> Tensorflex.matrix_pos mat, 5,97 @@ -1056,9 +1057,9 @@ same file but without header, `test_without_header.csv`: ```elixir iex(1)> no_header = Tensorflex.load_csv_as_matrix("test/test_without_header.csv", header: :false) %Tensorflex.Matrix{ - data: #Reference<0.4024686574.590479364.257078>, - ncols: 784, - nrows: 28000 + data: #Reference<0.4024686574.590479364.257078>, + ncols: 784, + nrows: 28000 } iex(2)> Tensorflex.matrix_pos no_header,5,97 @@ -1092,23 +1093,23 @@ same file but without header, `test_without_header.csv`: ```elixir iex(1)> m1 = Tensorflex.load_csv_as_matrix("sample1.csv", header: :false) %Tensorflex.Matrix{ - data: #Reference<0.3878093040.3013214209.247502>, - ncols: 5, - nrows: 3 + data: #Reference<0.3878093040.3013214209.247502>, + ncols: 5, + nrows: 3 } iex(2)> Tensorflex.matrix_to_lists m1 [ - [1.0, 2.0, 3.0, 4.0, 5.0], - [6.0, 7.0, 8.0, 9.0, 10.0], - [11.0, 12.0, 13.0, 14.0, 15.0] + [1.0, 2.0, 3.0, 4.0, 5.0], + [6.0, 7.0, 8.0, 9.0, 10.0], + [11.0, 12.0, 13.0, 14.0, 15.0] ] iex(3)> m2 = Tensorflex.load_csv_as_matrix("sample2.csv", header: :true, delimiter: "-") %Tensorflex.Matrix{ - data: #Reference<0.4024686574.590479361.258952>, - ncols: 4, - nrows: 3 + data: #Reference<0.4024686574.590479361.258952>, + ncols: 4, + nrows: 3 } iex(4)> Tensorflex.matrix_to_lists m2 @@ -1169,7 +1170,7 @@ image. The output tensor is created using the tensor allocation functions Returns a List of Lists (similar to the `matrix_to_lists/1` function) containing the generated predictions as per the output tensor dimensions. - ## Examples +## Examples - A blog post [here](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/) covers generating predictions and running sessions using an MLP model on the Iris From 60dc1e416fefa82c30e0d65242047bd51a563bd7 Mon Sep 17 00:00:00 2001 From: Anshuman Chhabra Date: Mon, 30 Jul 2018 01:11:28 +0530 Subject: [PATCH 5/5] Corrected improper indentation --- lib/tensorflex.ex | 323 +++++++++++++++++++++++----------------------- 1 file changed, 162 insertions(+), 161 deletions(-) diff --git a/lib/tensorflex.ex b/lib/tensorflex.ex index 52a69d7..afdaddd 100644 --- a/lib/tensorflex.ex +++ b/lib/tensorflex.ex @@ -10,20 +10,20 @@ projects. - Make sure that the C API version and Python API version (assuming you are using the Python API for first training your models) are the latest. As of -July 2018, the latest version is `r1.9`. + July 2018, the latest version is `r1.9`. - Since Tensorflex provides Inference capability for pre-trained graph models, it is assumed you have adequate knowledge of the pre-trained models -you are using (such as the input data type/dimensions, input and output -operation names, etc.). Some basic understanding of the [Tensorflow Python -API](https://www.tensorflow.org/api_docs/python/) can come in very handy. + you are using (such as the input data type/dimensions, input and output + operation names, etc.). Some basic understanding of the [Tensorflow Python + API](https://www.tensorflow.org/api_docs/python/) can come in very handy. - Tensorflex consists of multiple NIFs, so exercise caution while using it-- providing incorrect operation names for running sessions, incorrect -dimensions of tensors than the actual pre-trained graph requires, providing -different tensor datatypes than the ones required by the graph can all lead to -failure. While these are not easy errors to make, do ensure that you test your -solution well before deployment. + dimensions of tensors than the actual pre-trained graph requires, providing + different tensor datatypes than the ones required by the graph can all lead to + failure. While these are not easy errors to make, do ensure that you test your + solution well before deployment. """ alias Tensorflex.{NIFs, Graph, Tensor, Matrix} @@ -48,10 +48,11 @@ file and the binary definition data that is read in via the `.pb` file. _Reading in a graph_ As an example, we can try reading in the -[Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) -convolutional neural network based image classification graph model by Google. -The graph file is named `classify_image_graph_def.pb`: -```elixir + [Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) + convolutional neural network based image classification graph model by Google. + The graph file is named `classify_image_graph_def.pb`: + + ```elixir iex(1)> {:ok, graph} = Tensorflex.read_graph "classify_image_graph_def.pb" 2018-07-23 15:31:35.949345: W tensorflow/core/framework/op_def_util.cc:346] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization(). {:ok, @@ -61,8 +62,8 @@ The graph file is named `classify_image_graph_def.pb`: }} ``` Generally to check that the loaded graph model is correct and contains -computational operations, the `get_graph_ops/1` function is useful: -```elixir + computational operations, the `get_graph_ops/1` function is useful: + ```elixir iex(2)> Tensorflex.get_graph_ops graph ["DecodeJpeg/contents", "DecodeJpeg", "Cast", "ExpandDims/dim", "ExpandDims", "ResizeBilinear/size", "ResizeBilinear", "Sub/y", "Sub", "Mul/y", "Mul", @@ -116,7 +117,7 @@ computational operations, the `get_graph_ops/1` function is useful: Reads in a Tensorflex ```%Graph``` struct obtained from `read_graph/1`. Returns a list of all the operation names (as strings) that populate the -graph model. + graph model. ## Examples @@ -224,8 +225,8 @@ graph model. Creates a 2-D Tensorflex matrix from custom input specifications. Takes three input arguments: number of rows in matrix (`nrows`), number of -columns in matrix (`ncols`), and a list of lists of the data that will form the -matrix (`datalist`). + columns in matrix (`ncols`), and a list of lists of the data that will form the + matrix (`datalist`). Returns a `%Matrix` Tensorflex struct type. @@ -243,8 +244,8 @@ matrix (`datalist`). ``` All `%Matrix` Tensorflex matrices can be passed in to the other matrix -inspection and manipulation functions-- `matrix_pos/3`,`size_of_matrix/1`, -`matrix_to_lists/1`, and `append_to_matrix/2`: + inspection and manipulation functions-- `matrix_pos/3`,`size_of_matrix/1`, + `matrix_to_lists/1`, and `append_to_matrix/2`: ```elixir iex(1)> mat = Tensorflex.create_matrix(4,4,[[123,431,23,1],[1,2,3,4],[5,6,7,8],[768,564,44,5]]) @@ -318,8 +319,8 @@ inspection and manipulation functions-- `matrix_pos/3`,`size_of_matrix/1`, Used for accessing an element of a Tensorflex matrix. Takes in three input arguments: a Tensorflex `%Matrix` struct matrix, and the -row (`row`) and column (`col`) values of the required element in the matrix. -Both `row` and `col` here are __NOT__ zero indexed. + row (`row`) and column (`col`) values of the required element in the matrix. + Both `row` and `col` here are __NOT__ zero indexed. Returns the value as float. @@ -351,8 +352,8 @@ Both `row` and `col` here are __NOT__ zero indexed. Takes a Tensorflex `%Matrix` struct matrix as input. -Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of -the matrix and `ncols` represents the number of columns of the matrix. + Returns a tuple `{nrows, ncols}` where `nrows` represents the number of rows of + the matrix and `ncols` represents the number of columns of the matrix. ## Examples @@ -377,8 +378,8 @@ the matrix and `ncols` represents the number of columns of the matrix. Appends a single row to the back of a Tensorflex matrix. Takes a Tensorflex `%Matrix` matrix as input and a single row of data (with -the same number of columns as the original matrix) as a list of lists -(`datalist`) to append to the original matrix. + the same number of columns as the original matrix) as a list of lists + (`datalist`) to append to the original matrix. Returns the extended and modified `%Matrix` struct matrix. @@ -440,9 +441,9 @@ the same number of columns as the original matrix) as a list of lists Returns a list of lists representing the data stored in the matrix. __NOTE__: If the matrix contains very high dimensional data, typically -obtained from a function like `load_csv_as_matrix/2`, then it is not -recommended to convert the matrix back to a list of lists format due to a -possibility of memory errors. + obtained from a function like `load_csv_as_matrix/2`, then it is not + recommended to convert the matrix back to a list of lists format due to a + possibility of memory errors. ## Examples @@ -458,14 +459,14 @@ possibility of memory errors. @doc """ Creates a `TF_DOUBLE` tensor from Tensorflex matrices containing the values -and dimensions specified. + and dimensions specified. Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the -tensor should have and another `%Matrix` matrix (`matrix2`) containing the -dimensions of the required tensor. + tensor should have and another `%Matrix` matrix (`matrix2`) containing the + dimensions of the required tensor. Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding tensor data and type. + Tensorflex struct type that is used for holding tensor data and type. ## Examples: @@ -501,12 +502,12 @@ Tensorflex struct type that is used for holding tensor data and type. @doc """ Creates a `TF_DOUBLE` constant value one-dimensional tensor from the floating -point value specified. + point value specified. Takes in a float value as input. Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding tensor data and type. + Tensorflex struct type that is used for holding tensor data and type. ## Examples @@ -538,14 +539,14 @@ Tensorflex struct type that is used for holding tensor data and type. @doc """ Creates a `TF_FLOAT` tensor from Tensorflex matrices containing the values -and dimensions specified. + and dimensions specified. Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the -tensor should have and another `%Matrix` matrix (`matrix2`) containing the -dimensions of the required tensor. + tensor should have and another `%Matrix` matrix (`matrix2`) containing the + dimensions of the required tensor. -Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal + Tensorflex struct type that is used for holding tensor data and type. ## Examples: @@ -581,12 +582,12 @@ Tensorflex struct type that is used for holding tensor data and type. @doc """ Creates a `TF_FLOAT` constant value one-dimensional tensor from the floating -point value specified. + point value specified. Takes in a float value as input. Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding tensor data and type. + Tensorflex struct type that is used for holding tensor data and type. ## Examples @@ -618,18 +619,18 @@ Tensorflex struct type that is used for holding tensor data and type. @doc """ Creates a `TF_INT32` tensor from Tensorflex matrices containing the values -and dimensions specified. + and dimensions specified. Takes two arguments: a `%Matrix` matrix (`matrix1`) containing the values the -tensor should have and another `%Matrix` matrix (`matrix2`) containing the -dimensions of the required tensor. + tensor should have and another `%Matrix` matrix (`matrix2`) containing the + dimensions of the required tensor. -Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding tensor data and type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal + Tensorflex struct type that is used for holding tensor data and type. __NOTE__: In case floating point values are passed in the values matrix -(`matrix1`) as arguments for this function, the tensor will still be created -and all the float values will be typecast to integers. + (`matrix1`) as arguments for this function, the tensor will still be created + and all the float values will be typecast to integers. ## Examples: @@ -665,12 +666,12 @@ and all the float values will be typecast to integers. @doc """ Creates a `TF_INT32` constant value one-dimensional tensor from the integer -value specified. + value specified. Takes in an integer value as input. Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding tensor data and type. + Tensorflex struct type that is used for holding tensor data and type. ## Examples @@ -702,12 +703,12 @@ Tensorflex struct type that is used for holding tensor data and type. @doc """ Creates a `TF_STRING` constant value string tensor from the string value -specified. + specified. Takes in a string value as input. Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding tensor data and type. + Tensorflex struct type that is used for holding tensor data and type. ## Examples @@ -741,21 +742,21 @@ Tensorflex struct type that is used for holding tensor data and type. Allocates a `TF_INT32` tensor of specified dimensions. This function is generally used to allocate output tensors that do not hold -any value data yet, but _will_ after the session is run for Inference. Output -tensors of the required dimensions are allocated and then passed to the -`run_session/5` function to hold the output values generated as predictions. + any value data yet, but _will_ after the session is run for Inference. Output + tensors of the required dimensions are allocated and then passed to the + `run_session/5` function to hold the output values generated as predictions. Takes a Tensorflex `%Matrix` struct matrix as input. Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding the potential tensor data and -type. + Tensorflex struct type that is used for holding the potential tensor data and + type. ## Examples As an example, we can allocate an `int32` output tensor that will be a vector -of 250 values (`1x250` matrix). Therefore, after the session is run, the output -will be an `integer` vector containing 250 values: + of 250 values (`1x250` matrix). Therefore, after the session is run, the output + will be an `integer` vector containing 250 values: ```elixir iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.int32_tensor_alloc @@ -777,21 +778,21 @@ will be an `integer` vector containing 250 values: Allocates a `TF_FLOAT` tensor of specified dimensions. This function is generally used to allocate output tensors that do not hold -any value data yet, but _will_ after the session is run for Inference. Output -tensors of the required dimensions are allocated and then passed to the -`run_session/5` function to hold the output values generated as predictions. + any value data yet, but _will_ after the session is run for Inference. Output + tensors of the required dimensions are allocated and then passed to the + `run_session/5` function to hold the output values generated as predictions. Takes a Tensorflex `%Matrix` struct matrix as input. Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding the potential tensor data and -type. + Tensorflex struct type that is used for holding the potential tensor data and + type. ## Examples As an example, we can allocate a `float32` output tensor that will be a -vector of 250 values (`1x250` matrix). Therefore, after the session is run, the -output will be a `float` vector containing 250 values: + vector of 250 values (`1x250` matrix). Therefore, after the session is run, the + output will be a `float` vector containing 250 values: ```elixir iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float32_tensor_alloc @@ -812,22 +813,22 @@ output will be a `float` vector containing 250 values: @doc """ Allocates a `TF_DOUBLE` tensor of specified dimensions. - This function is generally used to allocate output tensors that do not hold -any value data yet, but _will_ after the session is run for Inference. Output -tensors of the required dimensions are allocated and then passed to the -`run_session/5` function to hold the output values generated as predictions. + This function is generally used to allocate output tensors that do not hold + any value data yet, but _will_ after the session is run for Inference. Output + tensors of the required dimensions are allocated and then passed to the + `run_session/5` function to hold the output values generated as predictions. - Takes a Tensorflex `%Matrix` struct matrix as input. + Takes a Tensorflex `%Matrix` struct matrix as input. - Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding the potential tensor data and -type. + Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal + Tensorflex struct type that is used for holding the potential tensor data and + type. ## Examples As an example, we can allocate a `float64` output tensor that will be a -vector of 250 values (`1x250` matrix). Therefore, after the session is run, the -output will be a `double` vector containing 250 values: + vector of 250 values (`1x250` matrix). Therefore, after the session is run, the + output will be a `double` vector containing 250 values: ```elixir iex(1)> {:ok, tensor} = Tensorflex.create_matrix(1,2,[[1,250]]) |> Tensorflex.float64_tensor_alloc @@ -851,9 +852,9 @@ output will be a `double` vector containing 250 values: Takes in a `%Tensor` struct tensor as input. Returns a tuple `{:ok, datatype}` where `datatype` is an atom representing -the list of Tensorflow `TF_DataType` tensor datatypes. Click -[here](https://github.com/anshuman23/tensorflex/blob/master/c_src/c_api.h#L98-L122) -to view a list of all possible datatypes. + the list of Tensorflow `TF_DataType` tensor datatypes. Click + [here](https://github.com/anshuman23/tensorflex/blob/master/c_src/c_api.h#L98-L122) + to view a list of all possible datatypes. ## Examples @@ -876,37 +877,37 @@ to view a list of all possible datatypes. @doc """ Loads `JPEG` images into Tensorflex directly as a `TF_UINT8` tensor of -dimensions `image height x image width x number of color channels`. + dimensions `image height x image width x number of color channels`. This function is very useful if you wish to do image classification using -Convolutional Neural Networks, or other Deep Learning Models. One of the most -widely adopted and robust image classification models is the -[Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) -model by Google. It makes classifications on images from over a 1000 classes -with highly accurate results. The `load_image_as_tensor/1` function is an -essential component for the prediction pipeline of the Inception model (and for -other similar image classification models) to work in Tensorflex. + Convolutional Neural Networks, or other Deep Learning Models. One of the most + widely adopted and robust image classification models is the + [Inception](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz) + model by Google. It makes classifications on images from over a 1000 classes + with highly accurate results. The `load_image_as_tensor/1` function is an + essential component for the prediction pipeline of the Inception model (and for + other similar image classification models) to work in Tensorflex. Reads in the path to a `JPEG` image file (`.jpg` or `.jpeg`). Returns a tuple `{:ok, %Tensor}` where `%Tensor` represents an internal -Tensorflex struct type that is used for holding the tensor data and type. Here -the created Tensor is a `uint8` tensor (`TF_UINT8`). + Tensorflex struct type that is used for holding the tensor data and type. Here + the created Tensor is a `uint8` tensor (`TF_UINT8`). __NOTE__: For now, only 3 channel RGB `JPEG` color images can be passed as -arguments. Support for grayscale images and other image formats such as `PNG` -will be added in the future. + arguments. Support for grayscale images and other image formats such as `PNG` + will be added in the future. ## Examples To exemplify the working of the `load_image_as_tensor/1` function we will -cover the entire prediction pipeline for the Inception model. However, this -makes use of many other Tensorflex functions such as `run_session/5` and the -other tensor functions so it would be advisable to go through them first. Also, -the Inception model can be downloaded -[here](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz). -We will be making use of the `cropped_panda.jpg` image file that comes along -with the model to test out the model in Tensorflex. + cover the entire prediction pipeline for the Inception model. However, this + makes use of many other Tensorflex functions such as `run_session/5` and the + other tensor functions so it would be advisable to go through them first. Also, + the Inception model can be downloaded + [here](http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz). + We will be making use of the `cropped_panda.jpg` image file that comes along + with the model to test out the model in Tensorflex. First the graph is loaded: @@ -930,8 +931,8 @@ with the model to test out the model in Tensorflex. }} ``` Then we create the output tensor which will hold out output vector values. -For the Inception model, the output is received as a `1008x1 float32` tensor, -as there are 1008 classes in the model: + For the Inception model, the output is received as a `1008x1 float32` tensor, + as there are 1008 classes in the model: ```elixir iex(3)> {:ok, output_tensor} = Tensorflex.create_matrix(1,2,[[1008,1]]) |> Tensorflex.float32_tensor_alloc @@ -968,10 +969,10 @@ as there are 1008 classes in the model: ``` Finally, we need to find which class has the maximum probability and identify -it's label. Since `results` is a List of Lists, it's better to read in the -flattened list. Then we need to find the index of the element in the new list -which as the maximum value. Therefore: -```elixir + it's label. Since `results` is a List of Lists, it's better to read in the + flattened list. Then we need to find the index of the element in the new list + which as the maximum value. Therefore: + ```elixir iex(5)> max_prob = List.flatten(results) |> Enum.max 0.8849328756332397 @@ -979,18 +980,18 @@ which as the maximum value. Therefore: 169 ``` We can thus see that the class with the maximum probability predicted -(`0.8849328756332397`) for the image is `169`. We will now find what the `169` -label corresponds to. For this we can look back into the unzipped Inception -folder, where there is a file called -`imagenet_2012_challenge_label_map_proto.pbtxt`. On opening this file, we can -find the string class identifier for the `169` class index. This is `n02510455` -and is present on Line 1556 in the file. Finally, we need to match this string -identifier to a set of identification labels by referring to the file -`imagenet_synset_to_human_label_map.txt` file. Here we can see that -corresponding to the string class `n02510455` the human labels are `giant -panda, panda, panda bear, coon bear, Ailuropoda melanoleuca` (Line 3691 in the -file). Thus, we have correctly identified the animal in the image as a panda -using Tensorflex. + (`0.8849328756332397`) for the image is `169`. We will now find what the `169` + label corresponds to. For this we can look back into the unzipped Inception + folder, where there is a file called + `imagenet_2012_challenge_label_map_proto.pbtxt`. On opening this file, we can + find the string class identifier for the `169` class index. This is `n02510455` + and is present on Line 1556 in the file. Finally, we need to match this string + identifier to a set of identification labels by referring to the file + `imagenet_synset_to_human_label_map.txt` file. Here we can see that + corresponding to the string class `n02510455` the human labels are `giant + panda, panda, panda bear, coon bear, Ailuropoda melanoleuca` (Line 3691 in the + file). Thus, we have correctly identified the animal in the image as a panda + using Tensorflex. """ def load_image_as_tensor(imagepath) do @@ -1008,33 +1009,33 @@ using Tensorflex. @doc """ Loads high-dimensional data from a `CSV` file as a Tensorflex 2-D matrix in a -super-fast manner. + super-fast manner. The `load_csv_as_matrix/2` function is very fast-- when compared with the -Python based `pandas` library for data science and analysis' function -`read_csv` on the `test.csv` file from MNIST Kaggle data -([source](https://www.kaggle.com/c/digit-recognizer/data)), the following -execution times were obtained: - - `read_csv`: `2.549233` seconds - - `load_csv_as_matrix/2`: `1.711494` seconds + Python based `pandas` library for data science and analysis' function + `read_csv` on the `test.csv` file from MNIST Kaggle data + ([source](https://www.kaggle.com/c/digit-recognizer/data)), the following + execution times were obtained: + - `read_csv`: `2.549233` seconds + - `load_csv_as_matrix/2`: `1.711494` seconds This function takes in 2 arguments: a path to a valid CSV file (`filepath`) -and other optional arguments `opts`. These include whether or not a header -needs to be discarded in the CSV, and what the delimiter type is. These are -specified by passing in an atom `:true` or `:false` to the `header:` key, and -setting a string value for the `delimiter:` key. By default, the header is -considered to be present (`:true`) and the delimiter is set to `,`. + and other optional arguments `opts`. These include whether or not a header + needs to be discarded in the CSV, and what the delimiter type is. These are + specified by passing in an atom `:true` or `:false` to the `header:` key, and + setting a string value for the `delimiter:` key. By default, the header is + considered to be present (`:true`) and the delimiter is set to `,`. Returns a `%Matrix` Tensorflex struct type. ## Examples: We first exemplify the working with the `test.csv` file which belongs to the -MNIST Kaggle CSV data -([source](https://www.kaggle.com/c/digit-recognizer/data)), which contains -`28000` rows and `784` columns (without the header). It is comma delimited and -also contains a header. From the `test.csv` file, we also create a custom file -withou the header present which we refer to as `test_without_header.csv` in the -examples below: + MNIST Kaggle CSV data + ([source](https://www.kaggle.com/c/digit-recognizer/data)), which contains + `28000` rows and `784` columns (without the header). It is comma delimited and + also contains a header. From the `test.csv` file, we also create a custom file + without the header present which we refer to as `test_without_header.csv` in the + examples below: ```elixir iex(1)> mat = Tensorflex.load_csv_as_matrix("test.csv") @@ -1052,9 +1053,9 @@ examples below: ``` On a visual inspection of the very large `test.csv` file, one can see that -the values in these particular positions are correct. Now we show usage for the -same file but without header, `test_without_header.csv`: -```elixir + the values in these particular positions are correct. Now we show usage for the + same file but without header, `test_without_header.csv`: + ```elixir iex(1)> no_header = Tensorflex.load_csv_as_matrix("test/test_without_header.csv", header: :false) %Tensorflex.Matrix{ data: #Reference<0.4024686574.590479364.257078>, @@ -1070,7 +1071,7 @@ same file but without header, `test_without_header.csv`: ``` Next we see the delimiter functionalities. First, assuming we have two simple -`CSV` files, `sample1.csv` and `sample2.csv` + `CSV` files, `sample1.csv` and `sample2.csv` _sample1.csv_: @@ -1148,39 +1149,39 @@ same file but without header, `test_without_header.csv`: @doc """ Runs a Tensorflow session to generate predictions for a given graph, input -data, and required input/output operations. + data, and required input/output operations. This function is the final step of the Inference (prediction) pipeline and -generates output for a given set of input data, a pre-trained graph model, and -the specified input and output operations of the graph. + generates output for a given set of input data, a pre-trained graph model, and + the specified input and output operations of the graph. Takes in five arguments: a pre-trained Tensorflow graph `.pb` model read in -from the `read_graph/1` function (`graph`), an input tensor with the dimensions -and data required for the input operation of the graph to run (`tensor1`), an -output tensor allocated with the right dimensions (`tensor2`), the name of the -input operation of the graph that needs where the input data is fed -(`input_opname`), and the output operation name in the graph where the outputs -are obtained (`output_opname`). The input tensor is generally created from the -matrices manually or using the `load_csv_as_matrix/2` function, and then passed -through to one of the tensor creation functions. For image classification the -`load_image_as_tensor/1` can also be used to create the input tensor from an -image. The output tensor is created using the tensor allocation functions -(generally containing `alloc` at the end of the function name). + from the `read_graph/1` function (`graph`), an input tensor with the dimensions + and data required for the input operation of the graph to run (`tensor1`), an + output tensor allocated with the right dimensions (`tensor2`), the name of the + input operation of the graph that needs where the input data is fed + (`input_opname`), and the output operation name in the graph where the outputs + are obtained (`output_opname`). The input tensor is generally created from the + matrices manually or using the `load_csv_as_matrix/2` function, and then passed + through to one of the tensor creation functions. For image classification the + `load_image_as_tensor/1` can also be used to create the input tensor from an + image. The output tensor is created using the tensor allocation functions + (generally containing `alloc` at the end of the function name). Returns a List of Lists (similar to the `matrix_to_lists/1` function) -containing the generated predictions as per the output tensor dimensions. + containing the generated predictions as per the output tensor dimensions. -## Examples - - - A blog post [here](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/) covers + These examples can be observed for understanding the prediction pipeline: + + * A blog post [here](http://www.anshumanc.ml/gsoc/2018/06/14/gsoc/) covers generating predictions and running sessions using an MLP model on the Iris -Dataset + Dataset - - Generating predictions from the Inception model by Google is covered in the - `load_image_as_tensor/1` function examples. + * Generating predictions from the Inception model by Google is covered in the + `load_image_as_tensor/1` function examples. - - Working with an RNN-LSTM example for sentiment analysis is covered - [here](https://github.com/anshuman23/tensorflex/pull/25). + * Working with an RNN-LSTM example for sentiment analysis is covered + [here](https://github.com/anshuman23/tensorflex/pull/25). """ def run_session(%Graph{def: graphdef, name: filepath}, %Tensor{datatype: input_datatype, tensor: input_ref}, %Tensor{datatype: output_datatype, tensor: output_ref}, input_opname, output_opname) do