Bump to v1.1.4 (#563)

* Set Twine version to fix pypi uploads on build

- twine v3.4.x and greater seems to have issues with our build system since we have an older version of `importlib_metadata`

* Revert changes from merge into main

* Bump version to 1.1.3

* Update docs

* fix code samples in documentation (#535)

The code samples to render properly have to be valid rst literal blocks (https://www.sphinx-doc.org/en/master/usage/restructuredtext/basics.html#literal-blocks). The most affected are multi-line code samples that are joined to a singe line, see https://sleap.ai/api/sleap.io.dataset.html.

* Minor typo fixed (#538)

* Add CLI options for specifying the GPU when training (#547)

* Add GPU specification to training CLI
* Update CLI docs
* Set linter version

* Minor QOL update (#562)

* Add score metadata to the exported HDF5 analysis files

* Fix and add test for labels.numpy()

* Add some tweaks and docs to video rendering

* Add model describer to sleap-inspect CLI

* Minor logging

* CLI tweak

* Tweak visual indicator of not-visible points

* Color untracked instances consistently

* Track removal utilities
- Add menu option: Tracks -> Delete All Tracks
- Add Labels.remove_all_tracks() method
- Tests

* Try adding token to codecov action

* Pin black formatter version and relint

* Bump minor version

* Update minor version in docs

* Disable verbose coverage uploading

* Update codecov badge branch

Co-authored-by: Matěj Šmíd <[email protected]>
Co-authored-by: Darío Hereñú <[email protected]>

Author: 3 people

.github/workflows/ci.yml

@@ -36,7 +36,7 @@ jobs:
         python-version: 3.6
     - name: Install Dependencies
       run: |
-        pip install black
+        pip install black==20.8b1
     - name: Run Black
       run: |
         black --check sleap tests
@@ -90,4 +90,8 @@ jobs:
           xvfb-run pytest --cov=sleap --cov-report=xml tests/
       - name: Upload coverage
         uses: codecov/codecov-action@v1
-        if: matrix.os == 'ubuntu-18.04'
+        if: matrix.os == 'ubuntu-18.04'
+        with:
+          token: ${{ secrets.CODECOV_TOKEN }}
+          fail_ci_if_error: true
+          verbose: false

README.rst

@@ -6,7 +6,7 @@
    :alt: Continuous integration status
 
 .. |Coverage| image::
-   https://codecov.io/gh/murthylab/sleap/branch/tf23/graph/badge.svg?token=YWQYBN6820
+   https://codecov.io/gh/murthylab/sleap/branch/develop/graph/badge.svg?token=YWQYBN6820
    :target: https://codecov.io/gh/murthylab/sleap
    :alt: Coverage
 
@@ -99,7 +99,7 @@ Contributors
 * **Joshua Shaevitz**, Physics and Lewis-Sigler Institute, Princeton University
 * **Mala Murthy**, Princeton Neuroscience Institute, Princeton University
 
-SLEAP is developed in the `Murthy <https://murthylab.princeton.edu>`_ and `Shaevitz <https://shaevitzlab.princeton.edu>`_ labs at the `Princeton Neuroscience Institute <https://pni.princeton.edu>`_ at Princeton University. This work was made possible through our funding sources, including: NIH BRAIN Initative R01 NS104899 and Princeton Innovation Accelerator Fund.
+SLEAP is developed in the `Murthy <https://murthylab.princeton.edu>`_ and `Shaevitz <https://shaevitzlab.princeton.edu>`_ labs at the `Princeton Neuroscience Institute <https://pni.princeton.edu>`_ at Princeton University. This work was made possible through our funding sources, including: NIH BRAIN Initiative R01 NS104899 and Princeton Innovation Accelerator Fund.
 
 .. end-inclusion-marker-do-not-remove
 

dev_requirements.txt

@@ -7,7 +7,7 @@ furo
 sphinx-copybutton
 myst-nb
 sphinx-autobuild
-black
+black==21.6b0
 pre-commit
 twine==3.3.0
 PyGithub

docs/conf.py

@@ -27,15 +27,15 @@
 copyright = "2019–2021, Murthy Lab @ Princeton University"
 
 # The short X.Y version
-version = "1.1.3"
+version = "1.1.4"
 
 # Get the sleap version
 # with open("../sleap/version.py") as f:
 #     version_file = f.read()
 #     version = re.search("\d.+(?=['\"])", version_file).group(0)
 
 # Release should be the full branch name
-release = "v1.1.3"
+release = "v1.1.4"
 
 html_title = f"SLEAP ({release})"
 html_short_title = "SLEAP"

docs/guides/cli.rst

@@ -66,6 +66,11 @@ Training
                             name settings.
       --prefix PREFIX       Prefix to prepend to run name.
       --suffix SUFFIX       Suffix to append to run name.
+      --cpu                 Run training only on CPU. If not specified, will use
+                            available GPU.
+      --first-gpu           Run training on the first GPU, if available.
+      --last-gpu            Run training on the last GPU, if available.
+      --gpu GPU             Run training on the i-th GPU on the system.
 
 
 Inference and Tracking
@@ -232,13 +237,14 @@ Dataset files
 
 :code:`sleap-inspect` gives you various information about a SLEAP dataset file such as a list of videos and a count of the frames with labels. If you're inspecting a predictions dataset (i.e., the output from running :code:`sleap-track` or inference in the GUI) it will also include details about how those predictions were created (i.e., the models, the version of SLEAP, and any inference parameters).
 
+You can also specify a model folder to get a quick summary of the configuration and metrics (if available).
 
 .. code-block:: none
 
     usage: sleap-inspect [-h] [--verbose] data_path
 
     positional arguments:
-      data_path   Path to labels json file
+      data_path   Path to labels file (.slp) or model folder
 
     optional arguments:
       -h, --help  show this help message and exit

docs/installation.rst

@@ -32,7 +32,7 @@ different projects, mitigating issues of managing complex dependencies. To creat
 
 ::
 
-    (base) C:\>  conda create -n sleap -c defaults -c sleap sleap=1.1.3 python=3.6 -y
+    (base) C:\>  conda create -n sleap -c defaults -c sleap sleap=1.1.4 python=3.6 -y
 
 Once the environment is finished installing, it can be activated using the following command:
 
@@ -76,7 +76,7 @@ You can then install SLEAP by running:
 
 ::
 
-    pip install sleap==1.1.3
+    pip install sleap==1.1.4
 
 **SLEAP** is now installed, you can run the labeling GUI:
 

docs/notebooks/Training_and_inference_on_an_example_dataset.ipynb

@@ -51,7 +51,7 @@
    },
    "outputs": [],
    "source": [
-    "!pip install sleap==1.1.3"
+    "!pip install sleap==1.1.4"
    ]
   },
   {

docs/notebooks/Training_and_inference_using_Google_Drive.ipynb

@@ -48,7 +48,7 @@
    },
    "outputs": [],
    "source": [
-    "!pip install sleap==1.1.3"
+    "!pip install sleap==1.1.4"
    ]
   },
   {

sleap/gui/app.py

@@ -635,8 +635,6 @@ def new_instance_menu_action():
 
         tracksMenu = self.menuBar().addMenu("Tracks")
         self.track_menu = tracksMenu.addMenu("Set Instance Track")
-        self.delete_tracks_menu = tracksMenu.addMenu("Delete Track")
-        self.delete_tracks_menu.setEnabled(False)
         add_menu_check_item(
             tracksMenu, "propagate track labels", "Propagate Track Labels"
         ).setToolTip(
@@ -649,12 +647,25 @@ def new_instance_menu_action():
             "Transpose Instance Tracks",
             self.commands.transposeInstance,
         )
+
+        tracksMenu.addSeparator()
+
         add_menu_item(
             tracksMenu,
             "delete track",
             "Delete Instance and Track",
             self.commands.deleteSelectedInstanceTrack,
         )
+        self.delete_tracks_menu = tracksMenu.addMenu("Delete Track")
+        self.delete_tracks_menu.setEnabled(False)
+        add_menu_item(
+            tracksMenu,
+            "delete all tracks",
+            "Delete All Tracks",
+            self.commands.deleteAllTracks,
+        ).setToolTip(
+            "Delete all tracks and update instances. Instances are not removed."
+        )
 
         tracksMenu.addSeparator()
 

sleap/gui/color.py

@@ -151,11 +151,11 @@ def get_pseudo_track_index(self, instance: "Instance") -> Union[Track, int]:
         if not instance.frame:
             return 0
 
-        non_track_instances = [
-            inst for inst in instance.frame.instances_to_show if inst.track is None
+        untracked_instances = [
+            inst for inst in instance.frame.user_instances if inst.track is None
         ]
 
-        return len(self.tracks) + non_track_instances.index(instance)
+        return len(self.tracks) + untracked_instances.index(instance)
 
     def get_track_color(self, track: Union[Track, int]) -> ColorTupleType:
         """Returns the color to use for a given track.

sleap/gui/commands.py

@@ -505,6 +505,10 @@ def deleteTrack(self, track: "Track"):
         """Delete a track and remove from all instances."""
         self.execute(DeleteTrack, track=track)
 
+    def deleteAllTracks(self):
+        """Delete all tracks."""
+        self.execute(DeleteAllTracks)
+
     def setTrackName(self, track: "Track", name: str):
         """Sets name for track."""
         self.execute(SetTrackName, track=track, name=name)
@@ -1939,6 +1943,14 @@ def do_action(context: CommandContext, params: dict):
         context.labels.remove_track(track)
 
 
+class DeleteAllTracks(EditCommand):
+    topics = [UpdateTopic.tracks]
+
+    @staticmethod
+    def do_action(context: CommandContext, params: dict):
+        context.labels.remove_all_tracks()
+
+
 class SetTrackName(EditCommand):
     topics = [UpdateTopic.tracks, UpdateTopic.frame]
 

sleap/gui/dialogs/formbuilder.py

@@ -6,17 +6,17 @@
 a little more convenient (it provides methods for adding a message for the
 dialog and for getting the results when the dialog is closed).
 
-Example of form widget:
+Example of form widget: ::
 
->>> widget = YamlFormWidget(yaml_file="example.yaml")
->>> widget.mainAction.connect(my_function)
+   >>> widget = YamlFormWidget(yaml_file="example.yaml")
+   >>> widget.mainAction.connect(my_function)
 
 my_function will get called with form data when user clicks the main button
 (main button has type "button" and default "main action")
 
-Example of modal dialog:
+Example of modal dialog: ::
 
->>> results = FormBuilderModalDialog(form_name="example").get_results()
+   >>> results = FormBuilderModalDialog(form_name="example").get_results()
 
 The results will be empty dictionary if the user hit "cancel", otherwise it
 will contain all data from form (dict keys matching names of fields).
@@ -43,9 +43,9 @@ class YamlFormWidget(QtWidgets.QGroupBox):
 
     Typically you'll want to save the YAML in `sleap/config/` and use the
     :py:meth:`from_name` method to make the form (e.g., if your form data is in
-    `sleap/config/foo.yaml`, then you can create form like so:
+    `sleap/config/foo.yaml`, then you can create form like so: ::
 
-    >>> widget = YamlFormWidget.from_name("foo")
+       >>> widget = YamlFormWidget.from_name("foo")
 
     Args:
         yaml_file: filename of YAML file to load.

sleap/gui/dialogs/importvideos.py

@@ -1,9 +1,9 @@
 """
 Interface to handle the UI for importing videos.
 
-Usage:
+Usage: ::
 
->>> import_list = ImportVideos().ask()
+   >>> import_list = ImportVideos().ask()
 
 This will show the user a file-selection dialog, and then a second dialog
 to select the import parameters for each file.
@@ -14,9 +14,9 @@
 to the relevant method of :class:`Video`.
 
 For each `item` in `import_list`, we can load the video by calling this
-method while passing the user-selected params as the named parameters:
+method while passing the user-selected params as the named parameters: ::
 
->>> vid = item["video_class"](**item["params"])
+   >>> vid = item["video_class"](**item["params"])
 
 """
 

sleap/gui/overlays/confmaps.py

@@ -5,9 +5,11 @@
 current frame) and then uses a `ConfMapsPlot` object to show the resulting
 confidence maps.
 
-Example:
+Example: ::
+
     >>> cm = ConfMapsPlot(conf_data.get_frame(0))
     >>> window.view.scene.addItem(cm)
+
 """
 
 from PySide2 import QtWidgets, QtCore, QtGui

sleap/gui/widgets/multicheck.py

@@ -1,10 +1,12 @@
 """
 Module for Qt Widget to show multiple checkboxes for selecting.
 
-Example:
+Example: ::
+
     >>> mc = MultiCheckWidget(count=5, selected=[0,1],title="My Items")
     >>> mc.selectionChanged.connect(window.plot)
     >>> window.layout.addWidget(mc)
+
 """
 
 from typing import List, Optional

sleap/gui/widgets/video.py

@@ -3,7 +3,8 @@
 
 All interactions should go through `QtVideoPlayer`.
 
-Example usage:
+Example usage: ::
+
     >>> my_video = Video(...)
     >>> my_instance = Instance(...)
 
@@ -523,8 +524,9 @@ def onSequenceSelect(
         selection), the `on_failure` callback is called (if given).
 
         Note:
-            If successful, we call
-            >>> on_success(list_of_instances)
+            If successful, we call ::
+
+               >>> on_success(list_of_instances)
 
         Args:
             seq_len: Number of instances we want to collect in sequence.
@@ -1218,24 +1220,31 @@ def adjustStyle(self):
         complete_color = (
             QColor(80, 194, 159) if self.node.point.complete else QColor(232, 45, 32)
         )
+        missing_color = QColor(128, 128, 128)
 
         if self.predicted:
             self._base_font.setBold(False)
+            self._base_font.setItalic(False)
             self.setFont(self._base_font)
             self.setDefaultTextColor(QColor(128, 128, 128))
         elif not self.node.point.visible:
-            self._base_font.setBold(False)
+            self._base_font.setBold(True)
+            self._base_font.setItalic(True)
             self.setFont(self._base_font)
-            # self.setDefaultTextColor(self.node.pen().color())
-            self.setDefaultTextColor(complete_color)
+            self.setPlainText(f"{self.node.name}*")
+            self.setDefaultTextColor(missing_color)
         elif self.node.point.complete:
             self._base_font.setBold(True)
+            self._base_font.setItalic(False)
+            self.setPlainText(f"{self.node.name}")
             self.setFont(self._base_font)
             self.setDefaultTextColor(complete_color)  # greenish
             # FIXME: Adjust style of node here as well?
             # self.node.setBrush(complete_color)
         else:
             self._base_font.setBold(False)
+            self._base_font.setItalic(False)
+            self.setPlainText(f"{self.node.name}")
             self.setFont(self._base_font)
             self.setDefaultTextColor(complete_color)  # redish
 
@@ -1352,10 +1361,10 @@ def __init__(
             self.pen_default.setCosmetic(
                 True
             )  # https://stackoverflow.com/questions/13120486/adjusting-qpen-thickness-when-scaling-qgraphicsview
-            self.pen_missing = QPen(line_color, 1)
+            self.pen_missing = QPen(line_color, 1)  # thin border
             self.pen_missing.setCosmetic(True)
             self.brush = QBrush(QColor(*self.color, a=128))
-            self.brush_missing = QBrush(QColor(*self.color, a=0))
+            self.brush_missing = QBrush(QColor(*self.color, a=0))  # no fill
 
         self.setPos(self.point.x, self.point.y)
         self.updatePoint(user_change=False)
@@ -1371,7 +1380,7 @@ def visible_radius(self):
         if self.point.visible:
             return self.radius / self.player.view.zoomFactor
         else:
-            return self.radius / (2.0 * self.player.view.zoomFactor)
+            return self.radius / (2.0 * self.player.view.zoomFactor)  # smaller marker
 
     def updatePoint(self, user_change: bool = False):
         """
@@ -1395,7 +1404,7 @@ def updatePoint(self, user_change: bool = False):
             self.setPen(self.pen_default)
             self.setBrush(self.brush)
         else:
-            radius = self.radius / 2.0
+            radius = self.radius / 2.0  # smaller marker
             self.setPen(self.pen_missing)
             self.setBrush(self.brush_missing)
             if not self.show_non_visible: