Fix typing for coords argument in DataArray

[DeltaSigma130]

Problem

Mypy reports errors on anything other than a list of tuples passed to coords in xr.DataArray constructor even though xr.Variable and xr.DataArray are valid objects.

Code

Contained in file type_test.py.

import numpy as np
import xarray as xr

t = np.linspace(0, 2*np.pi, 12)
tv = xr.Variable('time', t)  # line 5

d2sin = xr.DataArray(np.sin(2 * tv), coords=[('time', t)])  # using tuple
#    mypy column error detect here  ----v
dsin = xr.DataArray(np.sin(tv), coords=[tv])  # using xr.Variable
dcos = xr.DataArray(np.cos(tv), coords=[dsin['time']])  # using xr.DataArray

print(f"Types:\n\t{type(tv)=}\n\t{type(dsin['time'])=}")  # line 12
print("**d2sin**:", d2sin)
print("**dsin**:", dsin)
print("**dcos**:", dcos)

Running Python

(jupyter) PS C:\Users\<USER>\Documents\Projects\Jupyter\sim_optical_disk> python .\type_test.py
Types:
        type(tv)=<class 'xarray.core.variable.Variable'>
        type(dsin['time'])=<class 'xarray.core.dataarray.DataArray'>
**d2sin**: <xarray.DataArray (time: 12)>
array([ 0.00000000e+00,  9.09631995e-01,  7.55749574e-01, -2.81732557e-01,
       -9.89821442e-01, -5.40640817e-01,  5.40640817e-01,  9.89821442e-01,
        2.81732557e-01, -7.55749574e-01, -9.09631995e-01, -4.89858720e-16])
Coordinates:
  * time     (time) float64 0.0 0.5712 1.142 1.714 ... 4.57 5.141 5.712 6.283
**dsin**: <xarray.DataArray (time: 12)>
array([ 0.00000000e+00,  5.40640817e-01,  9.09631995e-01,  9.89821442e-01,
        7.55749574e-01,  2.81732557e-01, -2.81732557e-01, -7.55749574e-01,
       -9.89821442e-01, -9.09631995e-01, -5.40640817e-01, -2.44929360e-16])
Coordinates:
  * time     (time) float64 0.0 0.5712 1.142 1.714 ... 4.57 5.141 5.712 6.283
**dcos**: <xarray.DataArray (time: 12)>
array([ 1.        ,  0.84125353,  0.41541501, -0.14231484, -0.65486073,
       -0.95949297, -0.95949297, -0.65486073, -0.14231484,  0.41541501,
        0.84125353,  1.        ])
Coordinates:
  * time     (time) float64 0.0 0.5712 1.142 1.714 ... 4.57 5.141 5.712 6.283

Running MyPy

(jupyter) PS C:\Users\<USER>\Documents\Projects\Jupyter\sim_optical_disk> mypy .\type_test.py
type_test.py:9: error: List item 0 has incompatible type "Variable"; expected "Tuple[Any, ...]"
type_test.py:10: error: List item 0 has incompatible type "DataArray"; expected "Tuple[Any, ...]"
Found 2 errors in 1 file (checked 1 source file)

Environment

Output of xr.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.9.1 (default, Dec 11 2020, 09:29:25) [MSC v.1916 64 bit (AMD64)]
python-bits: 64
OS: Windows
OS-release: 10
machine: AMD64
processor: Intel64 Family 6 Model 158 Stepping 9, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None
LOCALE: English_United States.1252
libhdf5: 1.10.6
libnetcdf: 4.6.1

xarray: 0.18.0
pandas: 1.2.1
numpy: 1.20.2
scipy: 1.6.2
netCDF4: 1.5.7
pydap: None
h5netcdf: None
h5py: None
Nio: None
zarr: None
cftime: 1.5.0
nc_time_axis: None
PseudoNetCDF: None
rasterio: None
cfgrib: None
iris: None
bottleneck: None
dask: None
distributed: None
matplotlib: 3.3.2
cartopy: None
seaborn: None
numbagg: None
pint: None
setuptools: 52.0.0.post20210125
pip: 21.1.1
conda: None
pytest: None
IPython: 7.19.0
sphinx: None

[max-sixty]

Good spot, thanks @DeltaSigma130 .

I think this should be a fairly easy fix if anyone is up for taking this. I'm somewhat surprised that it's not covered in our existing code such that mypy fails within xarray.

[H0R5E]

I can confirm this issue when specifying dimensions with coordinates, as per this example:

import xarray as xr

z = -1
x = [1, 2, 3]
y = [4, 5, 6]
data = [0, 0, 0]

coords = {"z": z,
          "x": ("dim_0", x),
          "y": ("dim_0", y)}

xr.DataArray(data,
             coords=coords)

MyPy output:

error: Argument "coords" to "DataArray" has incompatible type "Dict[str, object]"; expected "Union[Sequence[Tuple[Any, ...]], Mapping[Hashable, Any], None]"

I can work around it by specifying the type of coords explicitly:

from typing import Any, Hashable, Mapping

import xarray as xr

z = -1
x = [1, 2, 3]
y = [4, 5, 6]
data = [0, 0, 0]

coords: Mapping[Hashable, Any] = {"z": z,
                                  "x": ("dim_0", x),
                                  "y": ("dim_0", y)}

xr.DataArray(data,
             coords=coords)

Success: no issues found in 1 source file

There is some discussion about str and Hashable in this issue, but type invariance is still a little over my head, TBH.

[H0R5E]

My issue might actually be fixed in main.

EDIT: Change was made in #5728