Enable xarray v2024.10.0+

[gselzer]

For some reason, xarray v2024.10.0 and above breaks CI on Ubuntu, with Python 3.12. At the end of this issue is a snippet of the CI failure, saved because the logs will delete later on. I placed an exclusive upper bound on this version in #303, even though I cannot reproduce the error on an available WSL instance.

This is the commit in xarray that made the breaking change.

@elevans @ctrueden thoughts on how to proceed here? Does this failure affect PyImageJ in any way?

-------------------------------------
| Testing ImageJ2 + original ImageJ |
-------------------------------------
============================= test session starts ==============================
platform linux -- Python 3.12.7, pytest-8.3.3, pluggy-1.5.0
PyQt5 5.15.11 -- Qt runtime 5.15.15 -- Qt compiled 5.15.14
rootdir: /home/runner/work/napari-imagej/napari-imagej
configfile: pyproject.toml
plugins: napari-plugin-engine-0.2.0, qt-4.4.0, cov-6.0.0, npe2-0.7.7, napari-0.5.4, env-1.1.5
collected 167 items

tests/test_java.py Oct 29, 2024 9:56:42 PM java.util.prefs.FileSystemPreferences$1 run
INFO: Created user preferences directory.

======= BEGIN JAVA VERSIONS =======
io.scif:scifio 0.45.0
net.imagej:ij 1.54f (no minimum)
net.imagej:imagej 2.15.0 (no minimum)
net.imagej:imagej-common 2.0.4
net.imagej:imagej-legacy 1.2.1
net.imagej:imagej-ops 2.0.0
net.imglib2:imglib2-imglyb 1.1.0
net.imglib2:imglib2-unsafe 1.0.0
org.scijava:scijava-common 2.97.0
org.scijava:scijava-search 2.0.4
sc.fiji:fiji NOT PRESENT
======== END JAVA VERSIONS ========
...
tests/test_scripting.py .
tests/test_settings.py ......
tests/types/test_converters.py .....................................
tests/types/test_enums.py .
tests/types/test_trackmate.py s
tests/types/test_type_conversions.py ......
tests/types/test_widget_mappings.py ................
tests/utilities/test_module_utils.py ..................................
tests/utilities/test_progress.py 
Frangi Vesselness: 100%|██████████| 3/3 [00:00<00:00, 23921.89it/s]
....
tests/widgets/test_info_bar.py .
tests/widgets/test_menu.py ..s.ss...F...
tests/widgets/test_napari_imagej.py .............
tests/widgets/test_parameter_widgets.py ...................
tests/widgets/test_result_runner.py ...
tests/widgets/test_result_tree.py .......
tests/widgets/test_searchbar.py ..

=================================== FAILURES ===================================
_________________________ test_modification_in_imagej __________________________

asserter = <function asserter.<locals>.assertFunc at 0x7f9d28dcdf80>
qtbot = <pytestqt.qtbot.QtBot object at 0x7f9d40f0c920>
ij = <java object 'net.imagej.ImageJ'>
gui_widget = <napari_imagej.widgets.menu.NapariImageJMenu object at 0x7f9dc8927140>

    def test_modification_in_imagej(asserter, qtbot, ij, gui_widget: NapariImageJMenu):
        if settings.headless():
            pytest.skip("Only applies when not running headlessly")
        if not settings.include_imagej_legacy:
            pytest.skip("Tests legacy behavior")
    
        to_button: ToIJButton = gui_widget.to_ij
        from_button: FromIJButton = gui_widget.from_ij
    
        # Show the button
        qtbot.mouseClick(gui_widget.gui_button, Qt.LeftButton, delay=1)
    
        # Add some data to the viewer
        sample_data = numpy.ones((100, 100, 3), dtype=numpy.uint8)
        image: Image = Image(data=sample_data, name="test_to")
        current_viewer().add_layer(image)
    
        # Press the button, handle the Dialog
        qtbot.mouseClick(to_button, Qt.LeftButton, delay=1)
    
        # Assert that the data is in the legacy UI
        asserter(lambda: ij.WindowManager.getCurrentImage() is not None)
        imp = ij.WindowManager.getCurrentImage()
        assert imp.getTitle() == "test_to"
        # Edit the data
        imp.getProcessor().invert()
        imp.updateAndDraw()
    
        # Press the button, handle the Dialog
        qtbot.mouseClick(from_button, Qt.LeftButton, delay=1)
    
        # Assert the returned data is inverted
        asserter(lambda: "test_to [1]" in current_viewer().layers)
        modified_layer = current_viewer().layers["test_to [1]"].data
>       assert numpy.all(modified_layer[0, :, :] == 254)

tests/widgets/test_menu.py:470: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
/opt/hostedtoolcache/Python/3.12.7/x64/lib/python3.12/site-packages/xarray/core/_typed_ops.py:457: in __eq__
    return self._binary_op(other, nputils.array_eq)
/opt/hostedtoolcache/Python/3.12.7/x64/lib/python3.12/site-packages/xarray/core/dataarray.py:4786: in _binary_op
    name = result_name([self, other])
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

objects = [<xarray.DataArray array([[[254, 254, 254],
        [254, 254, 254],
        [254, 254, 254],
        ...,
        [25...   (row) float64 800B 0.0 1.0 2.0 3.0 4.0 ... 95.0 96.0 97.0 98.0 99.0
  * col      (col) float64 24B 0.0 1.0 2.0, 254]

    def result_name(objects: Iterable[Any]) -> Any:
        # use the same naming heuristics as pandas:
        # https://github.com/blaze/blaze/issues/458#issuecomment-51936356
>       names = {getattr(obj, "name", _DEFAULT_NAME) for obj in objects}
E       TypeError: unhashable type: 'numpy.ndarray'

/opt/hostedtoolcache/Python/3.12.7/x64/lib/python3.12/site-packages/xarray/core/utils.py:1206: TypeError
=============================== warnings summary ===============================
../../../../../opt/hostedtoolcache/Python/3.12.7/x64/lib/python3.12/copy.py:151
  /opt/hostedtoolcache/Python/3.12.7/x64/lib/python3.12/copy.py:151: DeprecationWarning: Pickle, copy, and deepcopy support will be removed from itertools in Python 3.14.
    rv = reductor(4)

../../../../../opt/hostedtoolcache/Python/3.12.7/x64/lib/python3.12/copy.py:261
  /opt/hostedtoolcache/Python/3.12.7/x64/lib/python3.12/copy.py:261: DeprecationWarning: Pickle, copy, and deepcopy support will be removed from itertools in Python 3.14.
    y.__setstate__(state)

src/napari_imagej/types/converters/trackmate.py:26
  /home/runner/work/napari-imagej/napari-imagej/src/napari_imagej/types/converters/trackmate.py:26: SyntaxWarning: invalid escape sequence '\d'
    if match("TrackMate-\d.*\.jar", fname):  # noqa

-- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html
=========================== short test summary info ============================
FAILED tests/widgets/test_menu.py::test_modification_in_imagej - TypeError: unhashable type: 'numpy.ndarray'
======= 1 failed, 162 passed, 4 skipped, 3 warnings in 64.71s (0:01:04) ========

[elevans]

Can you reproduce this bug locally or just in CI? I'll see if I can replicate this. I had to make this change to pyimagej due to xarray API deprecation. I'm wondering if something similar is happening here.

[gselzer]

Nope, also happens locally (Windows 10, Powershell). Reproducible using:

1. On the main branch of napari-imagej, run mamba env create -f dev-envrionment.yml
2. mamba install -c conda-forge xarray==2024.11.0
3. mamba run pytest

I receive an error equivalent to the error on CI.