Clip min eigenvalues to 0 to block negative values in sqrt

[zoccoler]

This sometimes can happen with floating point representation of very small numbers.

This solves #88 .

[zoccoler]

Hi @haesleinhuepf ,

This is a simple fix to solve an issue referred in image.sc here.

Please take a look and feel free to merge it if you agree.

Side-note: I had to pin the napari version for tests to make them work. We should either pin it also in the plugin requirements or update the plugin to make it compatible with newer napari versions. Let me know if you have an opinion on this.

[haesleinhuepf]

Hi Marcelo @zoccoler ,

awesome, thanks!

Side-note: I had to pin the napari version for tests to make them work.

What do you think about pinning the version in the tests only? Have you tried the plugin with napari 0.5.0 ? If it still works there, but only the tests are failing, we could implement the version pin here instead by adding "napari==0.4.19":

napari-skimage-regionprops/.github/workflows/test_and_deploy.yml

Line 50 in e38cc72

python -m pip install setuptools tox tox-gh-actions

Thanks for working on this!

Best,
Robert

[zoccoler]

What do you think about pinning the version in the tests only?

I thought I did exactly this by pinning it here and not here. Is that correct?

Have you tried the plugin with napari 0.5.0 ?

I tried the plugin an environment with napari>0.5 (0.5.4), and there the following test fails:

================================================================ FAILURES ================================================================
___________________________________________ test_timelapse_analyse_all_timepoints_with_viewer ____________________________________________

make_napari_viewer = <function make_napari_viewer.<locals>.actual_factory at 0x0000014CBC8FDAB0>

    def test_timelapse_analyse_all_timepoints_with_viewer(make_napari_viewer):
        viewer = make_napari_viewer()

        image = np.asarray([
            [[[1, 2], [3, 4]]],
            [[[5, 6], [7, 8]]]
        ])

        labels = np.asarray([
            [[[1, 2], [3, 4]]],
            [[[1, 2], [3, 4]]]
        ])

        viewer.add_image(image)
        labels_layer = viewer.add_labels(labels)

        regionprops_table_all_frames(image, labels, size=False, napari_viewer=viewer)

        df = pd.DataFrame(labels_layer.properties)

>       assert df.shape[0] == 8
E       assert 4 == 8

napari_skimage_regionprops\_tests\test_timelapse.py:83: AssertionError

...

======================================================== short test summary info =========================================================
FAILED napari_skimage_regionprops/_tests/test_timelapse.py::test_timelapse_analyse_all_timepoints_with_viewer - assert 4 == 8
=============================================== 1 failed, 26 passed, 54 warnings in 45.52s ===============================================

It was the same error displayed here with Github actions. Thus, I made this commit and it worked.

I don't know what causes it, but it seems related to the napari version somehow.

[haesleinhuepf]

Awesome, thanks for working on this Marcelo!