single precision support in skimage.filters

[grlee77]

Description

This PR contains the commits from #5353, so that one should be reviewed prior to this (float32 support Hessian functions in skimage.feature were required in the ridge filters implementation here).

Checklist

• Docstrings for all functions
• Gallery example in ./doc/examples (new features only)
• Benchmark in ./benchmarks, if your changes aren't covered by an
  existing benchmark
• Unit tests
• Clean style in the spirit of PEP8

For reviewers

• Check that the PR title is short, concise, and will make sense 1 year
  later.
• Check that new functions are imported in corresponding __init__.py.
• Check that new features, API changes, and deprecations are mentioned in
  doc/release/release_dev.rst.

[pep8speaks]

Hello @grlee77! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

• In the file skimage/filters/rank/tests/test_rank.py:

Line 6:80: E501 line too long (80 > 79 characters)

• In the file skimage/filters/tests/test_thresholding.py:

Line 21:80: E501 line too long (80 > 79 characters)

• In the file skimage/filters/tests/test_unsharp_mask.py:

Line 118:80: E501 line too long (81 > 79 characters)

Comment last updated at 2021-08-24 23:04:51 UTC

[grlee77]

Hi @rfezzani, thanks for reviewing this previously. I fixed conflicts and addressed your comment about the global eps variable.

Since the last time you reviewed this, I added single precision support to the three threshold_* functions that return arrays (see commits 42c9a94 and 5ebcf52).

Also, it could go in a separate standalone PR if preferred, but there are a couple of minor fixes to prior single precision PRs in fbb1df9 (a dtype, not an image should be passed to _supported_float_type).

[rfezzani]

Thank you @grlee77, a second approval and we are good to go 😉 !

[grlee77]

This is the only one of the single precision PRs that has not yet been merged. It has one approval now and the changes made here are consistent with the already merged PRs in this area. Given that, I would be inclined to merge it as is rather than ask others to spend substantial time doing additional reviews. Are there any objections to this?

[rfezzani]

... Are there any objections to this?

Not from my side ^^