Inference docs - Githubissues

manoharan-lab / holopy

Hologram processing and light scattering in python

GNU General Public License v3.0

134 stars 50 forks source link

Inference docs #385

Closed barkls closed 3 years ago

barkls commented 3 years ago

Main thing is adding TransformedPrior to tutorial, viewable on readthedocs. Lots of smaller changes as well - see commits/changelist.

Fixes #299. Fixes #376.

pep8speaks commented 3 years ago

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

In the file holopy/inference/nmpfit.py:

Line 199:38: W292 no newline at end of file

In the file holopy/scattering/tests/test_2_color.py:

Line 41:31: E231 missing whitespace after ',' Line 41:36: E231 missing whitespace after ',' Line 41:45: E231 missing whitespace after ',' Line 111:29: E231 missing whitespace after ',' Line 111:47: E231 missing whitespace after ',' Line 111:66: E231 missing whitespace after ','

Comment last updated at 2021-02-16 22:37:05 UTC

barkls commented 3 years ago

I don't feel the need to make those particular PEP8 changes, but I can if you think it's a good idea @briandleahy.

barkls commented 3 years ago

I stopped raising exceptions on FutureWarning to get travis to pass. I think this is ok because what we really want to make sure we catch is DeprecationWarning. When a package bumps versions and uses FutureWarning, other packages we have as dependecies won't update right away, e.g. here h5py updated, but h5netcdf is calling it in a way that generates the FutureWarning. We could pin h5py until h5netcdf updates, but I think it's better to just ignore the FutureWarning.

barkls commented 3 years ago

Thanks for the review @briandleahy! Tests passing now. I'm happy with the hp.fit alpha where it is now (uniform over [0.5, 1], no explicit guess given), but I'm also happy to change it if you prefer.