Flatfield correction and metadata #150
Conversation
Codecov Report
@@ Coverage Diff @@
## master #150 +/- ##
==========================================
- Coverage 80.59% 76.39% -4.20%
==========================================
Files 50 53 +3
Lines 3756 4554 +798
==========================================
+ Hits 3027 3479 +452
- Misses 729 1075 +346
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
|
@mattersoflight @smguo @JohannaRahm This PR is now ready for review. Thanks! |
There was a problem hiding this comment.
Thanks for the PR @jennyfolkesson! The PR looks great. Please see my questions below.
| np.save(file_name, np.squeeze(im_pred), allow_pickle=True) | ||
| else: | ||
| raise ValueError( | ||
| 'Unsupported file extension: {}'.format(self.image_ext), | ||
| ) | ||
| if self.save_figs: | ||
| if self.save_figs and self.image_ext != '.npy': |
There was a problem hiding this comment.
Why can't figures be saved if the file extension is .npy?
There was a problem hiding this comment.
Good point. It's not the file format I was trying to catch but 3D images which are usually the ones saved as numpy format but that's totally unclear. I'll update to check for 3d images.
| flat_field_dir, | ||
| 'flat-field_channel-{}.npy'.format(channel_idx) | ||
| ) | ||
| mp_fn_args.append((im_path, ff_path, block_size, meta_row)) |
There was a problem hiding this comment.
It looks to me that if channel_ids is only a subset of channels in frames_metadata, then for channels not in channel_ids, their ff_path won't get updated. Is this the behavior you expect or I missed something?
There was a problem hiding this comment.
No that's not the behavior I want, thanks for catching that but. I'm now setting ff_path to None inside the for loop so no channels outside channel_ids flatfields should be messing up the calculations.
There was a problem hiding this comment.
Setting ff_path to None inside the for loop makes sense to me. One more comment: It looks like flatfield correction share the parameter channel_ids as other preprocessing steps, but in most use cases we might apply flatfield correction to fluorescence channels but usually not to label-free channels. Should we make a separate channel_ids for flatfield correction? It can be done in the next PR if it requires more code changes.
There was a problem hiding this comment.
Good point! Let's create an issue and I'll take care of it in a separate PR.
Here's another PR for you where:
Now that @JohannaRahm 's PR is merged with master I will rebase to master, then get the master_tests branch merged, then it's time for this one.
After this, I will start working on adding support for zarr input data.