Ability to save multi-animal pose tracks to single-animal files

[DhruvSkyy]

Description

What is this PR

• Bug fix
• Addition of a new feature
• Other

What does this PR do?

This PR adds the functionality to save xarrays as a dictionary of individual pandas dataframes for to_dlc_df and as multiple files for each individual for to_dlc_file. This is based on the split_individuals: bool parameter, if True it will split the data for each individual, if false it will save it as a multi-animal file.

References

This PR addresses Issue #39.

How has this PR been tested?

The PR was tested on a Juypter Notebook against the four scenarios, (credit to @niksirbi):

• The xarray.Dataset has multiple (>1) individuals and the user wants to save it to a multi-animal DeepLabCut dataframe. This currently works with the existing functions.
• The xarray.Dataset has only 1 individual but the user still wants to save it to a multi-animal DeepLabCut dataframe, including the redundant "individuals" level. This also works currently.
• The xarray.Dataset has a single individual and the user wants to save it to single-animal DeepLabCut dataframe (without the "individuals" level). This behaviour needs to be implemented, the output will be a single file.
• The xarray.Dataset has multiple (>1) individuals and the user wants to split them for saving into multiple single-animal files. * This is the trickiest case to implement, as the dataset would have to be split across individuals (behind the scenes) and each part is then saved to a single-animal file as in case 3. The output should be as many files as there are individuals.

Todo

Need to write tests.

Checklist:

• The code has been tested locally
• Tests have been added to cover all new functionality
• The documentation has been updated to reflect any changes
• The code has been formatted with pre-commit

[niksirbi]

Very well done @DruvSkyy 🎉

I like the current implementation of this feature. Thanks to our back-and-forth, we converged on something sensible I think.

I have left some minor comments for you to implement.

After that, I'll modify the tests to cover the new funcitonality. Let's see if they'll help us discover any bugs/edge cases.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (3b61db6) 98.19% compared to head (2115b67) 98.34%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #83      +/-   ##
==========================================
+ Coverage   98.19%   98.34%   +0.14%     
==========================================
  Files           8        8              
  Lines         333      362      +29     
==========================================
+ Hits          327      356      +29     
  Misses          6        6              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[niksirbi]

Hey @DhruvSkyy, I've made the following modification to this PR:

• I've reviewed all docstring and function/argument names in save_poses.py and made sure they are consistent throughout
• I've fixed the tests that were failing
• I've added new tests for the feature you've implemented. You can find these on tests/test_unit/test_save_poses.py, specifically:
  • test_auto_split_individuals
  • test_to_dlc_df_split_individuals
  • test_to_dlc_file_split_individuals

Take a look at the new tests and see if you can make sense of them. I'm happy to discuss these 1-to-1 if some things are unclear.

Other than that, this PR is now ready to merge I think, thanks a lot for your help!

[sonarqubecloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

No Coverage information
0.0% Duplication