Splitset.plot_stratification()

[aiqc]

Problem

We put all of this effort into stratifying [and balancing] samples, but we have no way to visualize that for end users.

Solution

• This post demonstrates visualizing the distribution of each split: https://towardsdatascience.com/straightforward-stratification-bb0dcfcaf9ef
• The sample indices are in splitset.samples you can use those to fetch the label classes with Label.to_numpy(sample_indices) and Label.to_pandas(sample_indices).
• Don't worry about folds for now.
• Reference the lines surrounding np.issubdtype related to bin_count for help determining categorical vs continuous labels