Add a violin plot visualizations to compare a pair of columns

[R-Palazzo]

CU-86b4g8r3b
Resolve #759

Here are the visuals; let me know if they work. (For Figs 1 and 3, I placed my mouse over the violin to see what is shown.) @npatki

[sdv-team]

Task linked: CU-86b4g8r3b SDMetrics - Add a violin plot visualizations to compare a pair of columns #759

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 95.66%. Comparing base (47ad8d4) to head (2c4186e).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #767   +/-   ##
=======================================
  Coverage   95.65%   95.66%           
=======================================
  Files         115      115           
  Lines        4581     4590    +9     
=======================================
+ Hits         4382     4391    +9     
  Misses        199      199           

Flag	Coverage Δ
integration	80.39% <10.00%> (-0.14%)	⬇️
unit	84.18% <100.00%> (+0.03%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[npatki]

Visuals are looking good!

As discussed in the eng meeting today, could you check what happens if there are different proportions of real vs. synthetic data? For example, if you sample synthetic data to be 5x the size of the real data, will the violin plot for synthetic data be much fatter than the real data? Ideally we don't want that. We want the plots to be normalized within the respective dataset (real or synthetic).

[pvk-developer]

LGTM!

[R-Palazzo]

Visuals are looking good!

As discussed in the eng meeting today, could you check what happens if there are different proportions of real vs. synthetic data? For example, if you sample synthetic data to be 5x the size of the real data, will the violin plot for synthetic data be much fatter than the real data? Ideally we don't want that. We want the plots to be normalized within the respective dataset (real or synthetic).

Hi @npatki thanks for pointing this out.
Based on the attached screenshot, the data is normalized (here the synthetic data is 100 times the size of the real data).

[npatki]

Great @R-Palazzo -- looks good to me!