The UniformSynthesizer produces multiple UserWarning messages when run on a demo dataset

[R-Palazzo]

Resolve #449
CU-86b72ragn

[sdv-team]

Task linked: CU-86b72ragn SDGym - The UniformSynthesizer produces multiple UserWarning messages when run on a demo dataset #449

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 71.94%. Comparing base (20e32eb) to head (3c033e8).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #452      +/-   ##
==========================================
+ Coverage   71.90%   71.94%   +0.03%     
==========================================
  Files          27       27              
  Lines        2168     2171       +3     
==========================================
+ Hits         1559     1562       +3     
  Misses        609      609              

Flag	Coverage Δ
integration	56.57% <100.00%> (+0.06%)	⬆️
unit	65.96% <100.00%> (+3.32%)	⬆️

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.

[amontanez24]

what is the actual cause of the warnings? Could we change the logic at all to just prevent them instead of filtering them out?

[R-Palazzo]

Hi @andrew,

The issue occurs because we first call hyper_transformer.detect_initial_config(real_data) and then update the configuration to match the metadata sdtypes. Sometimes, detect_initial_config assigns transformers that aren’t compatible with the updated sdtypes.

For example, a datetime column stored as an object might be detected as categorical and assigned the UniformEncoder. Later, when we update the sdtype to datetime, the configuration becomes invalid since UniformEncoder doesn’t support datetime sdtypes — which triggers the warning.

Instead of filtering out the warning, I decided to update the transformer assignment whenever the sdtype changes. This preserves the previous behavior while avoiding the warning.

[amontanez24]

Now that I understand the warning I think it's better to just go back to silencing it 😅 . The reason being that I think it's better to use the methods RDT exposes in the docs than directly changing field_transformers

[amontanez24]

This is annoying, but can you change it back to how you had it before?