Skip to content

Conversation

@symdec
Copy link

@symdec symdec commented Nov 14, 2025

Description

Fix an out-of-memory bug when fitting PrivBayes plugin on dataset with continuous columns.
Fixes #350.

You can consider that there is code duplication between case 1 and 2 (cf. the diff), but it was written this way for more clarity when reading the code (especially the conditions), in my opinion. I am open to merge these two conditions into one, if you prefer.

Affected Dependencies

No additional dependency.

How has this been tested?

  • Run a training of PrivBayes model on dataset with continuous columns, which triggered the bug before the fix and is ok now.
  • Existing privbayes tests have passed (tests/plugins/privacy/test_privbayes.py)

Checklist

@symdec symdec changed the title Fix PrivBayes data encoding bug Bug: Fix PrivBayes data encoding bug Nov 14, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Issue in PrivBayes data encoding (continous/categorical)

1 participant