search

mmcdermott / MEDS_transforms

A simple set of MEDS polars-based ETL and transformation functions
MIT License
20 stars 5 forks source link

Static DataFrame Missing Rows for Patients Without Static Data #205

Closed Oufattole closed 1 month ago

Oufattole commented 1 month ago

Patients with no static data are not represented in the static_df generated during the tokenization transform.

Expected Behavior: All patients should have a row in static_df, even if they lack static data.

Current Behavior: When there is no static data in the dataset (for example if you filter out all the static measurements in a prior filter stage), the static data in the schema folder is all empty.

mmcdermott commented 1 month ago

Related: #197

mmcdermott commented 1 month ago

Hmm -- it looks like maybe we just need to push a new version with this change? I see changes in dev from PR #198 that were added to solve #197 that look like they should fix this. @Oufattole does that look right to you?

mmcdermott commented 1 month ago

Seems like there was a PR to fix this that didn't go in for some reason: https://github.com/mmcdermott/MEDS_transforms/pull/199 I've reopened it, and if it seems viable to you @Oufattole we can merge and push a new version.

Oufattole commented 1 month ago

Sounds great thanks!

mmcdermott commented 1 month ago

@Oufattole I pushed a new version with the hotfix in; please test it and reopen this issue if the problem is not fixed.