Training cache breaks if previous attempt only managed one iteration

OHDSI / DeepPatientLevelPrediction

An R package for performing patient level prediction using deep learning in an observational database in the OMOP Common Data Model.

https://ohdsi.github.io/DeepPatientLevelPrediction

10 stars 4 forks source link

Training cache breaks if previous attempt only managed one iteration #75

Closed egillax closed 1 year ago

egillax commented 1 year ago

to reproduce:

run an example with any model, stop it after it has saved the predictions of an iteration to the training cache. Then rerun the same example. This will give the following error:

Error in trainCache$getLastGridSearchIndex():length(paramSearch): NA/NaN argument

Debugging showed that it's the trainCache$getLastGridSearchIndex() that's failing.

I can share the training cache file with you @lhjohn