search

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
11 stars 4 forks source link

Add parameter caching for training persistence and continuity #63

Closed lhjohn closed 1 year ago

lhjohn commented 1 year ago

Tier 0 implementation of a training cache. Stores the parameter grid result to the analysis path for each hyperparamter combination. Requires analysis path to be passed to fit function (https://github.com/OHDSI/PatientLevelPrediction/pull/382).

codecov[bot] commented 1 year ago

Codecov Report

Merging #63 (32b890a) into develop (2d07b6b) will increase coverage by 0.34%. The diff coverage is 100.00%.

:exclamation: Current head 32b890a differs from pull request most recent head 424836f. Consider uploading reports for the commit 424836f to get more accurate results

@@             Coverage Diff             @@
##           develop      #63      +/-   ##
===========================================
+ Coverage    99.17%   99.52%   +0.34%     
===========================================
  Files            6        7       +1     
  Lines          608      635      +27     
===========================================
+ Hits           603      632      +29     
+ Misses           5        3       -2
Impacted Files Coverage Δ
R/Estimator.R 100.00% <100.00%> (ø)
R/MLP.R 100.00% <100.00%> (ø)
R/ResNet.R 100.00% <100.00%> (ø)
R/TrainingCache-class.R 100.00% <100.00%> (ø)
R/Transformer.R 94.44% <100.00%> (+0.10%) :arrow_up:

... and 1 file with indirect coverage changes

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

lhjohn commented 1 year ago

I will add some more tests to improve coverage.