jreps

Adding Logistic Regression with enet

1

The PatientLevelPrediction package updates make it easy to add R classifies into the package. It would be useful to add Logistic Regression with enet into the package.

Add code to create plpModel for an existing GLM

2

The PLP validation skeleton contains code to create a plpModel version of any GLM that has been developed outside of PatientLevelPrediction (see https://github.com/OHDSI/SkeletonPredictionValidationStudy/blob/main/extras/addModel.R ) - This should be added to...

minor suggested updates to package

1

- [fixed] add checkmate to check all inputs for PLP - use roxygen templates for common inputs to use standard terminologies (should we have OHDSI specific templates)?

Replace the fixed AUC optimization to be flexible to custom metrics

incidence rate enhancements

6

### Expected behavior Incidence rate currently only lets you calculate proportion and rate using number of people with outcome (only one count per person) ### Enhancement behavior It would be...

NULL domain for analyses 705-712

1

705 to 712 in https://github.com/OHDSI/FeatureExtraction/blob/main/inst/csv/PrespecAnalyses.csv do not have a domain. Is there a reason for the domain not being Measurement?

It seems that users may be able to specify an incorrectly formatted covariateSettings and the code runs but then hangs when FeatureExtraction is called. It is difficult for users to...

explicit split settings

2

At the moment we can split the data into train/test and folds by patientId, rowId or time. It would be nice to have an explicit splitter where you can provide...

issue with shiny app

2

When I try and view models in the shiny app it fails due to: databaseSettings$databaseTable = 'database_meta_table' but the table name should be databaseSettings$databaseTable = 'database_meta_data'