execute: Execute the Study

In OHDSI/SkeletonPredictionStudy: A Skeleton for Prediction Development Studies that use the PatientLevelPrediction Framework

Execute the Study

Description

Execute the Study

Usage

execute(
  databaseDetails,
  outputFolder,
  createProtocol = F,
  createCohorts = F,
  runDiagnostic = F,
  viewDiagnostic = F,
  runAnalyses = F,
  createValidationPackage = F,
  useHydra = F,
  skeletonVersion = "v0.0.1",
  analysesToValidate = NULL,
  packageResults = F,
  minCellCount = 5,
  viewShiny = F,
  onlyFetchData = F,
  sampleSize = NULL,
  logSettings
)

Arguments

databaseDetails	Database details created using PatientLevelPrediction::createDatabaseDetails()
outputFolder	Name of local folder to place results; make sure to use forward slashes (/). Do not use a folder on a network drive since this greatly impacts performance.
createProtocol	Creates a protocol based on the analyses specification
createCohorts	Create the cohortTable table with the target population and outcome cohorts?
runDiagnostic	Runs a diagnostic of the T, O and tar settings for the cdmDatabaseSchema - can be used to check whether to change settings or whether the prediction may not do well.
viewDiagnostic	Opens a shiny app with the diagnostic results (run after runDiagnostic completes)
runAnalyses	Run the model development
createValidationPackage	Create a package for sharing the models
useHydra	Whether to use Hydra to create the validation package (requires hydra to be installed) or download the master version of the skeleton (requires internet access)
skeletonVersion	The version of the validation skeleton to use
analysesToValidate	A vector of analysis ids (e.g., c(1,3,10)) specifying which analysese to export into validation package. Default is NULL and all are exported.
packageResults	Should results be packaged for later sharing?
minCellCount	The minimum number of subjects contributing to a count before it can be included in packaged results.
viewShiny	View a shiny app with the results
onlyFetchData	Only fetch data for the analyses without fitting models. Setting this flag will overwrite your input provided to the runAnalyses and createCohorts parameters.
sampleSize	The number of patients in the target cohort to sample (if NULL uses all patients)
logSettings	The log setting PatientLevelPrediction::createLogSettings()

Details

This function executes the SkeletonPredictionStudy Study.

Examples

## Not run: 
connectionDetails <- createConnectionDetails(dbms = "postgresql",
                                             user = "joe",
                                             password = "secret",
                                             server = "myserver")
                                             
 databaseDetails <- PatientLevelPrediction::createDatabaseDetails(
 connectionDetails = connectionDetails,
 cdmDatabaseSchema = cdmDatabaseSchema,
 cdmDatabaseName = cdmDatabaseName,
 tempEmulationSchema = tempEmulationSchema,
 cohortDatabaseSchema = cohortDatabaseSchema,
 cohortTable = cohortTable,
 outcomeDatabaseSchema = cohortDatabaseSchema,
 outcomeTable = cohortTable,
 cdmVersion = cdmVersion
 )  
 
 logSettings <- PatientLevelPrediction::createLogSettings(
 verbosity = "INFO",
 timeStamp = T,
 logName = 'skeletonPlp'
 )                                          

execute(databaseDetails = databaseDetails,
        outputFolder = "c:/temp/study_results", 
        createProtocol = T,
        createCohorts = T,
        runDiagnostic = F,
        viewDiagnostic = F,
        runAnalyses = T,
        createValidationPackage = T,
        useHydra = F,
        skeletonVersion = 'v1.0.1',
        packageResults = F,
        minCellCount = 5,
        viewShiny = F,
        sampleSize = 10000,
        logSettings = logSettings
        )

## End(Not run)

OHDSI/SkeletonPredictionStudy documentation built on March 21, 2022, 3:12 a.m.