R/getDistributions.R

In ohdsi-studies/PioneerWatchfulWaiting: PIONEER / EHDEN / OHDSI prostate cancer study

#' @export
getAtEventDistribution <- function(connection, cohortDatabaseSchema, cdmDatabaseSchema, cohortTable,
                                          targetIds, outcomeId, databaseId, packageName, analysisName){
  targetIds <- paste(targetIds, collapse = ', ')
  if(length(outcomeId) == 0){
    sqlFileName <- paste(analysisName, 'sql', sep='.')
  }
  else{
    sqlFileName <- paste('TimeToOutcome', 'sql', sep='.')
  }
  analysisName <- substring(SqlRender::camelCaseToTitleCase(analysisName), 2)
  pathToSql <- system.file("sql", "sql_server", "quartiles", sqlFileName, package = packageName)
  pathToAggregSql <- system.file("sql", "sql_server", "quartiles", 'QuartilesAggregation.sql', package = packageName)
  sql <- readChar(pathToSql, file.info(pathToSql)$size)
  sqlAggreg <- readChar(pathToAggregSql, file.info(pathToAggregSql)$size)
  sql <- paste0(sql, sqlAggreg)
  sql <- SqlRender::render(sql, cohort_database_schema = cohortDatabaseSchema, cdm_database_schema = cdmDatabaseSchema,
                           cohort_table = cohortTable, target_ids = targetIds, analysis_name = analysisName, 
                           outcome_id = outcomeId, warnOnMissingParameters = FALSE)
  sql <- SqlRender::translate(sql, targetDialect = connection@dbms)
  
  data <- DatabaseConnector::querySql(connection, sql, snakeCaseToCamelCase = T)
  
  if (nrow(data) == 0) {
    ParallelLogger::logWarn("There is NO data for atEventDistribution")
    df <- data.frame(matrix(nrow = 0, ncol = 9))
    colnames(df) <- c("cohortDefinitionId", "iqr", "minimum", "q1", "median", "q3", "maximum", "analysisName", "database_id")
    return(df)
  }
  
  return(
    data.frame(cohortDefinitionId = data$cohortDefinitionId, iqr = data$iqr, minimum = data$minimum, q1 = data$q1, 
               median = data$median, q3 = data$q3, maximum = data$maximum, analysisName = data$analysisName, database_id = databaseId)
  )
}