index only if rsqlite

OHDSI · Feb 6, 2025 · 6510e1a · 6510e1a
1 parent 1752836
commit 6510e1a
Show file tree

Hide file tree

Showing 3 changed files with 71 additions and 56 deletions.
diff --git a/R/AndromedaHelperFunctions.R b/R/AndromedaHelperFunctions.R
@@ -1,7 +1,7 @@
-# @file AndromedaHelperFunctions.R 
-# 
-# Copyright 2025 Observational Health Data Sciences and Informatics 
-# 
+# @file AndromedaHelperFunctions.R
+#
+# Copyright 2025 Observational Health Data Sciences and Informatics
+#
 # This file is part of PatientLevelPrediction
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -26,19 +26,23 @@ limitCovariatesToPopulation <- function(covariateData, rowIds) {
   )
 
   covariateData$pop <- data.frame(rowId = rowIds)
-  Andromeda::createIndex(
-    tbl = covariateData$pop, columnNames = "rowId",
-    indexName = "pop_rowIds"
-  )
+  if (inherits(covariateData, "RSQLiteConnection")) {
+    Andromeda::createIndex(
+      tbl = covariateData$pop, columnNames = "rowId",
+      indexName = "pop_rowIds"
+    )
+  }
 
   on.exit(covariateData$pop <- NULL, add = TRUE)
 
   newCovariateData$covariates <- covariateData$covariates %>%
     dplyr::inner_join(covariateData$pop, by = "rowId")
-  Andromeda::createIndex(
-    tbl = newCovariateData$covariates, columnNames = "covariateId",
-    indexName = "covariates_ncovariateIds"
-  )
+  if (inherits(newCovariateData, "RSQLiteConnection")) {
+    Andromeda::createIndex(
+      tbl = newCovariateData$covariates, columnNames = "covariateId",
+      indexName = "covariates_ncovariateIds"
+    )
+  }
 
   metaData$populationSize <- length(rowIds)
   attr(newCovariateData, "metaData") <- metaData
@@ -74,21 +78,23 @@ batchRestrict <- function(covariateData, population, sizeN = 10000000) {
   batchSize = sizeN
   )
 
-  Andromeda::createIndex(
-    tbl = newCovariateData$covariates,
-    columnNames = "covariateId",
-    indexName = "covariates_ncovariateIds"
-  )
-  Andromeda::createIndex(
-    tbl = newCovariateData$covariates,
-    columnNames = "rowId",
-    indexName = "covariates_rowId"
-  )
-  Andromeda::createIndex(
-    tbl = newCovariateData$covariates,
-    columnNames = c("covariateId", "covariateValue"),
-    indexName = "covariates_covariateId_value"
-  )
+  if (inherits(newCovariateData, "RSQLiteConnection")) {
+    Andromeda::createIndex(
+      tbl = newCovariateData$covariates,
+      columnNames = "covariateId",
+      indexName = "covariates_ncovariateIds"
+    )
+    Andromeda::createIndex(
+      tbl = newCovariateData$covariates,
+      columnNames = "rowId",
+      indexName = "covariates_rowId"
+    )
+    Andromeda::createIndex(
+      tbl = newCovariateData$covariates,
+      columnNames = c("covariateId", "covariateValue"),
+      indexName = "covariates_covariateId_value"
+    )
+  }
 
   metaData$populationSize <- nrow(population)
   attr(newCovariateData, "metaData") <- metaData

diff --git a/R/ExtractData.R b/R/ExtractData.R
@@ -405,16 +405,18 @@ getPlpData <- function(
     rowIdField = "row_id",
     covariateSettings = covariateSettings
   )
-  # add indexes for tidyCov + covariate summary
-  Andromeda::createIndex(covariateData$covariates, c("rowId"),
-    indexName = "covariates_rowId"
-  )
-  Andromeda::createIndex(covariateData$covariates, c("covariateId"),
-    indexName = "covariates_covariateId"
-  )
-  Andromeda::createIndex(covariateData$covariates, c("covariateId", "covariateValue"),
-    indexName = "covariates_covariateId_value"
-  )
+  if (inherits(covariateData, "RSQLiteConnection")) {
+    # add indexes for tidyCov + covariate summary
+    Andromeda::createIndex(covariateData$covariates, c("rowId"),
+      indexName = "covariates_rowId"
+    )
+    Andromeda::createIndex(covariateData$covariates, c("covariateId"),
+      indexName = "covariates_covariateId"
+    )
+    Andromeda::createIndex(covariateData$covariates, c("covariateId", "covariateValue"),
+      indexName = "covariates_covariateId_value"
+    )
+  }
 
   if (max(databaseDetails$outcomeIds) != -999) {
     ParallelLogger::logTrace("Fetching outcomes from server")

diff --git a/R/Predict.R b/R/Predict.R
@@ -31,10 +31,11 @@
 #' @examples
 #' coefficients <- data.frame(
 #'   covariateId = c(1002),
-#'   coefficient = c(0.05))
+#'   coefficient = c(0.05)
+#' )
 #' model <- createGlmModel(coefficients, intercept = -2.5)
 #' data("simulationProfile")
-#' plpData <- simulatePlpData(simulationProfile, n=50)
+#' plpData <- simulatePlpData(simulationProfile, n = 50)
 #' prediction <- predictPlp(model, plpData, plpData$cohorts)
 #' # see the predicted risk values
 #' head(prediction)
@@ -119,8 +120,10 @@ predictPlp <- function(plpModel, plpData, population, timepoint) {
 
   attr(prediction, "metaData") <- metaData
   delta <- Sys.time() - start
-  ParallelLogger::logInfo("Prediction done in: ", 
-    signif(delta, 3), " ", attr(delta, "units"))
+  ParallelLogger::logInfo(
+    "Prediction done in: ",
+    signif(delta, 3), " ", attr(delta, "units")
+  )
   return(prediction)
 }
 
@@ -168,9 +171,11 @@ applyTidyCovariateData <- function(
   temp <- covariateData$covariateRef %>% dplyr::collect()
   allCovariateIds <- temp$covariateId
   covariateData$includeCovariates <- data.frame(covariateId = allCovariateIds[!allCovariateIds %in% deleteCovariateIds])
-  Andromeda::createIndex(covariateData$includeCovariates, c("covariateId"),
-    indexName = "includeCovariates_covariateId"
-  )
+  if (inherits(covariateData, "RSQLiteConnection")) {
+    Andromeda::createIndex(covariateData$includeCovariates, c("covariateId"),
+      indexName = "includeCovariates_covariateId"
+    )
+  }
   on.exit(covariateData$includeCovariates <- NULL, add = TRUE)
   # ---
 
@@ -187,11 +192,12 @@ applyTidyCovariateData <- function(
     }
     on.exit(covariateData$maxes <- NULL, add = TRUE)
 
-    # --- added for speed
-    Andromeda::createIndex(covariateData$maxes, c("covariateId"),
-      indexName = "maxes_covariateId"
-    )
-    # ---
+    if (inherits(covariateData, "RSQLiteConnection")) {
+      # --- added for speed
+      Andromeda::createIndex(covariateData$maxes, c("covariateId"),
+        indexName = "maxes_covariateId"
+      )
+    } # ---
 
     newCovariateData$covariates <- covariateData$covariates %>%
       dplyr::inner_join(covariateData$includeCovariates, by = "covariateId") %>% # added as join
@@ -208,14 +214,15 @@ applyTidyCovariateData <- function(
   newCovariateData$covariateRef <- covariateData$covariateRef %>%
     dplyr::inner_join(covariateData$includeCovariates, by = "covariateId")
 
-  # adding index for restrict to pop
-  Andromeda::createIndex(
-    newCovariateData$covariates,
-    c("rowId"),
-    indexName = "ncovariates_rowId"
-  )
-
 
+  if (inherits(newCovariateData, "RSQLiteConnection")) {
+    # adding index for restrict to pop
+    Andromeda::createIndex(
+      newCovariateData$covariates,
+      c("rowId"),
+      indexName = "ncovariates_rowId"
+    )
+  }
   class(newCovariateData) <- "CovariateData"
 
   delta <- Sys.time() - start