Merge pull request #112 from darwin-eu/mc_dev

remove arrow dependency
darwin-eu · Jan 16, 2024 · 14994c9 · 14994c9
2 parents 6ccc5f0 + 0b1009e
commit 14994c9
Show file tree

Hide file tree

Showing 24 changed files with 79 additions and 77 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: CodelistGenerator
 Title: Generate Code Lists for the OMOP Common Data Model
-Version: 2.1.1
+Version: 2.1.2
 Authors@R: c(
     person("Edward", "Burn", email = "[email protected]", 
     role = c("aut", "cre"),
@@ -30,7 +30,6 @@ Imports:
     PatientProfiles (>= 0.3.0),
     RJSONIO
 Suggests: 
-    arrow (>= 9.0.0),
     covr,
     dbplyr (>= 2.2.1),
     knitr,
@@ -49,3 +48,4 @@ Config/testthat/edition: 3
 VignetteBuilder: knitr
 URL: https://darwin-eu.github.io/CodelistGenerator/
 Language: en-US
+Config/testthat/parallel: true
diff --git a/R/mockVocabRef.R b/R/mockVocabRef.R
@@ -319,39 +319,39 @@ mockVocabRef <- function(backend = "database") {
   })
 
   cdm <- CDMConnector::cdm_from_con(db)
-  if (backend == "database") {
-    return(cdm)
-  }
-
-  if (backend %in% c("arrow", "data_frame")) {
-    dOut <- tempfile()
-    dir.create(dOut)
-    CDMConnector::stow(cdm, dOut)
-
-    if (backend == "arrow") {
-      cdmArrow <- CDMConnector::cdm_from_files(
-        path = dOut,
-        as_data_frame = FALSE, cdm_name = "mock_vocab"
-      )
-      DBI::dbDisconnect(attr(cdm, "dbcon"), shutdown = TRUE)
-      return(cdmArrow)
-    }
-
-    if (backend == "data_frame") {
-      if(utils::packageVersion("CDMConnector")<"1.1.0"){
-        cdmDF <- CDMConnector::cdm_from_files(
-          path = dOut,
-          as_data_frame = TRUE
-        )
-      } else {
-        cdmDF <- CDMConnector::cdm_from_files(
-          path = dOut, cdm_name = "mock",
-          as_data_frame = TRUE
-        )
-      }
-
-      DBI::dbDisconnect(attr(cdm, "dbcon"), shutdown = TRUE)
-      return(cdmDF)
-    }
-  }
+  return(cdm)
+  # if (backend == "database") {
+  #   return(cdm)
+  # }
+  #
+  # if (backend %in% c("arrow", "data_frame")) {
+  #   dOut <- tempfile()
+  #   dir.create(dOut)
+  #   CDMConnector::stow(cdm = cdm, path = dOut, format = "csv")
+  #
+  #   if (backend == "arrow") {
+  #     cdmArrow <- CDMConnector::cdm_from_files(
+  #       path = dOut,
+  #       as_data_frame = FALSE, cdm_name = "mock_vocab"
+  #     )
+  #     DBI::dbDisconnect(attr(cdm, "dbcon"), shutdown = TRUE)
+  #     return(cdmArrow)
+  #   }
+  #
+  #   if (backend == "data_frame") {
+  #     if(utils::packageVersion("CDMConnector")<"1.1.0"){
+  #       cdmDF <- CDMConnector::cdm_from_files(
+  #         path = dOut,
+  #         as_data_frame = TRUE
+  #       )
+  #     } else {
+  #       cdmDF <- CDMConnector::cdm_from_files(
+  #         path = dOut, cdm_name = "mock",
+  #         as_data_frame = TRUE
+  #       )
+  #     }
+  #     DBI::dbDisconnect(attr(cdm, "dbcon"), shutdown = TRUE)
+  #     return(cdmDF)
+  #   }
+  # }
 }
diff --git a/extras/precomputeVignetteData.R b/extras/precomputeVignetteData.R
@@ -14,7 +14,7 @@ devtools::load_all()
 
 # example with postgres database connection details
 db <- dbConnect(RPostgres::Postgres(),
-                dbname = "cdm_iqvia_pharmetrics_plus_202203",
+                dbname = "cdm_gold_202301",
                 port = Sys.getenv("DB_PORT") ,
                 host = "163.1.65.51",
                 user = Sys.getenv("DB_USER"),
@@ -46,9 +46,9 @@ rm(cdm)
 # intro vignette ----
 vocabVersion <- getVocabVersion(cdm = cdm_arrow)
 
-saveRDS(
+save(
   vocabVersion,
-  here("vignettes", "introVocab.RData")
+  file = here("vignettes", "introVocab.RData")
 )
 
 
@@ -70,12 +70,12 @@ saveRDS(
 dementiaCodes1 <- getCandidateCodes(cdm = cdm_arrow,
   keywords = "dementia",
   domains = "Condition",
-  searchViaSynonyms = FALSE,
-  fuzzyMatch = FALSE,
+  #searchViaSynonyms = FALSE,
+  #fuzzyMatch = FALSE,
   exclude = NULL,
   includeDescendants = TRUE,
   includeAncestor = FALSE,
-  verbose=TRUE
+  #verbose=TRUE
 )
 saveRDS(
   dementiaCodes1,
@@ -112,15 +112,15 @@ saveRDS(
 oaCodes1 <- getCandidateCodes(cdm = cdm_arrow,
   keywords = "osteoarthritis",
   domains = "Condition",
-  searchViaSynonyms = FALSE,
-  fuzzyMatch = FALSE,
+  #searchViaSynonyms = FALSE,
+  #fuzzyMatch = FALSE,
   exclude = c(
     "post-infection",
     "post-traumatic"
   ),
   includeDescendants = FALSE,
   includeAncestor = FALSE,
-  verbose = TRUE
+  #verbose = TRUE
 )
 saveRDS(
   oaCodes1,
@@ -131,8 +131,8 @@ saveRDS(
 oaCodes2 <- getCandidateCodes(cdm = cdm_arrow,
   keywords = "osteoarthritis",
   domains = "Condition",
-  searchViaSynonyms = FALSE,
-  fuzzyMatch = FALSE,
+  #searchViaSynonyms = FALSE,
+  #fuzzyMatch = FALSE,
   exclude = c(
     "post-infection",
     "post-traumatic"
@@ -149,9 +149,9 @@ saveRDS(
 oaCodes3 <- getCandidateCodes(cdm = cdm_arrow,
   keywords = "osteoarthritis",
   domains = c("Condition", "Observation"),
-  searchViaSynonyms = FALSE,
-  fuzzyMatch = FALSE,
-  maxDistanceCost = 0.1,
+  #searchViaSynonyms = FALSE,
+  #fuzzyMatch = FALSE,
+  #maxDistanceCost = 0.1,
   exclude = c(
     "post-infection",
     "post-traumatic"
@@ -169,9 +169,9 @@ oaCodes4 <- getCandidateCodes(cdm = cdm_arrow,
   keywords = "osteoarthritis",
   domains = "Condition",
   searchInSynonyms = TRUE,
-  searchViaSynonyms = TRUE,
-  fuzzyMatch = FALSE,
-  maxDistanceCost = 0.1,
+  #searchViaSynonyms = TRUE,
+  #fuzzyMatch = FALSE,
+  #maxDistanceCost = 0.1,
   exclude = c(
     "post-infection",
     "post-traumatic"
@@ -201,15 +201,13 @@ saveRDS(
   here("vignettes", "optionsData04.RData")
 )
 
-
-
 # fuzzy search
 oaCodes6 <- getCandidateCodes(cdm = cdm_arrow,
   keywords = "osteoarthritis",
   domains = "Condition",
-  searchViaSynonyms = FALSE,
-  fuzzyMatch = TRUE,
-  maxDistanceCost = 0.1,
+  #searchViaSynonyms = FALSE,
+  #fuzzyMatch = TRUE,
+  #maxDistanceCost = 0.1,
   exclude = c(
     "post-infection",
     "post-traumatic"
@@ -226,9 +224,9 @@ saveRDS(
 oaCodes7 <- getCandidateCodes(cdm = cdm_arrow,
   keywords = "osteoarthritis",
   domains = "Condition",
-  searchViaSynonyms = FALSE,
-  fuzzyMatch = TRUE,
-  maxDistanceCost = 0.2,
+  #searchViaSynonyms = FALSE,
+  #fuzzyMatch = TRUE,
+  #maxDistanceCost = 0.2,
   exclude = c(
     "post-infection",
     "post-traumatic"
@@ -245,9 +243,9 @@ saveRDS(
 oaCodes8 <- getCandidateCodes(cdm = cdm_arrow,
   keywords = "osteoarthritis",
   domains = "Condition",
-  searchViaSynonyms = FALSE,
-  fuzzyMatch = FALSE,
-  maxDistanceCost = 0.2,
+  #searchViaSynonyms = FALSE,
+  #fuzzyMatch = FALSE,
+  #maxDistanceCost = 0.2,
   exclude = c(
     "post-infection",
     "post-traumatic"
@@ -265,8 +263,7 @@ ac_codes_1 <- getCandidateCodes(cdm = cdm_arrow,
                                      keywords="acetaminophen",
                                      domains="drug",
                                      standardConcept="standard",
-                                     includeDescendants = TRUE,
-                                     verbose = TRUE)
+                                     includeDescendants = TRUE)
 saveRDS(
   ac_codes_1,
   here("vignettes", "medData01.RData")
@@ -277,8 +274,7 @@ ac_codes_2a <- getCandidateCodes(cdm = cdm_arrow,
                                              "acetaminophen intravenous"),
                                  domains="drug",
                                  standardConcept="standard",
-                                 includeDescendants = TRUE,
-                                 verbose = TRUE)
+                                 includeDescendants = TRUE)
 saveRDS(
   ac_codes_2a,
   here("vignettes", "medData02a.RData")
@@ -287,10 +283,9 @@ saveRDS(
 ac_codes_2b <- getCandidateCodes(cdm = cdm_arrow,
                                 keywords="acetaminophen",
                                 domains="drug",
-                                doseForm = c("injection", "intravenous"),
+                                #doseForm = c("injection", "intravenous"),
                                 standardConcept="standard",
-                                includeDescendants = TRUE,
-                                verbose = TRUE )
+                                includeDescendants = TRUE)
 saveRDS(
   ac_codes_2b,
   here("vignettes", "medData02b.RData")
@@ -305,11 +300,10 @@ c <- compareCodelists(ac_codes_2, ac_codes_2a)
 ac_codes_3 <- getCandidateCodes(cdm = cdm_arrow,
                                 keywords="acetaminophen",
                                 domains="drug",
-                                conceptClassId = c("Quant Clinical Drug"),
-                                doseForm = c("injection", "intravenous"),
+                                #conceptClassId = c("Quant Clinical Drug"),
+                                #doseForm = c("injection", "intravenous"),
                                 standardConcept="standard",
-                                includeDescendants = TRUE,
-                                verbose = TRUE)
+                                includeDescendants = TRUE)
 saveRDS(
   ac_codes_3,
   here("vignettes", "medData03.RData")

diff --git a/tests/testthat.R b/tests/testthat.R
@@ -1,3 +1,11 @@
+# This file is part of the standard setup for testthat.
+# It is recommended that you do not modify it.
+#
+# Where should you do additional test configuration?
+# Learn more about the roles of various files in:
+# * https://r-pkgs.org/tests.html
+# * https://testthat.r-lib.org/reference/test_package.html#special-files
+
 library(testthat)
 library(CodelistGenerator)
 

diff --git a/vignettes/a01_Introduction_to_CodelistGenerator.Rmd b/vignettes/a01_Introduction_to_CodelistGenerator.Rmd
@@ -60,7 +60,7 @@ getVocabVersion(cdm = cdm)
 ```
 
 ```{r,  message=FALSE, warning=FALSE,echo=FALSE}
-vocabVersion <- readRDS(here("vignettes", "introVocab.RData"))
+vocabVersion <- load(here("vignettes", "introVocab.RData"))
 vocabVersion
 ```
 

diff --git a/vignettes/introData01.RData b/vignettes/introData01.RData
diff --git a/vignettes/introData02.RData b/vignettes/introData02.RData
diff --git a/vignettes/introData03.RData b/vignettes/introData03.RData
diff --git a/vignettes/introData04.RData b/vignettes/introData04.RData
diff --git a/vignettes/introData05.RData b/vignettes/introData05.RData
diff --git a/vignettes/introVocab.RData b/vignettes/introVocab.RData
diff --git a/vignettes/medData01.RData b/vignettes/medData01.RData
diff --git a/vignettes/medData02a.RData b/vignettes/medData02a.RData
diff --git a/vignettes/medData02b.RData b/vignettes/medData02b.RData
diff --git a/vignettes/medData03.RData b/vignettes/medData03.RData
diff --git a/vignettes/medDataConceptClass.RData b/vignettes/medDataConceptClass.RData
diff --git a/vignettes/medDataDoseForms.RData b/vignettes/medDataDoseForms.RData
diff --git a/vignettes/optionsData01.RData b/vignettes/optionsData01.RData
diff --git a/vignettes/optionsData02.RData b/vignettes/optionsData02.RData
diff --git a/vignettes/optionsData03.RData b/vignettes/optionsData03.RData
diff --git a/vignettes/optionsData04.RData b/vignettes/optionsData04.RData
diff --git a/vignettes/optionsData05.RData b/vignettes/optionsData05.RData
diff --git a/vignettes/optionsData06.RData b/vignettes/optionsData06.RData
diff --git a/vignettes/optionsData07.RData b/vignettes/optionsData07.RData