Closes #51 (#52)

* #51 reformatted qs_ophtha code so that it doesn't depend on exisitng qs dataset. also improved test data

* #51 chore: Update roxygen

data-raw/qs_ophtha.R

@@ -1,31 +1,31 @@
-# Dataset: qs
-# Description: Create QS test SDTM dataset with VFQ records for ophthalmology studies
+# Dataset: qs_ophtha
+# Description: NEI Visual Functioning Questionnaire 25 (VFQ) test SDTM data
+# for ophthalmology studies, based on QS dataset from CDISC pilot study
 
 # Load libraries -----
-library(dplyr)
-library(stringr)
-library(metatools)
 library(haven)
 library(admiral)
+library(dplyr)
+library(stringr)
+
+# Make qs_ophtha dataset ----
 
-# Create qs_ophtha ----
-## Get qs from CDISC pilot study ----
-sdtm_path <- "https://github.com/cdisc-org/sdtm-adam-pilot-project/blob/master/updated-pilot-submission-package/900172/m5/datasets/cdiscpilot01/tabulations/sdtm/" # nolint
-raw_qs <- read_xpt(paste0(sdtm_path, "qs", ".xpt?raw=true")) %>%
-  convert_blanks_to_na()
+# Get CDISC qs dataset so that we have the right structure ----
+raw_qs <- read_xpt("https://github.com/cdisc-org/sdtm-adam-pilot-project/blob/master/updated-pilot-submission-package/900172/m5/datasets/cdiscpilot01/tabulations/sdtm/qs.xpt?raw=true") # nolint
+qs <- convert_blanks_to_na(raw_qs)
 
-## set seed to get same results each run ----
+## set seed to get same results each run
 set.seed(999)
 
-## add VFQ data ----
-qs1 <- raw_qs %>%
+## create new QS data - keep standard variables from QS ----
+qs1 <- qs %>%
   # select standard variables
   select(STUDYID, DOMAIN, USUBJID, QSBLFL, VISITNUM, VISIT, VISITDY, QSDTC, QSDY) %>%
   # keep unique subjects and visits per subject
   group_by(USUBJID, VISITDY) %>%
   unique()
 
-#### create dummy parameters and results ----
+## create dummy parameters and results ----
 dummy_param <- data.frame(QSTEST = c(
   "Your Overall Health Is",
   "Eyesight Using Both Eyes Is",
@@ -63,7 +63,7 @@ dummy_param <- data.frame(QSTEST = c(
     QSSCAT = "Original Response"
   )
 
-### dummy answers ----
+## dummy answers ----
 
 # difficulty in performing tasks
 difficulty_res <- c(
@@ -92,15 +92,15 @@ yesno <- setNames(yn_res, yn_resn)
 answers <- c(difficulty_res, freq_res, qual_res, yn_res)
 answersn <- c(difficulty_resn, freq_resn, qual_resn, yn_resn)
 
-### assign answers to questions randomly for each subjects ----
+## assign answers to questions randomly for each subjects ----
 
-# take unique subjects
-subjects <- qs1 %>%
+# take unique subject/visit combinations
+subjects_visits <- qs1 %>%
   ungroup() %>%
-  select(USUBJID) %>%
+  select(USUBJID, VISIT) %>%
   distinct()
 
-dummy_param_res_by_subj <- merge(subjects, dummy_param) %>%
+dummy_param_res_by_subj_visit <- merge(subjects_visits, dummy_param) %>%
   mutate(QSORRES = case_when(
     str_detect(QSTEST, "Difficult") ~ sample(difficulty, size = nrow(.), replace = T),
     str_detect(QSTEST, "How") ~ sample(frequency, size = nrow(.), replace = T),
@@ -116,14 +116,14 @@ dummy_param_res_by_subj <- merge(subjects, dummy_param) %>%
     QSDRVFL = ""
   )
 
-### Merge standard QS with parameters and result variables from ophtha QS data ----
-
-qs2 <- merge(qs1, dummy_param_res_by_subj, by = "USUBJID") %>%
+## merge standard QS with parameters and result variables from temp QS data ----
+qs2 <- merge(qs1, dummy_param_res_by_subj_visit, by = c("USUBJID", "VISIT")) %>%
   group_by(USUBJID) %>%
   # create QSSEQ based on VFQ QS parameters
   mutate(QSSEQ = row_number()) %>%
   arrange(USUBJID, QSSEQ)
 
+## create QSSTRESN ----
 qs3 <- qs2 %>%
   group_by(QSTEST) %>%
   # create numeric var for std result
@@ -134,29 +134,7 @@ qs3 <- qs2 %>%
   ) %>%
   ungroup()
 
-## Keep only VFQ records ----
-qs_novfq <- raw_qs %>% filter(QSCAT != "NEI VFQ-25")
-
-qs_ophtha1 <- rbind(qs_novfq, qs3)
-
-## xxSEQ and keep relevant variables ----
-qs_ophtha2 <- qs_ophtha1 %>%
-  select(-QSSEQ) %>%
-  arrange(STUDYID, USUBJID, QSCAT, QSTESTCD, QSDTC, VISITNUM)
-
-qs_ophtha3 <- qs_ophtha2 %>%
-  group_by(STUDYID, USUBJID) %>%
-  mutate(QSSEQ = row_number()) %>%
-  select(
-    STUDYID, DOMAIN, USUBJID, QSSEQ, QSTESTCD, QSTEST, QSCAT, QSSCAT, QSORRES, QSORRESU, QSSTRESC, QSSTRESN, QSSTRESU,
-    QSBLFL, QSDRVFL, VISITNUM, VISIT, VISITDY, QSDTC, QSDY
-  ) %>%
-  ungroup()
-
-qs_ophtha <- qs_ophtha3
-
-# Assign dataset label ----
-attr(qs_ophtha, "label") <- "Questionnaires"
+qs_ophtha <- qs3
 
 # Save dataset ----
 usethis::use_data(qs_ophtha, overwrite = TRUE)