REDCapCAST/test-data/test_splitr/debug.r

62 lines
1.6 KiB
R
Raw Normal View History

# Installing the latest REDCapRITS from GitHub ------------------------------
#devtools::install_github("SpectrumHealthResearch/REDCapRITS/R@s3methods")
devtools::install_github("SpectrumHealthResearch/REDCapRITS/R@longitudinal-data")
# Debugging reading in longitudinal datasets ------------------------------
# Reading in the files
file_paths <- file.path(
"../test-data/test_splitr/",
c(
records = "WARRIORtestForSoftwa_DATA_2018-06-21_1431.csv",
metadata = "WARRIORtestForSoftwareUpgrades_DataDictionary_2018-06-21.csv"
)
)
redcap <- lapply(file_paths, read.csv, stringsAsFactors = FALSE)
names(redcap) <- c("records", "metadata")
str(redcap)
# A bunch of blank rows
redcap[["metadata"]] <- redcap[["metadata"]][as.logical(nchar(redcap[["metadata"]][,1])),]
str(redcap, 1)
setdiff(redcap[["metadata"]][,1], names(redcap[["records"]]))
# Viewing the files
View(redcap[["records"]])
View(redcap[["metadata"]])
# Playing with the names --------------------------------------------------
vars_in_data <- names(redcap$records)
universal_fields <- c(
vars_in_data[1],
grep("^redcap_(?!(repeat)).*", vars_in_data, value = TRUE, perl = TRUE)
)
repeat_instrument_fields <- grep("^redcap_repeat.*", vars_in_data, value = TRUE)
# Give it a shot ----------------------------------------------------------
testCheck <- with(redcap, REDCap_split(records, metadata))
lapply(testCheck, names)
commonFields <- Reduce(intersect, lapply(testCheck, names))
commonFields
library(dplyr)
lapply(testCheck, glimpse) %>% invisible
testCheck[[1]] %>%
left_join(testCheck$informed_consent, by = commonFields) %>%
glimpse