R/check_qs_dup.R

In sdtmchecks: Data Quality Checks for Study Data Tabulation Model (SDTM) Datasets

#' @title Check for multiple dates at the same visit in QS
#'
#' @description Identifies multiple dates at the same visit in QS
#'
#' @param QS QS SDTM dataset with variables USUBJID, QSCAT, VISIT, QSDTC
#'
#' @return Boolean value for whether the check passed or failed, with 'msg'
#'   attribute if the test failed
#'
#' @export
#'
#' @importFrom dplyr %>% filter select rename mutate distinct slice group_by n
#'
#' @author Yuliia Bahatska
#'
#' @examples
#'
#' QS1 <- data.frame(USUBJID = c(rep(101, 5), rep(102, 5)),
#'                 QSCAT = "DLQI",
#'                 QSDTC = rep(c("2017-01-01T08:25", "2017-01-05T09:25",
#'                  "2017-01-15T10:25","2017-01-20T08:25","2017-01-25T08:25"), 2),
#'                 VISITNUM = rep(1:5,2),
#'                 VISIT = rep(c( "Visit 1", "Visit 2", "Visit 3", "UNSCheduled!!!","VIsit 5"), 2),
#'                 stringsAsFactors = FALSE)
#' check_qs_dup(QS = QS1)
#'
#' # multiple dates for the same visit in QS
#' QS2 <- QS1
#' QS2$VISIT[QS2$USUBJID == 101] <- "Visit 1"
#' check_qs_dup(QS = QS2)
#'
#' # multiple visit labels for the same date
#' QS3 <- QS1
#' QS3$QSDTC[QS3$USUBJID == 101] <- "2017-01-01"
#' QS3
#' check_qs_dup(QS = QS3)
#'
#'




check_qs_dup <- function(QS) {

    # Checks whether required variables are in dataset
    if (QS %lacks_any% c("USUBJID", "QSCAT", "QSDTC", "VISIT")) {

        fail(lacks_msg(QS, c("USUBJID", "QSCAT", "QSDTC", "VISIT")))


    } else {

        # Remove time
         QS <- mutate(QS, QSDTC1 = substr(QSDTC,1,10))
         # Get unique records by category, date, visit

        df<-QS %>%
            select(USUBJID, QSCAT, QSDTC1, VISIT) %>%
            rename(QSDTC=QSDTC1) %>%
            group_by(USUBJID, QSCAT, QSDTC) %>%
            slice(1)
        
        # Get duplicates records by category, date
        df<- subset(df,!grepl("UNSCHEDU",toupper(df$VISIT)),)
        df2<-df %>% 
            group_by(USUBJID, QSCAT, VISIT) %>%
            filter( n() > 1 )

        # Outputs a resulting message depending on whether there are duplicates
        if (nrow(df2) != 0) {

            fail("Multiple dates for the same visit in QS. ", df2)

        } else {

            pass()

        }
    }
}