Version 1.0.1 - accumulated changes from development

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: REDCapDM
 Type: Package
 Title: 'REDCap' Data Management
-Version: 1.0.0
+Version: 1.0.1
 Authors@R: c(
     person("João", "Carmezim", email = "jcarmezim@igtp.cat", role = c("aut", "cre")),
     person("Pau", "Satorra", role = c("aut")),
@@ -33,7 +33,8 @@ Imports:
     stringi,
     cli,
     forcats,
-    lifecycle
+    lifecycle,
+    magrittr
 Suggests: 
     knitr,
     rmarkdown,

NAMESPACE

@@ -17,6 +17,7 @@ export(rd_transform)
 export(redcap_data)
 import(cli)
 importFrom(lifecycle,deprecated)
+importFrom(magrittr,`%>%`)
 importFrom(rlang,":=")
 importFrom(rlang,.data)
 importFrom(rlang,eval_tidy)

NEWS.md

@@ -1,3 +1,13 @@
+# REDCapDM 1.0-1
+
+## Bug fixes
+
+- Fixed a bug in `rd_checkbox` that mis-evaluated the branching logic of certain checkboxes if they had previously been transformed with `rd_factor`
+
+## New features
+
+-  Added a `sep` argument to the `redcap_data()` function, allowing users to specify the field separator when reading CSV dictionary files.
+
 # REDCapDM 1.0-0
 
 ## Bug fixes

R/rd_checkbox.R

@@ -11,7 +11,7 @@
 #' @param event_form Only applicable for longitudinal projects (presence of events). Event-to-form mapping for longitudinal projects.
 #' @param checkbox_labels Character vector of length 2 for labels of unchecked/checked values. Default: `c("No", "Yes")`.
 #' @param checkbox_names Logical. If `TRUE` (default), checkbox columns are renamed using choice labels.
-#' @param na_logic Controls how missing values are set based on branching logic. Must be one of `"none"` (do nothing), `"missing"` (set to `NA` only when the logic evaluation is `NA`), or `"eval"` (set to `NA` when the logic evaluates to `FALSE`). Defaults to `"none"`.
+#' @param na_logic Controls how missing values are set based on branching logic. Must be one of `"none"` (do nothing), `"missing"` (set to `NA` only when the logic evaluation is `NA`), or `"eval"` (set to `NA` when the logic evaluates to `FALSE`). Defaults to `"missing"`.
 #'
 #' @return A list with:
 #' \describe{
@@ -49,7 +49,7 @@
 #' @export
 #' @importFrom stats setNames na.omit
 
-rd_checkbox <- function(project = NULL, data = NULL, dic = NULL, event_form = NULL, checkbox_labels = c("No", "Yes"), checkbox_names = TRUE, na_logic = "none") {
+rd_checkbox <- function(project = NULL, data = NULL, dic = NULL, event_form = NULL, checkbox_labels = c("No", "Yes"), checkbox_names = TRUE, na_logic = "missing") {
   results <- NULL
   rlogic_eval <- NULL
 
@@ -121,7 +121,7 @@ rd_checkbox <- function(project = NULL, data = NULL, dic = NULL, event_form = NU
     }
   }
 
-  # Update results with the this transformation
+  # Update results with the transformation
   reason <- if (na_logic == "eval")
     "when the logic isn't satisfied or it's missing"
   else if (na_logic == "missing")
@@ -162,6 +162,12 @@ rd_checkbox <- function(project = NULL, data = NULL, dic = NULL, event_form = NU
         warning("The project contains repeated instruments, and this function cannot accurately evaluate the branching logic of checkboxes in such cases.", call. = FALSE)
       }
 
+      # Handle the fact that a rd_factor was performed before this function
+      if (length(var_check_factors) == 0 & all(!grepl("data\\$", dic$branching_logic_show_field_only_if))) {
+        warning(
+          "The checkboxes are already in factor form. To properly evaluate checkbox branching logic, please run `rd_dictionary()` first.", call. = FALSE
+        )
+      }
 
       caption <- "Checkbox variables advisable to be reviewed"
       review <- NULL
@@ -385,10 +391,23 @@ rd_checkbox <- function(project = NULL, data = NULL, dic = NULL, event_form = NU
       branching_logic_show_field_only_if = stringr::str_replace_all(.data$branching_logic_show_field_only_if, replace2)
     )
 
-  # Returning checkboxes to numeric version
+  # Returning checkboxes to numeric version & applying the same missing values to their factor version
   if (length(var_check_factors) > 0) {
+
+    correspondence <- correspondence |>
+      dplyr::mutate(out_factor = paste0(out, ".factor"))
+
     data <- data |>
-      dplyr::mutate(dplyr::across(correspondence$out, ~ as.numeric(.x)))
+      dplyr::mutate(
+        dplyr::across(dplyr::any_of(correspondence$out), as.numeric),
+        dplyr::across(dplyr::any_of(correspondence$out_factor), ~{
+
+          var_fac <- data[[dplyr::cur_column()]]
+          var_num <- data[[sub("\\.factor$", "", dplyr::cur_column())]]
+
+          dplyr::case_when(!is.na(var_num) ~ var_fac)
+          })
+      )
   }
 
   # Apply the labels to the data

R/rd_query.R

@@ -93,6 +93,9 @@ rd_query <- function(project = NULL, variables = NA, expression = NA, negate = F
   data <- as.data.frame(data)
   dic <- as.data.frame(dic)
 
+  # Ensure variables is a character string
+  variables <- as.character(variables)
+
   # Initialize the query structure
   queries <- as.data.frame(matrix(ncol = 10, nrow = 0))
   colnames(queries) <- c("Identifier", "DAG", "Event", "Instrument", "Field", "Repetition", "Description", "Query", "Code", "Link")

R/redcap_data.R

@@ -28,6 +28,7 @@
 #' @param event_path Path to the event-form mapping file (CSV or XLSX) for longitudinal projects (downloadable via the `Designate Instruments for My Events` tab within the `Project Setup` section of REDCap).
 #' @param uri REDCap API base URI (use with `token`).
 #' @param token REDCap API token (use with `uri`).
+#' @param sep Character string specifying the field separator for the exported dictionary CSV file. Defaul `","`.
 #' @param filter_field Optional character vector of field names to request from the API.
 #' @param survey_fields Logical; include survey-related fields when pulling via API. Default `FALSE`.
 #'
@@ -59,7 +60,7 @@
 #' @export
 #' @importFrom stats setNames
 
-redcap_data <- function(data_path = NA, dic_path = NA, event_path = NA, uri = NA, token = NA, filter_field = NULL, survey_fields = FALSE) {
+redcap_data <- function(data_path = NA, dic_path = NA, event_path = NA, uri = NA, token = NA, sep = ",", filter_field = NULL, survey_fields = FALSE) {
 
   event_form <- NULL
 
@@ -121,7 +122,7 @@ redcap_data <- function(data_path = NA, dic_path = NA, event_path = NA, uri = NA
       dic <- openxlsx::read.xlsx(dic_path, colNames = FALSE, detectDates = TRUE, sheet = 1)
     } else if (extension_dic == "csv") {
       # Read CSV file
-      dic <- utils::read.csv(dic_path, encoding = "UTF-8", header = FALSE)
+      dic <- utils::read.csv(dic_path, encoding = "UTF-8", header = FALSE, sep = sep)
     } else {
       stop("Unsupported dictionary format. Only CSV and XLSX are supported.", call. = FALSE)
     }

R/utils-transform.R

@@ -457,6 +457,7 @@ split_event <- function(data,dic,event_form,which=NULL){
 #' @param event_form Data frame  containing the correspondence of each event with each form.
 #' @param which Character string specifying a form if only data for that form is desired.
 #' @param wide Logical indicating if the dataset should be returned in a wide format (`TRUE`) or long format (`FALSE`).
+#' @importFrom magrittr `%>%`
 
 split_form <- function(data, dic, event_form = NULL, which = NULL, wide=FALSE){
 
@@ -468,7 +469,7 @@ split_form <- function(data, dic, event_form = NULL, which = NULL, wide=FALSE){
   if(longitudinal & is.null(event_form)){
     stop("To split the data by form the event_form has to be provided in a longitudinal project", call. = FALSE)
   }
-  
+
   #Check if the project has repeated instruments
   if("redcap_repeat_instrument" %in% names(data)) {
     repeat_instrument <- dplyr::case_when(
@@ -554,29 +555,29 @@ split_form <- function(data, dic, event_form = NULL, which = NULL, wide=FALSE){
       dplyr::mutate(df = purrr::map(.data$vars,  ~ data |>
                                       dplyr::select(tidyselect::all_of(unique(c(basic_redcap_vars, .x))))))
   }
-  
+
   if(repeat_instrument)  {
     form_check <- data %>%
-    dplyr::distinct(redcap_repeat_instrument, redcap_repeat_instrument.factor)
-    
+    dplyr::distinct(.data$redcap_repeat_instrument, .data$redcap_repeat_instrument.factor)
+
     ndata <- ndata %>%
       dplyr::left_join(form_check, by = dplyr::join_by("form" == "redcap_repeat_instrument")) %>%
       dplyr::relocate("form_factor" = "redcap_repeat_instrument.factor", .after = form) %>%
       dplyr::mutate(df = purrr::map2(.data$form_factor, .data$df, ~ {
         if (is.na(.x)) {
           .y %>%
-            dplyr::filter(is.na(redcap_repeat_instrument.factor)) %>%
+            dplyr::filter(is.na(.data$redcap_repeat_instrument.factor)) %>%
             dplyr::select(-dplyr::starts_with("redcap_repeat_instrument"))
         } else {
           .y %>%
-            dplyr::filter(redcap_repeat_instrument.factor == .x) %>%
+            dplyr::filter(.data$redcap_repeat_instrument.factor == .x) %>%
             dplyr::mutate(redcap_repeat_instrument = redcap_repeat_instrument.factor) %>%
             dplyr::select(-redcap_repeat_instrument.factor)
         }
       })) %>%
       dplyr::select(-"form_factor")
-  } 
-    
+  }
+
 
   if(repeat_instrument)  {
     form_check <- data |>

README.md

@@ -27,7 +27,6 @@ install.packages("remotes") # Run this line if the 'remotes' package isn't insta
 remotes::install_github("bruigtp/REDCapDM")
 ```
 
-
 ### Getting Started
 
 To learn more about the package’s functionality, visit the [**REDCapDM website**](https://bruigtp.github.io/REDCapDM/articles/REDCapDM.html). The site includes detailed descriptions of the package's functions and access to vignettes that demonstrate how to use REDCapDM effectively in your projects.