Merge pull request #205 from aim-rsf/user_testing

last checks before release

Author: RayStick

R/inputs_collect.R

@@ -14,7 +14,6 @@
 #' @importFrom cli cli_alert_info
 #' @importFrom utils read.csv
 #' @importFrom tools file_path_sans_ext
-#' @importFrom dplyr left_join
 #' @keywords internal
 #' @family metadata_map_internal
 #' @dev generate help files for unexported objects, for developers
@@ -105,8 +104,6 @@ data_load <- function(metadata_file, domain_file, look_up_file, quiet = FALSE) {
       if (!all(colnames(lookup) == expected_column_names)) {
         stop("look_up file does not have expected column names")
       }
-      # Add Domain_Code column into lookup table
-      lookup <- lookup %>% left_join(domains, by = "Domain_Name")
       # Check for look_up rows not covered by domain_list
       no_match <- lookup[!lookup$Domain_Name %in% domains$Domain_Name, ]
       if (nrow(no_match) != 0) {
@@ -148,16 +145,16 @@ data_load <- function(metadata_file, domain_file, look_up_file, quiet = FALSE) {
 #' @dev generate help files for unexported objects, for developers
 
 output_copy <- function(dataset_name, output_dir, quiet = FALSE) {
-  o_search <- paste0("^MAPPING_", gsub(" ", "", dataset_name), "*")
+  o_search <- paste0("^MAPPING_", gsub(" ", "", dataset_name), ".*\\.csv$")
   csv_list <- data.frame(file = list.files(output_dir, pattern = o_search))
   if (nrow(csv_list) != 0) {
     df_list <- lapply(file.path(output_dir, csv_list$file), read.csv)
     df_prev <- do.call("rbind", df_list) # combine all df
     ## make a new date column, order by earliest, remove duplicates & auto
-    df_prev$time2 <- as.POSIXct(df_prev$timestamp, format = "%Y-%m-%d-%H-%M-%S")
+    df_prev$time2 <- as.POSIXct(df_prev$timestamp, format = "%Y-%m-%d %H:%M:%S")
     df_prev <- df_prev[order(df_prev$time2), ]
     df_prev <- df_prev %>% distinct(variable, .keep_all = TRUE)
-    df_prev <- df_prev[-(which(df_prev$note %in% "AUTO CATEGORISED")), ]
+    df_prev <- df_prev[!(df_prev$note %in% "AUTO CATEGORISED"), ]
     df_prev_exist <- TRUE
     if (!quiet) {
       cli_alert_info(paste0("Copying from previous session(s):\n",
@@ -168,6 +165,6 @@ output_copy <- function(dataset_name, output_dir, quiet = FALSE) {
     df_prev_exist <- FALSE
   }
 
-  copy_prev <- list(df_prev = df_prev, df_prev_exist = df_prev_exist)
-  copy_prev
+  list(df_prev = df_prev, df_prev_exist = df_prev_exist)
+
 }

R/metadata_map.R

@@ -51,7 +51,7 @@ select.list <- NULL
 #'     metadata_map(output_dir = temp_output_dir)
 #' }
 #' @export
-#' @importFrom dplyr %>% filter
+#' @importFrom dplyr %>% filter left_join
 #' @importFrom cli cli_alert_info cli_alert_success
 #' @importFrom utils packageVersion write.csv browseURL menu select.list
 #' @importFrom ggplot2 ggsave
@@ -94,7 +94,10 @@ metadata_map <- function(
   # SECTION 1 - DEFINE & PREPARE INPUTS ----
 
   ## Use 'data_load.R' to collect inputs (defaults or user inputs)
-  data <- data_load(metadata_file, domain_file, look_up_file)
+  data <- data_load(metadata_file, domain_file, look_up_file, quiet)
+
+  ## Add Domain_Code column into lookup table
+  data$lookup <- data$lookup %>% left_join(data$domains, by = "Domain_Name")
 
   ## Extract Dataset from metadata_file
   dataset <- data$metadata
@@ -129,7 +132,7 @@ metadata_map <- function(
   if (length(existing_files) > 0) {
     cli_alert_warning(paste("A bar plot already exists for this dataset, saved",
                             "in your output directory.\nSkipping creation",
-                            "of a new plot and opening existing plot.\n\n"))
+                            "of a new plot.\n\n"))
   } else {
     bar_title <- paste0("\n'", dataset_name, "' contains ", n_tables,
                         " table(s)")
@@ -173,13 +176,12 @@ metadata_map <- function(
 
   #### Use 'output_copy.R' to copy from previous output(s) if they exist
   if (table_copy == TRUE) {
-    copy_prev <- output_copy(dataset_name, output_dir)
+    copy_prev <- output_copy(dataset_name, output_dir, quiet)
     df_prev_exist <- copy_prev$df_prev_exist
     df_prev <- copy_prev$df_prev
   } else {
     df_prev_exist <- FALSE
   }
-
   table_note <- readline(paste("Optional note about this table: "))
 
   ####  Extract table from metadata
@@ -202,9 +204,10 @@ metadata_map <- function(
     table_df,
     df_prev_exist,
     df_prev,
-    lookup = data$lookup,
+    data$lookup,
     n_codes,
-    output_df
+    output_df,
+    quiet
   )
 
   output_df$timestamp <- timestamp_now

R/user_interactions.R

@@ -49,7 +49,7 @@ user_categorisation <- function(var, desc, type,
       suppressWarnings(decision_int_max <- max(decision_int, na.rm = TRUE))
       suppressWarnings(decision_int_min <- min(decision_int, na.rm = TRUE))
       if (decision_int_na == TRUE || decision_int_max > domain_code_max ||
-            decision_int_min < 0) {
+            decision_int_min < 1) {
         cli_alert_warning(paste("Formatting is invalid or integer out of",
                                 "range. Provide one integer or a comma",
                                 "seperated list of integers."))
@@ -123,7 +123,7 @@ user_categorisation_loop <- function(start_v, end_v, table_df, df_prev_exist,
 
     ##### search if variable matches any variable from previous table
     if (df_prev_exist == TRUE) {
-      data_v_index <- which(df_prev$var ==
+      data_v_index <- which(df_prev$variable ==
                               table_df$Column.name[data_v])
       df_prev_subset <- df_prev[data_v_index, ]
     } else {

README.md

@@ -87,7 +87,7 @@ Add a note for processing this table:
 
 Optional note about this table: Demo run
 ```
-This table has 35 variables (see 'n of 35' below) but the demo run will only process the first 5 variables. If it skips over a variable (as is the case with 1, 2 and 3) this means it has been auto-categorised. Variable 4 has not been auto-categorised and is asking you to categorise it. You will be asked to categorise a variable with one (or more) of the numbers shown in the [key that has appeared in your plots tab](https://github.com/aim-rsf/mapmetadata/tree/main/inst/outputs/plots_tab_demo_domains.png). We input '8' which means 'Health Info' as defined by the key. The demo simplifies domains for demonstration purposes; for a research study, your domains are likely to be much more specific e.g. ‘Prenatal, antenatal, neonatal and birth’ or ‘Health behaviours and diet’. 
+This table has 35 variables (see 'n of 35' below) but the demo run will only process the first 5 variables (by default). If it skips over a variable (as is the case with 1, 2 and 3) this means it has been auto-categorised. Variable 4 has not been auto-categorised and is asking you to categorise it. You will be asked to categorise a variable with one (or more) of the numbers shown in the [key that has appeared in your plots tab](https://github.com/aim-rsf/mapmetadata/tree/main/inst/outputs/plots_tab_demo_domains.png). We input '8' which means 'Health Info' as defined by the key. The demo simplifies domains for demonstration purposes; for a research study, your domains are likely to be much more specific e.g. ‘Prenatal, antenatal, neonatal and birth’ or ‘Health behaviours and diet’. 
 
 ```
 ℹ Table variable 1 of 35 (5 left to process)

codemeta.json

@@ -226,7 +226,7 @@
     },
     "SystemRequirements": null
   },
-  "fileSize": "2425.471KB",
+  "fileSize": "2451.978KB",
   "citation": [
     {
       "@type": "SoftwareSourceCode",

tests/testthat/test-metadata_map.R

@@ -37,7 +37,8 @@ test_that("metadata_map function works correctly with user input", {
 
   local_mocked_bindings(
     user_categorisation_loop = function(start_v, end_v, table_df, df_prev_exist,
-                                        df_prev, lookup, df_plots, output_df) {
+                                        df_prev, lookup, df_plots, output_df,
+                                        quiet) {
       output_df <- read.csv(demo_output)
       output_df$timestamp <- NA
       output_df$table <- NA

vignettes/mapmetadata.Rmd

@@ -41,7 +41,7 @@ It is recommended to view the descriptions of each table to give you more contex
 
 # Understanding mapping outputs
 
-Running [metadata_map](https://aim-rsf.github.io/mapmetadata/reference/metadata_map.html) will run this function in demo mode (as explained in the README) and generate four files in your project directory. 
+Running [metadata_map](https://aim-rsf.github.io/mapmetadata/reference/metadata_map.html) will run this function in demo mode (as explained in the README) and generate six files in your project directory. 
 
 By default, the demo mode processes the first 5 variables in a table. Running `metadata_map(demo_number = 20)` will process the first 20 variables in a table. 
 Here you can [view outputs generated from this longer demo run](https://github.com/aim-rsf/mapmetadata/tree/main/inst/outputs/), which include:
@@ -166,6 +166,6 @@ If it finds other MAPPING files with the same ID_Name in the output directory, a
 Selection: 4
 ℹ Processing Table 4 of 13 (CHILD)
 
-ℹ Copying from previous session(s): MAPPING_360_NCCHD_CHILD_2025-02-14-18-14-01.csv
+ℹ Copying from previous session(s): MAPPING_360_NCCHD_BLOOD_TEST_2025-02-14-18-14-01.csv
 ```
-Instead of 'AUTO CATEGORISED' the note will say 'COPED FROM: CHILD' but you always have the option to manually override this copying during your review, or turn it off altogether with the option `table_copy = FASLE`.
+Instead of 'AUTO CATEGORISED' the note will say 'COPED FROM: BLOOD_TEST' but you always have the option to manually override this copying during your review, or turn it off altogether with the option `table_copy = FASLE`.