v1.1.0 - Merge branch 'post-version-1.0'

.gitignore

@@ -1,46 +1,54 @@
 ### macOS ###
 *.DS_Store
-
 ### R ###
 >>>>>>> rVMAC/master
 # History files
 .Rhistory
 .Rapp.history
-
 # Session Data files
 .RData
-
 # Example code in package build process
 *-Ex.R
-
 # Output files from R CMD build
 /*.tar.gz
-
 # Output files from R CMD check
 /*.Rcheck/
-
 # RStudio files
 .Rproj.user/
-
 <<<<<<< HEAD
 # produced vignettes
 vignettes/*.html
 vignettes/*.pdf
-
 =======
 >>>>>>> rVMAC/master
 # OAuth2 token, see https://github.com/hadley/httr/releases/tag/v0.3
 .httr-oauth
-
 # knitr and R markdown default cache directories
 /*_cache/
 /cache/
-
 # Temporary files created by R markdown
 *.utf8.md
 *.knit.md
 <<<<<<< HEAD
 =======
-
 # packrat
 .Rproj.user
+20180912_differing_variable_labels\.csv
+20180919_comprehensive_redcap_factors\.csv
+20180919_comprehensive_redcap_factors_d20180901\.csv
+abbreviated_differing_values_w_rounding\.csv
+abbreviated_differing_values\.csv
+differing_level_labels\.csv
+DocumentationPeek\.pdf
+full_differing_values\.csv
+variables_in_current_but_not_in_old\.txt
+variables_in_old_but_not_in_current\.txt
+redcap_api_tokens.R
+test_merge_20181127.Rmd
+test_merge_20181127.nb.html
+20181201_processed_data.rds
+epoch4aslmanualscandate.csv
+epoch_4_asl_manual_scan_date.csv
+merged_finally.rds
+mergedacrossepochs.rds
+processed_dec_data.rds

DESCRIPTION

@@ -3,10 +3,10 @@ Type: Package
 Title: A research package for improving statistical collaboration between
     VMAC and VUMC Department of Biostatistics for the Vanderbilt Memory and Aging 
     Project (VMAP)
-Version: 1.0
+Version: 1.1.0
 Authors@R: c(
     person(c("Omair", "A."), "Khan", 
-      email = c("omair.a.khan@vumc.org", "omair@prettynumbe.rs"), 
+      email = c("omair.a.khan@vumc.org"), 
       role = c("aut", "cre")),
     person("Dandan", "Liu", 
       email = "dandan.liu@vanderbilt.edu", 
@@ -33,5 +33,6 @@ Imports: knitr,
   Hmisc,
   formattable,
   REDCapR,
-  secret
+  secret,
+  arsenal
 RoxygenNote: 6.1.0

R/check_shared_vars.R

@@ -12,13 +12,13 @@ check_shared_vars <- function(data1, data2, merge.by) {
     warning(
       paste0(
         "[warning] The datasets (", deparse(substitute(data1)), " and ", deparse(substitute(data2)), ") share the following variable names: ",
-        paste(sharedNames, collapse = ", "), ". \n"
+        paste(sharedNames, collapse = ", "), ". \n\n"
       ),
       immediate. = TRUE
     )
   } else {
-    message(
-      paste0("The datasets (", deparse(substitute(data1)), " and ", deparse(substitute(data2)), ") do not share any variable names except: ", deparse(substitute(merge.by)), ". \n")
+    cat(
+      paste0("The datasets (", deparse(substitute(data1)), " and ", deparse(substitute(data2)), ") do not share any variable names except: ", deparse(substitute(merge.by)), ". \n\n")
     )
   }
 }

R/coerce_and_stack.R

@@ -36,10 +36,11 @@ coerce_and_stack <- function(new_data, old_data) {
         }
       }
     }
+    different_modes.df <- different_modes.df[-1, ]
 
-    if (!is.null(nrow(different_modes.df))) {
+    if (all(!is.null(nrow(different_modes.df)) & nrow(different_modes.df) > 0)) {
       cat("There are differences in storage modes in the following variables:\n")
-      print(different_modes.df[-1, ], row.names = FALSE)
+      print(different_modes.df, row.names = FALSE)
       cat("Numeric variables were coerced to character.\n")
     } else {
       cat("Variables in the new and old data sets share the same storage modes.\n")

R/consolidate_variable_names_and_labels.R

@@ -0,0 +1,30 @@
+consolidate_variable_names_and_labels <- function(data_freeze = MAPfreeze.list) {
+  within_epoch.df <- NULL
+  all_variables_labels.df <- NULL
+
+  for (epoch.i in names(data_freeze)[-1]) {
+    for (dataset.i in names(data_freeze[[epoch.i]][["data"]])) {
+      my.df <- NULL
+
+      if (grepl("\\.static", dataset.i)) {
+        my.df <- dplyr::tibble(
+          epoch = epoch.i,
+          dataset = dataset.i,
+          field_name = names(data_freeze[[epoch.i]][["data"]][[dataset.i]]),
+          field_label = NA_character_
+        )
+      } else {
+        my.df <- dplyr::tibble(
+          epoch = epoch.i,
+          dataset = dataset.i,
+          field_name = data_freeze[[epoch.i]][["metadata"]][[dataset.i]] %>% pull(field_name),
+          field_label = data_freeze[[epoch.i]][["metadata"]][[dataset.i]] %>% pull(field_label)
+        )
+      }
+      within_epoch.df <- rbind(within_epoch.df, my.df)
+    }
+    all_variables_labels.df <- rbind(all_variables_labels.df, within_epoch.df)
+  }
+
+  return(all_variables_labels.df)
+}

R/convert_dates.R

@@ -5,39 +5,38 @@
 #' @export
 
 convert_dates <- function(data) {
-  datevars <- setdiff(names(data)[grepl("\\.date", names(data))], "usa.date")
+  current_data.df <- data
 
-  if ("dob" %in% names(data)) {
-    datevars <- c(datevars, "dob")
+  date.var <- setdiff(names(current_data.df)[grepl("(\\_|\\.)date$", names(current_data.df))], c("usa.date", "usa_date"))
+
+  if ("dob" %in% names(current_data.df)) {
+    date.var <- c(date.var, "dob")
   }
 
-  for (vname in datevars) {
-    datevar <- data[, vname]
+  if (length(date.var) > 0) {
+    for (var.i in date.var) {
+      current_date.var <- current_data.df[, var.i]
 
-    if (is.factor(datevar)) {
-      datevar <- as.character(datevar)
-    }
+      if (is.factor(current_date.var)) {
+        current_date.var <- as.character(current_date.var)
+      }
 
-    if (all(is.na(datevar))) {
-      warning(paste0(vname, " is entirely missing values.\n"))
-      next
-    }
+      if (!is.character(current_date.var) & !is.Date(current_date.var) & !all(is.na(current_date.var))) {
+        warning(paste0(var.i, " is not formatted properly and will not be converted to 'Date' class. This variable should be manually checked.\n\n"))
+        next
+      }
 
-    if (!is.character(datevar) & !is.Date(datevar)) {
-      warning(paste0(vname, " is not in character format and may not need to be converted.\n"))
-      next
-    }
+      if (!is.Date(current_date.var) & any(nchar(na.omit(current_date.var)) %nin% c(0, 10))) {
+        warning(paste0(var.i, " has non-date values and will not be converted to 'Date' class. This variable should be manually checked.\n\n"))
+        next
+      }
 
-    datevar[datevar %in% c("1111-11-11", "", as.Date("1111-11-11"))] <- NA
-    #ifelse(datevar %in% c("1111-11-11", ""), NA, datevar)
+      current_date.var[nchar(current_date.var) == 0] <- NA
+      current_date.var[grep("1111", current_date.var)] <- NA
 
-    if (any((!is.na(datevar)) & nchar(datevar) != 10)) {
-      warning(paste0(vname, " is not in yyyy-mm-dd format.\n"))
-      next
+      current_data.df[, var.i] <- lubridate::as_date(current_date.var)
     }
-
-    data[, vname] <- as.Date(datevar)
   }
 
-  return(data)
+  return(current_data.df)
 }

R/derive_AD_signature.R

@@ -106,7 +106,7 @@ derive_AD_signature <- function(data) {
 
   data <- merge(data, mcevoy.all.temp, by = c("map.id", "epoch"), all.x = TRUE)
 
-  data$AD.sig.schwarz <- rowSums(data[, schwarz], na.rm = FALSE)
+  data$AD.sig.schwarz <- rowMeans(data[, schwarz], na.rm = FALSE)
 
   data <- within(data, {
     label(avg.hippocampus) <- "Avg hippocampus vol - T1 3T FreeSurfer"

R/derive_abp_qc.R

@@ -8,10 +8,10 @@ derive_abp_qc <- function(data) {
   data <- within(data, {
     # ABP QC:
     abp.within.90d.of.np <- ifelse(
-      is.na(days.np.date.minus.abp.date),
+      is.na(days.np.date.minus.abp.date.diff),
       NA,
       ifelse(
-        abs(days.np.date.minus.abp.date) <= 90,
+        abs(days.np.date.minus.abp.date.diff) <= 90,
         1,
         0
       )