Handle NAs and fixed a bug

IKNL · Jan 17, 2024 · 9623438 · 9623438
1 parent b1380b8
commit 9623438
Show file tree

Hide file tree

Showing 5 changed files with 36 additions and 31 deletions.
diff --git a/vtg.crosstab/src/R/RPC_CT.R b/vtg.crosstab/src/R/RPC_CT.R
@@ -27,13 +27,14 @@ RPC_CT <- function(data, subset_rules, master) {
     return(data)
   }
 
-  vtg::log$info("Rows before NA removal: {nrow(data)}")
   used_variables <- all.vars(master$formula)
-  data <- na.omit(data[, used_variables])
-  vtg::log$info("Rows after NA removal: {nrow(data)}")
+  data <- data[, used_variables]
+  data[is.na(data)] <- "N/A"
 
   for (i in used_variables) {
     data[, i] <- factor(data[, i], levels = master$var_cat[[i]])
+    vtg::log$debug(i)
+    vtg::log$debug(paste(master$var_cat[[i]], collapse = ", "))
   }
   ct <- xtabs(master$formula, data = data)
   return(ct)

diff --git a/vtg.crosstab/src/R/RPC_get_vars.R b/vtg.crosstab/src/R/RPC_get_vars.R
@@ -27,10 +27,10 @@ RPC_get_vars <- function(data, subset_rules, master) {
     return(data)
   }
 
-  vtg::log$info("Rows before NA removal: {nrow(data)}")
   used_variables <- all.vars(master$formula)
-  data <- na.omit(data[, used_variables])
-  vtg::log$info("Rows after NA removal: {nrow(data)}")
+  data <- data[, used_variables]
+  data[is.na(data)] <-"N/A"
+
 
   f <- master$formula
   vars <- apply(data, 2, unique, simplify = F)

diff --git a/vtg.crosstab/src/R/dct.R b/vtg.crosstab/src/R/dct.R
@@ -24,8 +24,6 @@ dct <- function(client, f, margin = NULL, percentage = F,
   # this will call itself without the `use.master.container` option
   #
   if (client$use.master.container) {
-    vtg::log$debug(glue::glue("Running `dct` in master container using
-                                  image '{image.name}'.."))
     result <- client$call(
       "dct",
       f = f,
@@ -43,7 +41,7 @@ dct <- function(client, f, margin = NULL, percentage = F,
   # of the `use.master.container` option)
   client$setOrganizations(organizations_to_include)
 
-  ct <- init_formula(f)
+  ct <- vtg.crosstab::init_formula(f)
 
   vtg::log$info("")
   vtg::log$info("###############################################")
@@ -64,6 +62,7 @@ dct <- function(client, f, margin = NULL, percentage = F,
   # VARIABLE CATEGORIES - COLLECT UNIQUE VARIABLE CATEGORIES FROM NODES
   #######################################################################
   ct <- vtg.crosstab::variable_categories(nodes = nodes, master = ct)
+  vtg::log$debug("ct: {ct}")
 
   vtg::log$info("")
   vtg::log$info("###############################################")

diff --git a/vtg.crosstab/src/R/variable_categories.R b/vtg.crosstab/src/R/variable_categories.R
@@ -12,9 +12,14 @@
 #' @export
 #'
 variable_categories <- function(nodes, master){
-    categories <- sapply(all.vars(master$formula), function(i){
-        unique(as.vector(sapply(nodes, function(j) j[[i]])))
-    }, simplify=F)
-    master$var_cat <- categories
-    return(master)
+
+  used_variables <- all.vars(master$formula)
+
+  for (i in used_variables) {
+    for (j in nodes) {
+      master$var_cat[[i]] <- unique(c(master$var_cat[[i]], j[[i]]))
+    }
+  }
+  return(master)
+
 }
diff --git a/vtg.crosstab/test.R b/vtg.crosstab/test.R
@@ -1,29 +1,29 @@
 rm(list = ls(all.names = TRUE))
+devtools::load_all("./src")
+devtools::load_all("../vtg.preprocessing")
+library(vtg.crosstab)
 
-# This seems to be equivalent to "import x as y"
-library(namespace)
-tryCatch({
-    invisible(registerNamespace('vtg', loadNamespace('vtg')))
-}, error = function(e) {
-    vtg::writeln("Package 'vantage.infrastructure' already loaded.")
-})
+data1 <- read.csv("/mnt/c/data/euracan-node-a.csv")
+data2 <- read.csv("/mnt/c/data/euracan-node-b.csv")
+data2 <- data2[data2$e34_cstage != 5, ]
 
-library(vtg.crosstab)
+# change the first row and make b04_sex collumn 999
+data2$b04_sex[1] <- 999
+
+dataset <- list(data1, data2)
+
+# Data = rbind(data1, data2)
 
-data <- data.frame(Type = paste0("T", rep(1:4, 9*4)),
-                   Subj = gl(9, 4, 36*4),event=rbinom(36*4,1,.5))
-data$Subj=as.numeric(data$Subj)
-D1  <- data[data$Subj%in%c(1,2,3), ]
-D2  <- data[data$Subj%in%c(4,5,6), ]
-D3  <- data[data$Subj%in%c(7,8,9), ]
+data_local <- rbind(vtg.preprocessing::extend_data(data1),
+                    vtg.preprocessing::extend_data(data2))
 
-dataset = list(D1,D2,D3)
+# dataset = rbind(data1, data2)
 
-formula = as.formula(~ Type + Subj + event)
+formula = as.formula(~ e34_cstage + b04_sex)
 
 crosstab.mock <- function(dataset,formula){
     client=vtg::MockClient$new(datasets = dataset,pkgname = 'vtg.crosstab')
-    result=vtg.crosstab::dct(client = client,f = formula,
+    result=vtg.crosstab::dct(client = client, f = formula,
                              organizations_to_include = NULL)
     return(result)
 }