From af836b269db69f1bb1032beb22c3a8824835c6b1 Mon Sep 17 00:00:00 2001
From: Nicholas Tierney <nicholas.tierney@gmail.com>
Date: Tue, 19 May 2020 18:25:34 +1000
Subject: [PATCH] a few more uses of colSums(is.na(data) to improve speed

---
 R/gg-miss-which.R | 11 ++++++-----
 R/where-na.R      |  2 +-
 2 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/R/gg-miss-which.R b/R/gg-miss-which.R
index c9f154da..649fb41f 100644
--- a/R/gg-miss-which.R
+++ b/R/gg-miss-which.R
@@ -21,12 +21,13 @@
 
 gg_miss_which <- function(x){
 
+  col_na <- colSums(is.na(x)) == 0
+  col_na_val <- dplyr::if_else(col_na, "complete", "missing")
+
+
   # tell us which columns have missing data
-  ggobject <- x %>%
-    miss_var_summary() %>%
-    dplyr::select(variable,
-                  value = n_miss) %>%
-    dplyr::mutate(value = dplyr::if_else(value == 0, "complete", "missing")) %>%
+  ggobject <- tibble::tibble(variable = names(col_na),
+                             value = col_na_val) %>%
     dplyr::mutate(nheight = 1) %>%
     ggplot(data = .,
            aes(x = variable,
diff --git a/R/where-na.R b/R/where-na.R
index 36077dde..97bb5e69 100644
--- a/R/where-na.R
+++ b/R/where-na.R
@@ -79,7 +79,7 @@ miss_var_which <- function(data){
   }
 
   # else return variables that contain ANY missing values
-  na_vars <- names(which(purrr::map_lgl(data,anyNA) == TRUE))
+  na_vars <- names(which(colSums(is.na(data)) > 0))
 
   return(na_vars)