How do I process an entire row when using rowwise()?-CodePudding

My dataset consists of numeric columns, with some missing (i.e. NA) values. I want to find the rows that consist entirely of NA values. For example:

library(tidyverse)
library(magrittr)

frm <- tribble(
  ~A, ~B, ~C,
  11, 22, 33,
  14, NA, 37,
  10, 29, 36,
  NA, NA, NA,
  18, 28, 38
)

I could process each row in a for-loop, using is.na() and all(), but I'd like to find a "tidy" solution. This is the best I could do:

frm %>%
 rowwise %>%
 summarize(all_values_missing = is.na(A) & is.na(B) & is.na(C))

But this approach doesn't scale to datasets with lots of columns and nontrivial column-names. Any ideas would be much appreciated!

CodePudding user response：

Use the c_across function to help. For example

frm %>%
  rowwise %>%
  summarize(all_values_missing = all(is.na(c_across())))

If you only need a subset of columns the c_across() will accept tidy selectors as well.

CodePudding user response：

What do you want your return value to be?

You can use pmap_lgl to iterate rowwise over your dataframe and return a logical vector of whether the row was all NA:

> pmap_lgl(frm, ~all(is.na(.x)))
[1] FALSE FALSE FALSE  TRUE FALSE

You can embed this in your dataframe in a tidy manner via:

> frm %>%
    mutate(all_na = pmap_lgl(., ~all(is.na(.x))))
# A tibble: 5 × 4
      A     B     C all_na
  <dbl> <dbl> <dbl> <lgl> 
1    11    22    33 FALSE 
2    14    NA    37 FALSE 
3    10    29    36 FALSE 
4    NA    NA    NA TRUE  
5    18    28    38 FALSE