R function to add multiple new columns based on values from a group of columns

user_10360595 0
Viewed 165

I am trying to figure out a clean way to add multiple new columns corresponding to multiple existing columns all at once in R. I want each new column cell to be a 0 if the corresponding cell in the corresponding column is an NA value, or a 1 if the corresponding cell has any other value (see example tables below).

I have been able to accomplish this by creating the new columns first, then with a messy which function and a loop. I was wondering if there was a clean way to do this and add in all the new columns with values at once.

Start data

A B
5 NA
NA 4

After function is applied

A B A New B New
5 NA 1 0
NA 4 0 1
dplyr plyr data-manipulation r
2 Answers

With tidyverse, you can use across then apply a function to all columns (i.e., everything()), or you could specify select columns as well. Then, I use .names to create the two new columns.

library(tidyverse)

df %>%
  mutate(across(everything(), ~ifelse(is.na(.x), 0,1), .names = "{col} New"))

Output

   A  B A New B New
1  5 NA     1     0
2 NA  4     0     1

Data

df <- structure(list(A = c(5L, NA), B = c(NA, 4L)), 
                class = "data.frame", row.names = c(NA, -2L))
user_152931910

In base R,

df[paste0(names(df), 'new')] <- +(!is.na(df))
df

#   A  B Anew Bnew
#1  5 NA    1    0
#2 NA  4    0    1

!is.na(df) gives you logical values (TRUE for non-NA values and FALSE for NA values). Adding + at the beginning changes the logical values to integer values (TRUE = 1, FALSE = 0).

user_39629140
Related