CODEWITHSUNDEEP
rdataframedplyr
ChazC
ChazC

Reputation: 61

Most efficient way to mutate multiple new columns at once using the same formula but different existing columns?

Say I have this dataframe 'mydata':

mydata <- data.frame(x = 1:5, y = -2:2, z=8:12, a = 7)

mydata
  x  y  z a
1 1 -2  8 7
2 2 -1  9 7
3 3  0 10 7
4 4  1 11 7
5 5  2 12 7

I want to create a new column that subtracts column 'x' from column 'a', and then do the same thing for columns y and z. If possible, I would then like columns x, y and z to be removed from the data frame. This would be the ideal resulting data frame:

  a new_x new_y new_z
1 7    -6    -9     1
2 7    -5    -8     2
3 7    -4    -7     3
4 7    -3    -6     4
5 7    -2    -5     5

I have about 30 columns to do this and about 10,000 rows. I am currently using mutate for each new column, like this:

mydata <- mydata %>% mutate(new_x = x-a, new_y = y-a, new_z = z-a)

Is there an efficient way to accomplish this (preferably via dplyr) that isn't so repetitive?

Thank you!!

Upvotes: 1

Views: 86

Answers (1)

akrun
akrun

Reputation: 887088

We can use across to loop over the columns, rename the columns with .name by adding the prefix new_. By default, it returns the original columns as well unless we make use of .keep

library(dplyr)
mydata %>% 
  mutate(a, across(x:z, ~ .x - a, .names = 'new_{.col}'), .keep  = 'unused')

-output

  a new_x new_y new_z
1 7    -6    -9     1
2 7    -5    -8     2
3 7    -4    -7     3
4 7    -3    -6     4
5 7    -2    -5     5

Upvotes: 2

Related Questions