mutate-across: truncated names for output columns

Question

When using mutate-across, I have understood that the .names argument may be used to name the output columns. If the desired column names are simply a truncation of the original names, how does one perform the truncation? In the example below, I have used rename_with to achieve the desired outcome and I am wondering if the .names argument could be directly tweaked to produce the same result.

df <- mtcars %>% head(5) %>% dplyr::select(am, wt_mtcars = wt , qsec_mtcars=qsec) 
df
                  wt_mtcars qsec_mtcars
Mazda RX4             2.620       16.46
Mazda RX4 Wag         2.875       17.02
Datsun 710            2.320       18.61
Hornet 4 Drive        3.215       19.44
Hornet Sportabout     3.440       17.02



df %>% 
mutate (across (contains("mtcars"), ~ .*10, .names = "{col}_1")) %>% ## perform a meaningless operation in .fns
  rename_with(~str_remove(., "_mtcars_1"), contains("_mtcars_1"))

                  am wt_mtcars qsec_mtcars    wt  qsec
Mazda RX4          1     2.620       16.46 26.20 164.6
Mazda RX4 Wag      1     2.875       17.02 28.75 170.2
Datsun 710         1     2.320       18.61 23.20 186.1
Hornet 4 Drive     0     3.215       19.44 32.15 194.4
Hornet Sportabout  0     3.440       17.02 34.40 170.2

score 2 · Accepted Answer · answered Jun 14 '21 at 05:09

You can pass the function to apply as string in .names -

library(dplyr)

df %>% 
  mutate(across(contains("mtcars"), ~ .*10,
                .names = '{stringr::str_remove(.col, "_mtcars")}'))

#                  am wt_mtcars qsec_mtcars    wt  qsec
#Mazda RX4          1     2.620       16.46 26.20 164.6
#Mazda RX4 Wag      1     2.875       17.02 28.75 170.2
#Datsun 710         1     2.320       18.61 23.20 186.1
#Hornet 4 Drive     0     3.215       19.44 32.15 194.4
#Hornet Sportabout  0     3.440       17.02 34.40 170.2

Here .col represents current column name.

mutate-across: truncated names for output columns

1 Answers1