CODEWITHSUNDEEP
rdplyrtidyr
Victor Yuan
Victor Yuan

Reputation: 47

Adding prefix to column names generated by pivot_longer names_to .value

Is it possible to add a prefix to the column names that are generated from using the .value option in the argument names_to in pivot_longer ?

Example:

df <- tibble(x1 = 0.1, x2 = 0.2, y1 = 0.05, y2 = 0.15)
df %>%
  pivot_longer(cols = everything(),
               names_to = c('coord_type', '.value'),
               names_pattern = '(\\D+)(\\d+)')

# A tibble: 2 × 3
  coord_type   `1`   `2`
  <chr>      <dbl> <dbl>
1 x           0.1   0.2 
2 y           0.05  0.15

I would like the 1 and 2 columns to have prefixes "coord_" in front of them. Of course, I could simply rename them after the pivot_longer like so:

df %>%
  pivot_longer(cols = everything(),
               names_to = c('coord_type', '.value'),
               names_pattern = '(\\D+)(\\d+)') %>%
  rename_with(.fn = ~glue("coord_{.x}"), matches("[12]"))

# A tibble: 2 × 3
  coord_type coord_1 coord_2
  <chr>        <dbl>   <dbl>
1 x             0.1     0.2 
2 y             0.05    0.15

But I'm interested if pivot_longer has some option to do so concisely.

Upvotes: 2

Views: 122

Answers (1)

akrun
akrun

Reputation: 887158

We could use names_repair, which can take a function

library(stringr)
library(tidyr)
pivot_longer(df, cols = everything(),
               names_to = c('coord_type', '.value'),
               names_pattern = '(\\D+)(\\d+)', 
       names_repair = ~ str_replace(.x, "^(\\d+)", "coord_\\1"))

-output

# A tibble: 2 × 3
  coord_type coord_1 coord_2
  <chr>        <dbl>   <dbl>
1 x             0.1     0.2 
2 y             0.05    0.15

Upvotes: 2

Related Questions