Search code examples
rdataframetidyrstringr

Separate string into many columns

I'd like to separate each letter or symbol in a string for composing a new data.frame with dimension equals the number of letters. I tried to use the function separate from tidyr package, but the result is not desired.

df <- data.frame(x = c('house', 'mouse'), y = c('count', 'apple'), stringsAsFactors = F)

#unexpected result df[1, ] %>% separate(x, c('A1', 'A2', 'A3', 'A4', 'A5'), sep ='') A1 A2 A3 A4 A5 y 1 count

Expected output

A1  A2  A3  A4  A5
 h   o   u   s   e
 m   o   u   s   e

Solutions using stringr are welcome.

Solution

  • We can use regex lookaround in sep to match the boundary between each character

    library(dplyr)
library(tidyr)
library(stringr)
df %>%
   select(x) %>% 
   separate(x, into = str_c("A", 1:5), sep= "(?<=[a-z])(?=[a-z])")
#  A1 A2 A3 A4 A5
#1  h  o  u  s  e
#2  m  o  u  s  e