r - How to split a dataframe column by the first instance of a character in its values

Question

I have a dataframe (or vector?)

x <- data.frame(a=c("A_B_D", "B_C"))

I want to split the vector in x$a into two new columns by the first instance of "_" to get

x$b 
[1] "A" "B_D"

and

x$c
[2] "B" "C"

i tried variants of gsub, but couldnt come to a solution.

Answer 1

One idea is to replace the first _ with another delimiter and split on the new delimiter. This works because using sub will only replace the first found delimiter (whereas gsub replaces all), i.e.

strsplit(sub('_', ',', x$a), ',', fixed = TRUE)
#[[1]]
#[1] "A"   "B_D"

#[[2]]
#[1] "B" "C"

To create two new columns in your original data frame,

within(x, new <- data.frame(do.call(rbind, strsplit(sub('_', ',', x$a), ',', fixed = TRUE))))
#      a new.X1 new.X2
#1 A_B_D      A    B_D
#2   B_C      B      C