CODEWITHSUNDEEP
rfunctiondataframetibble
marcel
marcel

Reputation: 409

Trouble applying a user-defined string replace function to contents of a tibble

I have a user-defined function to replace text patterns. This seems to work on a data frame but not a tibble. 

fixcontents <- function(mydf, mypattern1, mypattern2, mycol) {
  mydf[ ,mycol] <- sub(mypattern1, mypattern2, mydf[ ,mycol])
  return(mydf)
}

mydf1 <- data.frame(col1 = c(1,2), col2 = c("aaa", "bbb"))
mytbl1 <- tibble(col1 = c(1,2), col2 = c("aaa", "bbb"))

fixcontents(mydf1, "(b{3})", "\\1X", 2) # works
mydf1

  col1 col2
1    1  aaa
2    2 bbbX

fixcontents(mytbl1, "(b{3})", "\\1_", 2) # does not work (??)
mytbl1

# A tibble: 2 x 2
   col1 col2 
  <dbl> <chr>
1     1 aaa  
2     2 bbb

Why this behavior, and how do you manipulate data in a tibble?

Upvotes: 1

Views: 35

Answers (1)

Ronak Shah
Ronak Shah

Reputation: 388982

Subsetting using [ works differently on tibbles. Subsetting dataframe returns a vector whereas tibble returns tibble back. 

mydf1[, 2]
#[1] aaa bbb
#Levels: aaa bbb

mytbl1[, 2]
# A tibble: 2 x 1
#  col2 
#  <chr>
#1 aaa  
#2 bbb

Try using [[ to subset

fixcontents <- function(mydf, mypattern1, mypattern2, mycol) {
  mydf[[mycol]] <- sub(mypattern1, mypattern2, mydf[[mycol]])
  return(mydf)
}

fixcontents(mydf1, "(b{3})", "\\1X", 2)
#  col1 col2
#1    1  aaa
#2    2 bbbX

fixcontents(mytbl1, "(b{3})", "\\1_", 2)
# A tibble: 2 x 2
#   col1 col2 
#  <dbl> <chr>
#1     1 aaa  
#2     2 bbb_

Upvotes: 1

Related Questions