Failing gsub in for loops

user_7742354 0
Viewed 49

I need to remove a series of substrings from a vector, that might or might not occur multiple times. The substrings to drop are less than the number of strings.

I would like to use a loop but gsub seems to fail in a for loop.

drop <- c("red ","blue ","yellow ")
auto <- data.frame(entry=c("red car","red yellow car","car"))

for(i in 1:length(drop)){
  auto$entry_simple <- gsub(drop[i],"",auto$entry)
}

The loops works only for the last entry how? This is the result

           entry entry_simple
1        red car      red car
2 red yellow car      red car
3            car          car

Instead of 

           entry entry_simple
1        red car          car
2 red yellow car          car
3            car          car
stringr gsub r
2 Answers

We can use str_remove which is vectorized

library(dplyr)
library(stringr)
auto %>% 
   mutate(entry_simple = str_remove(entry, drop))
#       entry entry_simple
#1    red car          car
#2   blue car          car
#3 yellow car          car

If we look at the loop, the gsub is done on the entire column 'entry' and the output is assigned to 'entry_simple' i.e. in each iteration, the 'entry_simple' is gettting changed

lapply(drop, function(x) gsub(x, "", auto$entry))
#[[1]]
#[1] "car"        "blue car"   "yellow car"

#[[2]]
#[1] "red car"    "car"        "yellow car"

#[[3]]
#[1] "red car"  "blue car" "car"

leaving the last one i.e. 'red car' 'blue car', 'car' as the final output

It seems the OP wanted to replace for each corresponding row. In that case, just use the index on the 'x' value for gsub and on the lhs of <-

auto$entry_simple <- auto$entry
for(i in seq_along(drop)) auto$entry_simple[i] <- gsub(drop[i], "", auto$entry[i])
auto
#       entry entry_simple
#1    red car          car
#2   blue car          car
#3 yellow car          car

Update

Based on the OP's updated post

auto$entry_simple <- auto$entry
for(i in 1:length(drop)) auto$entry_simple <- gsub(drop[i],"",auto$entry_simple)
user_37322710

This works. Is it all down to "seq_along"?

for(i in seq_along(drop)) auto$entry <- gsub(drop[i], "",auto$entry)
user_77423540
Related