Get data out of "c(\"a\", \"b\")" format

Question

I have a string "c(\"AV\", \"IM\")", which I'm trying to transform into the string "AV IM".

My issue is that I can't unlist() or flatten() this, as it's a character, and neither paste() nor stringr::str_c() work, since it's technically still 1 character value.

Any ideas how I can do this?

Tidyverse solutions preferred, if possible.

EDIT: I know this can be solved via regex, but I feel like this is more a "fundamental" problem to be solved string-level than it is a regex problem, if that makes any sense.

iI you're the one responsible for getting your data in this shape (maybe using `as.character(substitute(...))` ?) you'd be better off solving this upstream. — moodymudskipper, Jul 08 '18 at 14:37
i'm with you there. Definitely want to fix this early. The problem is I'm not really sure how. I think it's a string-level issue (not character-level, where regex makes sense). — Evan O., Jul 09 '18 at 00:52

Aaron left Stack Overflow · Answer 1 · 2018-07-09T15:10:48.843

4

Not sure how you got here, but this as presented would be an eval/parse situation. However, as noted in many other answers on this site, there's almost always a better way of preparing your data so you end up in a more R-friendly form. See, for starters, What specifically are the dangers of eval(parse(...))?.

> a <- "c(\"AV\", \"IM\")"
> (b <- eval(parse(text=a)))
[1] "AV" "IM"
> paste(b, collapse=" ")
[1] "AV IM"

edited Jul 09 '18 at 15:10

answered Jul 08 '18 at 13:32

Aaron left Stack Overflow

36,704
7
77
142

1

I think here it's precisely a case where `eval(parse(text` is the cleanest solution. What is not clean is what was done to the data before to get it into this form. – moodymudskipper Jul 08 '18 at 14:35
1

Yes, this is exactly what I meant, thanks for the clarification, I'll edit. – Aaron left Stack Overflow Jul 09 '18 at 15:08

score 1 · Answer 2 · answered Jul 08 '18 at 13:40

1

You can also consider to use regular expression to replace all symbols and the beginning c.

s <- "c(\"AV\", \"IM\")"

s_vec <- strsplit(s, split = ",")[[1]]

gsub("[[:punct:]]|^c", "", s_vec)
# [1] "AV"  " IM"

answered Jul 08 '18 at 13:40

www

38,575
12
48
84

score 1 · Answer 3 · answered Jul 08 '18 at 15:44

1

Well it is not quite easy how you got here. You can use eval-parse, though it is not vectorized. And also it is slow. Thus you need a regular expression:

 a <- "c(\"AV\", \"IM\")"
 stringr::str_extract_all(a,"\\w+(?!\\()")
[[1]]
[1] "AV" "IM"

answered Jul 08 '18 at 15:44

Onyambu

67,392
3
24
53

Jack Brookes · Answer 4 · 2018-07-08T13:50:06.373

0

Other answers output a vector. My understanding is you want a space-delimited list of your strings.

library(dplyr)

a <- "c(\"AV\", \"IM\")"

a %>%
  gsub("c(", "", ., fixed=TRUE) %>% 
  gsub("\"", "", ., fixed=TRUE) %>% 
  gsub(",",  "", ., fixed=TRUE) %>% 
  gsub(")",  "", ., fixed=TRUE)

Output

"AV IM"

EDIT Or simply (from @www's answer):

a %>%
  gsub("[[:punct:]]|^c",  "", .)

edited Jul 08 '18 at 13:50

answered Jul 08 '18 at 13:41

Jack Brookes

3,720
2
11
22

Get data out of "c(\"a\", \"b\")" format

4 Answers4