Regex Capturing \" in R

Question

I have some data with height strings which are formatted like so

"6'2\""

I'm capturing the first digit just fine, but I can't get rid of the

\"

from the end of the string

I've tried several ways of getting at it but nothing has worked yet. Here's where I'm currently at

inches <- str_extract(htString,"(\\d{1,2})[\\\"]?$")

[1] "11"
[1] "3\""

If the inches string is 2 digits long, I'm able to capture the right characters, otherwise, I'm capturing the \"

Thanks for any help!

Edit: Thanks for the help. The following code ended up working for me. It could be cleaned up I'm sure.

for(i in 1:nrow(hs)){
  htString <- hs[i,]$HtRec

  ft <- str_extract(htString, "^(\\d{1,2})[\']?")
  ft <- substring(ft, 1, 1)

  inches <- str_extract(htString,"(\\d{1,2})[\"]?$")
  inches <- str_extract_all(inches, "\\d+")

  ft <- as.numeric(ft)
  inches <- as.numeric(inches)
  htInches <- (ft * 12) + inches
  hs[i,]$HtRec <- htInches
}

The output you see may not be literal. What I mean by this is that in `[1] "3\""` R is itself escaping the double quotes. In fact, there may not even be any backslashes in your data at all. To confirm this, just write the data frame to a text file and check. — Tim Biegeleisen, Jul 12 '17 at 01:50
I imagine Tim is right. You literally have the text string `6'2"` — thelatemail, Jul 12 '17 at 01:57
@TimBiegeleisen I don't think that is the case because if i do `as.numeric(inches)` then I get back NAs. edit: Oh, I see what you're saying. I'm not capturing the last `"` correctly then. — zfisher, Jul 12 '17 at 01:57
`as.numeric("text")` always gives `NA` - you don't have a number so you get `NA` — thelatemail, Jul 12 '17 at 01:58
@ZachFisher No...this is expected. Would you expect `as.numeric` with a quote in the input to be castable to a number? The slash is irrelevant; it won't work. — Tim Biegeleisen, Jul 12 '17 at 01:59
@ZachFisher - feel free to answer and accept your own question. You get a couple of points of karma and the question gets closed off. Win-win :-) — thelatemail, Jul 12 '17 at 02:10

score 0 · Answer 1 · edited Jun 20 '20 at 09:12

0

I am not expert in R, but If I try:

(\d+).(\d+)

on https://regex101.com/ with test string: "65'2\"" it seems that I can match both numbers:

Group 1. 1-3 65

Group 2. 4-5 2

This uses the concept of capture groups. It seems that str_match() will help you for working with capture groups, take a look at: Regex group capture in R with multiple capture-groups

edited Jun 20 '20 at 09:12

Community

1
1

answered Jul 12 '17 at 01:53

Fabien

4,862
2
19
33

score 0 · Answer 2 · answered Jul 12 '17 at 02:13

Thanks for the help. The following code ended up working for me using @thelatemail's answer. It could be cleaned up for sure.

for(i in 1:nrow(hs)){
  htString <- hs[i,]$HtRec

  ft <- str_extract(htString, "^(\\d{1,2})[\']?")
  ft <- substring(ft, 1, 1)

  inches <- str_extract(htString,"(\\d{1,2})[\"]?$")
  inches <- str_extract_all(inches, "\\d+")

  ft <- as.numeric(ft)
  inches <- as.numeric(inches)
  htInches <- (ft * 12) + inches
  hs[i,]$HtRec <- htInches
}

Regex Capturing \" in R

2 Answers2