R - partial string matching for new variable

Question

I have quite a big dataset which has 2 text variables A and B. length(A) <= length(B). B can be either variable A with some extra characters (without order) or can be totally different from A. So i need to to create new variable within my data table under this condition: If B contains A then C = TRUE. I believe partial string matching is more suitable for me here than normal string comparison.

My dataframe example:

Home      Pick  
Barc      Barcelona 0  
F Munch   FC munchen   
Lakers    Portland

I need to add new variable Side in this way:

Home     Pick         Side    
Barc     Barcelona 0  True  
F Munch  FC munchen   True  
Lakers   Portland     False

i am trying to solve with this:

data_n$Side <- stringMatch(data_n$Home, data_n$Pick, normalize = "YES")

but it gives all negative results.
Hoverer

stringMatch('barcel', 'Barcelona 0', normalize='YES')

gives needed answer. Any hints where i make mistake?

you should include any extra packages that you are using. agrep is a useful base solution for partial matching — rawr
– rawr, Commented May 24, 2014 at 14:55
doesn't stringMatch give a value within the interval of 0 to 1? by the way, using your dataset, I got a value of 0.444 for the dataframe and 0.545 for the single example. — Reuben L.
– Reuben L., Commented May 24, 2014 at 15:01
Hi, thanks for the answer. It actually doesn make such a big difference which funkction to use (agrep, stringmatch or any other). The problem is that i can not make it work on my 17000 rows data file. How should i do that?? :( — Residium
– Residium, Commented May 25, 2014 at 12:17

Rich Scriven · Accepted Answer · 2014-05-24 15:24:42Z

1

I'm not sure of its reliability, but agrepl, the partial pattern matching function, seems to work on your data. Assume dat is your original data, then

## read in the original data
> txt <- "Home\tPick
  Barc\tBarcelona 0
  F Munch\tFC munchen
  Lakers\tPortland"
> dat <- read.table(text = txt, sep = '\t', header = TRUE)
##      Home        Pick
## 1    Barc Barcelona 0
## 2 F Munch  FC munchen
## 3  Lakers    Portland

using agrepl

> d1 <- dat[,1]
> d2 <- dat[,2]
> dat$Side <- sapply(seq(nrow(dat)), function(i){
      agrepl(d1[i], d2[i], ignore.case = TRUE)
      })
> dat
##      Home        Pick  Side
## 1    Barc Barcelona 0  TRUE
## 2 F Munch  FC munchen  TRUE
## 3  Lakers    Portland FALSE

edited May 24, 2014 at 15:24

answered May 24, 2014 at 15:03

Rich Scriven

99.8k11 gold badges190 silver badges252 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Residium Over a year ago

Thanks a lot, it seems to be working from the first sight. As i undersand agrepl function is base funkcion and no additional packages are needed? I get error: Error in FUN(1:17076[[1L]], ...) : could not find function "agrepl"

Residium Over a year ago

just updated R and works fine now. Not 100% accurate but still helped a lot! thanks

Collectives™ on Stack Overflow

R - partial string matching for new variable

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related