apply strsplit rowwise

Question

Dear all, Im trying to split a string on "." and create additional columns with the two strings before and after ".".

tes<-c("1.abc","2.di","3.lik")
dat<-c(5,3,2)
h<-data.frame(tes,dat)
h$num<-substr(h$tes,1,1)

h$prim<-unlist(strsplit(as.character(h$tes),"\\."))[2]
h$prim<-sapply(h$tes,unlist(strsplit(as.character(h$tes),"\\."))[2])

I´d like h$prim to contain "abc","di","lik"..However I´m not able to figure it out. I guess strsplit is not vectorised, but then I thought the sapply version should have worked. However I assume it should be easy:-)

Regards, //M

Answer 1

This should do the trick

R> sapply(strsplit(as.character(h$tes), "\\."), "[[", 2)
[1] "abc" "di"  "lik"

Answer 2

This is the same as rcs' answer, but may be easier to understand:

> sapply(strsplit(as.character(h$tes), "\\."), function(x) x[[2]])
[1] "abc" "di"  "lik"

Answer 3

With the stringr package it's even easier:

library(stringr)
str_split_fixed(h$tes, fixed("."), 2)[, 2]

Answer 4

This question appears several time on StackOverflow.

In exact form as yours:

• Selecting first element of strsplit
• Selecting second element separate by space
• Selecting second element separate by dot I recommend this question to see in how many ways it could be achieved.

Some similar question in this topic:

• Split without separator
• First part as in your question but was change to more complex case

And if you care about speed then you should consider tip from John answer about fixed parameter to strsplit.

Answer 5

Alternatively, you can save yourself the work of pulling out the 2nd element if you add both columns at the same time:

tes <- c("1.abc","2.di","3.lik")
dat <- c(5,3,2)
h <- data.frame(tes, dat, stringsAsFactors=FALSE)
values <- unlist(strsplit(h$tes, ".", fixed=TRUE))
h <- cbind(h, matrix(values, byrow=TRUE, ncol=2,
                     dimnames=list(NULL, c("num", "prim"))))