How to use "and" with grepl?

Question

This what I have:

f=5.20
y=168.9850
dat=c("dat.txt","dat_5.20.txt","data_5.20_168.9850.txt")
Filter(function(x) grepl(f, x), dat)
# [1] "dat_5.20.txt"           "data_5.20_168.9850.txt"

I need to grep only the one obtained f and y

How to use both f and y in grepl?

The desired result would be:

"data_5.20_168.9850.txt"

Tim Biegeleisen · Answer 1 · 2018-02-26T16:00:10.017

2

One pure regex way of doing this would be to just use two lookahead assertions which independently check for the presence of each of the number strings:

f <- "5\\.20"
y <- "168\\.9850"
dat <- c("dat.txt","dat_5.20.txt","data_5.20_168.9850.txt")

grepl(paste0("(?=.*", f, ")(?=.*", y, ")"), dat, perl=TRUE)

[1] FALSE FALSE  TRUE

The pattern used here is (?=.*5\.20)(?=.*168\.9850).

edited Feb 26 '18 at 16:00

answered Feb 26 '18 at 15:48

Tim Biegeleisen

502,043
27
286
360

@MikeH. No, we don't. Every regex API is different, and some match the entire input by default while others don't. Hard to keep track of this. – Tim Biegeleisen Feb 26 '18 at 16:00

Mike H. · Answer 2 · 2018-02-26T16:18:28.303

2

I suppose if you had a long set of search strings and you didn't want to have to type out everything you could do:

dat[Reduce("&", lapply(c(f,y), function(x, dat) grepl(x, dat), dat = dat))]

However, you could probably also get around typing everything out using @TimBiegeleisen's method by doing something like: paste0("(?=.*", c(f,y), ")", collapse = "") and using the result as your search string.

edited Feb 26 '18 at 16:18

answered Feb 26 '18 at 15:53

Mike H.

13,960
2
29
39

Both of these could be a little less wordy. `dat[Reduce("&", lapply(c(f,y), grepl, x = dat))]` and `paste0("(?=.*", c(f,y), ")", collapse = '')` – IceCreamToucan Feb 26 '18 at 16:02
@Renu, good point about the second. For the first I did it that way so that `dat` was not hardcoded into the function. It is passed as a parameter instead – Mike H. Feb 26 '18 at 16:18

score 1 · Accepted Answer · answered Feb 26 '18 at 15:46

We can do two grep's using any of these alternatives:

grep(y, grep(f, dat, value = TRUE), value = TRUE)
## [1] "data_5.20_168.9850.txt"

dat[grepl(f, dat) & grepl(y, dat)]
## [1] "data_5.20_168.9850.txt"

dat[ intersect(grep(f, dat), grep(y, dat)) ]
## [1] "data_5.20_168.9850.txt"

How to use "and" with grepl?

3 Answers3