Return matched pattern from vector - R

Question

I'm trying to write code which adds a new column to a dataframe, returning a pattern which has been matched to the respective cell in a different column.

For example, I have a column where the values are a string with a mix of useful and non-useful information, like this:

data.frame(A = c("148apple32394", "386pear3", "23banana3808"))

A
1 148apple32394
2      386pear3
3  23banana3808

I would like to compare this column to a vector of possible patterns, ie:

patterns <- c("apple", "banana", "pear")

and return a new column containing whatever pattern matched, the end result being:

A                B
1 148apple32394  apple
2      386pear3  pear
3  23banana3808  banana

I know grep doesn't work well with vectors of patterns, so is there another good function which might work? Ideally I would like to implement the solution using mutate()

Thanks!

Answer 1

You could use str_extract with the patterns that are collapsed by | to detect and extract the patterns like this:

df = data.frame(A = c("148apple32394", "386pear3", "23banana3808"))
patterns <- c("apple", "banana", "pear")
library(dplyr)
library(stringr)
df %>%
  mutate(B = str_extract(A, paste(patterns, collapse = "|")))
#>               A      B
#> 1 148apple32394  apple
#> 2      386pear3   pear
#> 3  23banana3808 banana

^{Created on 2023-03-10 with reprex v2.0.2}