# Vectorise Text Function for dplyr::mutate

I'm trying to compare two strings to each other from two colums in a tibble. I want to know if any element of the first string is in the second. I wrote a small function to do this and it should be returning a single value per row. I can use rowwise to do this, but how can I vectorise the function so I don't have to use rowwise?

I see Dean Attali's use of Vectorize(): https://deanattali.com/blog/mutate-non-vectorized/
Seems like cheating. Is there a best practice for vectorizing functions out there?

``````library(tidyverse)
# split input into character vector, compare each one to other input as string and if any match output is TRUE
string_compare <-
function(string_1, string_2) {
map_lgl(unlist(str_split(string_1, "")), function(x) grepl(x, string_2)) %>% any(.)
}

data_tbl <- tibble(NP = c("A", "AB", "BC", "AB"), value = c("A", "C", "A", "BC"))

# errors out
data_tbl %>% mutate(comparison = string_compare(NP, value))
# individual comparisons
string_compare("A", "A")
string_compare("AB", "C")
string_compare("BC", "A")
string_compare("AB", "BC")

# rowwise works
data_tbl %>% rowwise() %>% mutate(comparison = string_compare(NP, value))

#vectorize works, seems like cheating
string_compare_v <- Vectorize(string_compare)
data_tbl %>% mutate(comparison = string_compare_v(NP, value))

``````

Hiya! o/ purrr's `map2` is your friend here ``````data_tbl %>%
mutate(comparison = purrr::map2_lgl(NP, value, string_compare))
``````
3 Likes

/facepalm
An answer staring me in the face as usual.

Thanks @bearloga!

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.