Mutate vector according to list names

November 30, 2021

Apologies if this has been asked elsewhere and I wasn’t able to find it. I have a data frame with a character vector, say

library(tidyverse)

mydf <- data.frame(value = c('c','e','a','d'))

and a list, the names imperfectly correspond to the character vector value in mydf

    mylist <- list ('a' = '1',
                    'b' = '2',
                    'c' = '3',
                    'd' = '4')

How can I define a new variable in mydf that takes on the elements of an item in the list if value equals the name of a given list item?

In other words, the desired output is

mydf2 <- data.frame(value = c('c','e','a','d'),
                    value2 = c('3', 'e', '1', '4'))

I assume that this can be done with case_when and perhaps bang-bang notation but I can’t quite square it.

>Solution :

unlist the list and use the named vector to match and replace the values in ‘value’ column and then with coalesce replace the non-matching i.e. NA to original column values

library(dplyr)
mydf3 <- mydf %>%
    mutate(value2 = coalesce(unlist(mylist)[value], value))

-output

mydf3
  value value2
1     c      3
2     e      e
3     a      1
4     d      4

Or another option is a join after converting the named list to a two column data.frame (stack) and then either use coalesce or case_when/ifelse

mydf %>% 
   left_join(stack(mylist), by = c("value"= "ind")) %>% 
   transmute(value, value2 =  case_when(is.na(values) ~ value, TRUE ~ values))
  value value2
1     c      3
2     e      e
3     a      1
4     d      4