Finding matching value in one column across several column in an R dataframe

Question

I have a dataframe that comes from a survey where the order the answer choices appeared is randomized, such that we can imagine this dataframe:

example_df <- data.frame(
   Choice = c ('A', 'B', 'A', 'C'), 
   Option1 = c('A', 'A', 'C', 'B'),
   Option2 = c('B', 'C', 'A', 'A'),
   Option3 = c('C', 'B', 'B', 'C'), 
   stringsAsFactors = FALSE
)

Which will look like this:

choice  Option1 Option2 Option3
A         A       B       C
B         A       C       B
A         C       A       B
C         B       A       C

What I'd like back is a new column in the dataframe where the value in the new column is the name of the column that the value in Choice appears in.

In the example, the new column would be:

The example df above would produce this new column:

Option1
Option3
Option2
Option3

Answer 1

We can use max.col

example_df$newcol <- names(example_df)[-1][max.col(example_df[-1] == 
         example_df$Choice)]

-output

example_df
#  Choice Option1 Option2 Option3  newcol
#1      A       A       B       C Option1
#2      B       A       C       B Option3
#3      A       C       A       B Option2
#4      C       B       A       C Option3

---

Or with tidyverse

library(dplyr)
example_df  %>%
   rowwise %>%
   mutate(newcol = names(.)[-1][match(Choice, 
          c_across(starts_with('Option')))]) %>%
   ungroup

-output

# A tibble: 4 x 5
#  Choice Option1 Option2 Option3 newcol 
#  <chr>  <chr>   <chr>   <chr>   <chr>  
#1 A      A       B       C       Option1
#2 B      A       C       B       Option3
#3 A      C       A       B       Option2
#4 C      B       A       C       Option3