I have a series of TRUE and FALSE variables representing search findings, ex: Cellphones, Knifes, Money, etc. My goal is to change the values TRUE for the name of the variable. Note that I would like to do this for 15 or more variables.
df <- data.frame(cellphone = c(TRUE, TRUE, FALSE, TRUE, FALSE),
money = c(FALSE, FALSE, FALSE, TRUE, FALSE),
knife = c(TRUE,TRUE,FALSE, FALSE, FALSE),
whatIneed = c("cellphone", "cellphone", "", "cellphone",""))
cellphone money knife whatIneed
1 TRUE FALSE TRUE cellphone
2 TRUE FALSE TRUE cellphone
3 FALSE FALSE FALSE
4 TRUE TRUE FALSE cellphone
5 FALSE FALSE FALSE
CodePudding user response:
In base R
, an option is to loop over the subset of columns that are logical rowwise, get the first column name based on the logical vector
df$whatIneed <- apply(df[1:3], 1, function(x) names(x)[x][1])
df$whatIneed
[1] "cellphone" "cellphone" NA "cellphone" NA
If we want to do this individually on each column
library(dplyr)
df %>%
mutate(across(where(is.logical),
~ case_when(.x ~ cur_column()), .names = "{.col}_name"))
-output
cellphone money knife whatIneed cellphone_name money_name knife_name
1 TRUE FALSE TRUE cellphone cellphone <NA> knife
2 TRUE FALSE TRUE cellphone cellphone <NA> knife
3 FALSE FALSE FALSE <NA> <NA> <NA>
4 TRUE TRUE FALSE cellphone cellphone money <NA>
5 FALSE FALSE FALSE <NA> <NA> <NA>