是否有用于嵌套 ifelse 函数的 R function ？

Question

I am trying to combine many variables to make a dummy variable for whether somebody belongs to one occupation category, using ifelse. I was wondering if there was a function to simplify this code and make it easier to repeat going forward. For example, my code is currently:

occupation_blue_collar <- ifelse(occupation=="Blue Collar", T, 
                          ifelse(occupation =="Blue Collar and Ex-Military", T, 
                          ifelse(occupation == "Blue Collar and Non-military Government", T,
                          ifelse(occupation== "Blue Collar and School Student", T,
                          ifelse(occupation== "Blue Collar and University Student", T,
                          ifelse(occupation== "Blue Collar and White Collar", T,       
                                        F))))))

I have to do this over many variables and many categories, so I was hoping there was a way to simplify. Thanks!

Answer 1

You could simplify your ifelse statement by using stringr::str_detect in your test expression -

ifelse(str_detect(occupation, “Blue Collar”, TRUE, FALSE))

If you have many variables then dplyr::case_when would be better -

case_when(str_detect(occupation, “Blue Collar”) ~ TRUE,
          str_detect(occupation, “White Collar) ~ TRUE,
          TRUE ~ FALSE)

Answer 2

See case_when which should meet your needs

library(dplyr)
mtcars %>% 
  mutate(cg = case_when(carb <= 3 ~ "low",
                        carb > 3  ~ "high"))

Answer 3

occupation=="Blue Collar" | occupation =="Blue Collar and Ex-Military" |...

Where | is the "or" operator.

Since you seem to have a lot of repeating words (ie "Blue Collar") you should look into regex to see if you can automate some of this repetition.

Answer 4

Few different ways to address this concern, but I think the easiest way to do this is to just use OR logic in your ifelse statement

ifelse(occupation == "Blue Collar" | occupation == "Blue Collar and Ex-Military" | occupation == "Blue Collar and Non-military Government" | occupation == "Blue Collar and School Student" | occupation == "Blue Collar and University Student", "T", "F")

However, if you have to do this many times across occupations, there is an even better way to do this. I would create a csv with the header of occupations_blue_collar, and fill the column with all the possible columns you want. Then read in the csv, and use ifelse(occupation %in% df$occupations_blue_collar, "T", "F") . Rinse and repeat for your other occupations!

Edit: as @markus has pointed out, if all of the values you want to be in occupations_blue_collar have the words 'blue collar' in them, then ifelse(grepl("blue collar", occupation), 'T', 'F') would be the most efficient way to process this, To filter out 'blue collar and white collar', you can use ifelse(grepl('blue collar', occupation) & occupation,= 'blue collar and white collar', 't', 'f') or ifelse(grepl('blue collar', occupation) &,grepl('white collar', occupation), 't', 'f')`

Edit 2: changing || to | as @wusel suggested.