Recode same factor levels values among multiple variables "with mutate_at" in Tidyverse

1
votes

I have several factor variables with values "1" and "2" that need to be recoded to "Yes" and "No". Tidyverse mutate_at with fct_recode appear to be the right tools. However, 

dataframe %>%
 mutate_at(vars (var1, var9, var17) =
              fct_recode(vars(var1, var9, var17), 
                         "Yes" = "1",
                         "No" = "2"))

gives 

Error: unexpected ')' in:
"                         "Yes" = "1",
                         "No" = "2"))"

And, 

dataframe %>%
 mutate_at(vars (var1, var9, var17), 
              funs(fct_recode(vars(var1, var9, var17), 
                         "Yes" = "1",
                         "No" = "2")))

gives

Error: `f` must be a factor (or character vector).

Can someone point at my mistake or do I use mutate_at, vars or fct_recode wrong? Is there a better way in tidyverse to recode same factor levels on multiple variables, a very common task in tidying data.

1
rdplyrtidyverseforcats

1 Answers

8
votes

The problem is your syntax in mutate_at(). 

library(dplyr)

First, I create a sample dataset:

set.seed(666)
dataframe <- data.frame(var1 = sample(c("1", "2"), 5, replace = TRUE),
                        var9 = sample(c("1", "2"), 5, replace = TRUE),
                        var17 = sample(c("1", "2"), 5, replace = TRUE))

  var1 var9 var17
1    2    2     2
2    1    2     1
3    2    1     1
4    1    1     1
5    1    1     1

Then I use mutate_at() like this: .vars are unchanged, but .funs is only the function, additional parameter to funs are passed in ...:

dataframe %>% 
  mutate_at(.vars = vars(var1, var9, var17),
            .funs = forcats::fct_recode,
            "Yes" = "1",
            "No" = "2")

The end result:

  var1 var9 var17
1   No   No    No
2  Yes   No   Yes
3   No  Yes   Yes
4  Yes  Yes   Yes
5  Yes  Yes   Yes