Changing NA to “N” in selected factor columns in R data frame

Question

I have a the following data frame with factor columns.

set.seed(1234)
df <- data.frame(a=sample(c("1","2",NA), 10, replace=T),
                b=sample(c("1","2",NA), 10, replace=T), 
                c=sample(c("1","2","3",NA), 10, replace=T))

which is

df
      a    b    c
1     1 <NA>    2
2     2    2    2
3     2    1    1
4     2 <NA>    1
5  <NA>    1    1
6     2 <NA> <NA>
7     1    1    3
8     1    1 <NA>
9     2    1 <NA>
10    2    1    1

Now, I want to create a new level "N" for selected columns and convert all NA in those column to "N". I make a vector of selected column names by

selected <- c("b", "c")

and then try to use apply in the following way

 apply(df, 2, function(x) {(if x %in% selected) x <- factor(x, levels=c(levels(x), 'N'))})

But it gives error:

Error: unexpected symbol in "apply(df, 2, function(x) {(if x"

In my original data, I have lots of columns. So I want to avoid doing it column by column.

Answer 1

The 'levels' of the 'selected' columns before the operation is:

 lapply(df[selected], levels)
 #$b
 #[1] "1" "2"

 #$c
 #[1] "1" "2" "3"

We can 'loop' over the columns in the 'selected' with lapply , include 'N' as one more level in each column, and replace the 'NA' values with 'N'.

 df[selected] <- lapply(df[selected], function(x) {
          levels(x) <- c(levels(x), 'N')
           replace(x, which(is.na(x)), 'N')
            })

Or another option is recode from car , where we can directly change 'NA' to 'N'. It will automatically update the levels.

 library(car)
 df[selected] <- lapply(df[selected], recode, "NA='N'")
 lapply(df[selected], levels)
 #$b
 #[1] "1" "2" "N"

 #$c
 #[1] "1" "2" "3" "N"

Another useful function is addNA if we want to add "NA" one of the levels

df[selected] <- lapply(df[selected], addNA)

NOTE: The output of apply on a non-numeric column will be 'character' class. I guess that is not you wanted.