How do I loop through columns in R?

Question

I am trying to clean up my code for cleaning missing data. I have a dataset with 6 columns and the code works if I were to do them individually like this:

mammographic_masses <- mammographic_masses %>%
  mutate(birad = replace(birad, birad== "na", NA)) %>%
  mutate(birad = replace(birad, birad== "N/A", NA))

But when I try to do it in a for loop like this:

for (i in ncol(mammographic_masses)){
  print(class(mammographic_masses[[i]]))
  mammographic_masses <- mammographic_masses %>%
    mutate(mammographic_masses[[,i]] = replace(mammographic_masses[[,i]], mammographic_masses[[,i]] == "na", NA)) %>%
    mutate(mammographic_masses[[,i]] = replace(mammographic_masses[[,i]], mammographic_masses[[,i]] == "N/A", NA))
}

I get an error:

Error: unexpected '=' in:
"  mammographic_masses <- mammographic_masses %>%
    mutate(mammographic_masses[[,i]] ="
>     mutate(mammographic_masses[[,i]] = replace(mammographic_masses[[,i]], mammographic_masses[[,i]] == "N/A", NA))
Error: unexpected '=' in "    mutate(mammographic_masses[[,i]] ="
> }
Error: unexpected '}' in "}"

I also was reading up on other ways such as apply etc. but i cant figure a way to loop it per column

Answer 1

Instead of looping, use mutate_all .

library(dplyr)

mammographic_masses %>%
  mutate_all(function(x) {is.na(x) <- x %in% c("na", "N/A"); x})
#     V1   V2   V3   V4
#1     d    b <NA>    c
#2     d    b <NA> <NA>
#3  <NA> <NA>    d    b
#4     a <NA> <NA>    b
#5     a    b    d <NA>
#6     d    c    b    c
#7     b    b    d <NA>
#8  <NA> <NA> <NA>    d
#9     a    d    d <NA>
#10 <NA>    b    d <NA>

Test data creation code

set.seed(2020)
n <- 10
mammographic_masses <- replicate(4, sample(c(letters[1:4], "na", "N/A"), n, TRUE))
mammographic_masses <- as.data.frame(mammographic_masses)

Answer 2

To provide the "new" dplyr preferred across as well as str_replace_all (good point @Gregor)

library(dplyr)
library(stringr)


mammographic_masses %>%
  mutate(across(everything(), 
         ~ str_replace_all(., c("na" = NA_character_, "N/A" = NA_character_))))

#>      V1   V2   V3   V4
#> 1     d    b <NA>    c
#> 2     d    b <NA> <NA>
#> 3  <NA> <NA>    d    b
#> 4     a <NA> <NA>    b
#> 5     a    b    d <NA>
#> 6     d    c    b    c
#> 7     b    b    d <NA>
#> 8  <NA> <NA> <NA>    d
#> 9     a    d    d <NA>
#> 10 <NA>    b    d <NA>

Same data as Rui

set.seed(2020)
n <- 10
mammographic_masses <- replicate(4, sample(c(letters[1:4], "na", "N/A"), n, TRUE))
mammographic_masses <- as.data.frame(mammographic_masses)

Answer 3

We can also use na_if

library(dplyr)
mammographic_masses %>%
     mutate(across(everything(), ~ na_if(na_if(., "na"), "N/A")))

data

set.seed(2020)
n <- 10
mammographic_masses <- replicate(4, sample(c(letters[1:4], "na", "N/A"), n, TRUE))
mammographic_masses <- as.data.frame(mammographic_masses)