CODEWITHSUNDEEP
rdplyr
Syl33
Syl33

Reputation: 97

Change NA to "" in a whole data.frame

I would like to change all the NA to "" in a data.frame made of string. I tried to use the mutate_each function of Dplyr package but it doesn't work (I get numbers in place of the strings).

Df_with_NA <- as.data.frame(rbind(c("toto", "titi", NA, NA), c("tata", "tztz", "tutu", NA), c("toto","titi", "tutu", "tyty")))

empty_as_na <- function(x){
  ifelse(is.na(x), "", x)
}

Df_with_empty_string_instead_of_NA <- Df_with_NA %>% mutate_each(funs(empty_as_na))

Can you tell me what is wrong ?

Thank

Upvotes: 1

Views: 1231

Answers (2)

d.b
d.b

Reputation: 32538

#First convert elements of 'Df_with_NA' to character and store in 'df'
#This step is necessary because you didn't use stringsAsFactors = FALSE
#when creating 'Df_with_NA'
df = sapply(Df_with_NA, as.character)

#Then replace NA with ""
df[is.na(df)] = ""
df
#     V1     V2     V3     V4    
#[1,] "toto" "titi" ""     ""    
#[2,] "tata" "tztz" "tutu" ""    
#[3,] "toto" "titi" "tutu" "tyty"

Upvotes: 2

gung - Reinstate Monica
gung - Reinstate Monica

Reputation: 11893

You should be just fine using plain, old apply(). Consider: 

Df_with_NA <- as.data.frame(rbind(c("toto", "titi", NA, NA), 
                                  c("tata", "tztz", "tutu", NA), 
                                  c("toto","titi", "tutu", "tyty")))
Df_with_NA
#     V1   V2   V3   V4
# 1 toto titi <NA> <NA>
# 2 tata tztz tutu <NA>
# 3 toto titi tutu tyty
empty_as_na <- function(x){
  ifelse(is.na(x), "", x)
}

apply(Df_with_NA, 2, empty_as_na)
#      V1     V2     V3     V4    
# [1,] "toto" "titi" ""     ""    
# [2,] "tata" "tztz" "tutu" ""    
# [3,] "toto" "titi" "tutu" "tyty"

Upvotes: 0

Related Questions