Replace a value NA with the value from another column in R

Question

I want to replace the NA value in dfABy from the column A, with the value from the column B, based on the year of column year. For example, my df is:

                 >dfABy 
                 A    B   Year
                 56   75  1921
                 NA   45  1921
                 NA   77  1922
                 67   41  1923
                 NA   65  1923

The result what I will attend is:

                 > dfABy
                 A    B   Year
                 56   75  1921
                *45*  45  1921
                *77*  77  1922
                 67   41  1923
                *65*  65  1923

P.S: with the * the value replacing in column A from column B for every year

Solution 1

Perhaps the easiest to read/understand answer in R lexicon is to use ifelse. So borrowing Richard's dataframe we could do:

df <- structure(list(A = c(56L, NA, NA, 67L, NA),
                     B = c(75L, 45L, 77L, 41L, 65L),
                     Year = c(1921L, 1921L, 1922L, 1923L, 1923L)),.Names = c("A", 
                                                                                                                            "B", "Year"), class = "data.frame", row.names = c(NA, -5L))
df$A <- ifelse(is.na(df$A), df$B, df$A)

Solution 2

Now corrected per @Max. (original worked with initial implementation)

The new dplyr function, coalesce, can really simplify these situations.

library(dplyr)

dfABy %>% 
    mutate(A = coalesce(A,B))

Solution 3

The solution provided by GGAnderson did return an error message. Using it inside mutate() however worked fine.

df <- structure(list(A = c(56L, NA, NA, 67L, NA),
                     B = c(75L, 45L, 77L, 41L, 65L),
                     Year = c(1921L, 1921L, 1922L, 1923L, 1923L)),
                .Names = c("A", "B", "Year"), 
                class = "data.frame", 
                row.names = c(NA, -5L))
df
df%>% 
  coalesce(A,B) #returns error

df %>%
mutate(A = coalesce(A,B)) #works

(I am new to Stackoverflow; My low reputation does not allow to comment on GGAnderson´s answer directly)

Solution 4

You could use simple replacement with [<-, subsetting for the NA elements.

df$A[is.na(df$A)] <- df$B[is.na(df$A)]

Or alternatively, within()

within(df, A[is.na(A)] <- B[is.na(A)])

Both give

   A  B Year
1 56 75 1921
2 45 45 1921
3 77 77 1922
4 67 41 1923
5 65 65 1923

Data:

df <- structure(list(A = c(56L, NA, NA, 67L, NA), B = c(75L, 45L, 77L, 
41L, 65L), Year = c(1921L, 1921L, 1922L, 1923L, 1923L)), .Names = c("A", 
"B", "Year"), class = "data.frame", row.names = c(NA, -5L))

Solution 5

Easy

library(dplyr)

dfABy %>%
  mutate(A_new = 
           A %>% 
             is.na %>%
             ifelse(B, A) )

View more solutions

Comments

[Gregor Thomas]

Strongly related: stackoverflow.com/q/19253820/903061 (maybe consider as duplicate?)

[Rich Scriven]

Why add a new column? If you did mutate(A = ...) instead of A_new you will replace the values in the first column, which is what OP wants.

[bramtayl]

I in general tend to program with immutability in mind.

[Sebastian Hesse]

Doesn't work but the other answers corrected it

[GGAnderson]

now corrected - indeed we shouldn't leave obsolete answers hanging around