How to find the column mean by excluding NA’s and if all values are NA then output NA in R data frame?

To find the column mean by excluding NA’s can be easily done by using na,rm but if we want to have NA if all the values are NA then it won’t be that straight forward. Therefore, in such situation, we can use ifelse function and return the output as NA if all the values are NA as shown in the below examples.

Example1

Consider the below data frame −

 Live Demo

x1<-sample(c(NA,2,3),20,replace=TRUE)
x2<-rep(NA,20)
df1<-data.frame(x1,x2)
df1

Output

   x1   x2
1  2   NA
2  NA  NA
3  NA  NA
4  2   NA
5  2   NA
6  NA  NA
7  3   NA
8  NA  NA
9  2   NA
10 3   NA
11 2   NA
12 NA  NA
13 3   NA
14 2   NA
15 3   NA
16 NA  NA
17 2   NA
18 2   NA
19 2   NA
20 2   NA

Finding the mean of columns x1 and x2 and return NA if all values are NA otherwise getting the mean of remaining values −

ifelse(all(is.na(df1$x1)),NA,mean(df1$x1,na.rm=T))

[1] 2.285714

ifelse(all(is.na(df1$x2)),NA,mean(df1$x2,na.rm=T))

[1] NA

Example2

 Live Demo

y1<-sample(c(NA,rpois(1,5)),20,replace=TRUE)
y2<-rep(NA,20)
df2<-data.frame(y1,y2)
df2

Output

   y1   y2
1  8   NA
2  NA  NA
3  NA  NA
4  NA  NA
5  NA  NA
6  NA  NA
7  8   NA
8  8   NA
9  NA  NA
10 NA  NA
11 8   NA
12 8   NA
13 8   NA
14 8   NA
15 NA  NA
16 NA  NA
17 NA  NA
18 8   NA
19 NA  NA
20 NA  NA

Finding the mean of columns y1 and y2 and return NA if all values are NA otherwise getting the mean of remaining values −

ifelse(all(is.na(df2$y1)),NA,mean(df2$y1,na.rm=T))

[1] 8

ifelse(all(is.na(df2$y2)),NA,mean(df2$y2,na.rm=T))

[1] NA