Generate random number of missing values in R

Question

I have a data frame such like that:

df<-data.frame(time1=rbinom(100,1,0.3),
               time2=rbinom(100,1,0.4),
               time3=rbinom(100,1,0.5),
               time4=rbinom(100,1,0.6))

How could I generate random missing values for each time variable with up to 20% number of missing? Namely, in this case, the total number of missing less than 20 in each column and they are missed in random from subjects (rows).

Answer 1

You could do:

insert_nas <- function(x) {
  len <- length(x)
  n <- sample(1:floor(0.2*len), 1)
  i <- sample(1:len, n)
  x[i] <- NA 
  x
}

df2 <- sapply(df, insert_nas)
df2

This will give you up to maximal 20% missings per column

colSums(is.na(df2)) / nrow(df2)

time1 time2 time3 time4 
 0.09  0.16  0.19  0.14