R loop matrix comparing whether the first column is the same as the other column

Question

I wanted would like to check whether in each column of a matrix the values are the same as in the first column:

 x1  <- c("x", "y", "x")
 x2  <- c("x", "y", "y")
 x3  <- c("y", "y","x")

 dat <- data.frame(x1,x2,x3)

   x1 x2 x3
1  x  x  y
2  y  y  y
3  x  y  x

So if x2 != 2, there should be a "1", otherwise a NA. How can I do that? The result in this case would be:

x2     x3 
NA     1
NA     NA
1      NA

My (not working) solution was:

fun <- function (x) {
for(j in 2: ncol(x)){
  ifelse(x[,1]== j, NA,1)
}
}

fun(dat)

I would need a function to perform this with lapply. How can I do that? Thanks a lot!

Answer 1

Different approach using the fact that column-operations are often easy because of argument recycling:

 dat <- data.frame(x1,x2,x3, stringsAsFactors=FALSE) # keeps as character 

 ne.dat1 <-  (dat != dat[,1])[ , -1]  # True/False rather than 1,NA
 is.na(ne.dat1) <- !ne.dat1
 ne.dat1
       x2   x3
[1,]   NA TRUE
[2,]   NA   NA
[3,] TRUE   NA

Answer 2

You could do:

 m1 <- (!(as.character(dat[,1])==dat[,-1])) +1
 m1[] <- c(NA,1)[m1]
 m1
 #    x2 x3
 #[1,] NA  1
 #[2,] NA NA
 #[3,]  1 NA

Answer 3

This is easier with the wrapper sapply as it applies applies the names to the result:

sapply(dat[-1], 
       function(x) 
          ifelse(as.character(x)==as.character(dat[[1]]), NA, 1)
)
##      x2 x3
## [1,] NA  1
## [2,] NA NA
## [3,]  1 NA