CODEWITHSUNDEEP
r
mathematical.coffee
mathematical.coffee

Reputation: 56915

R array - subsetting with dummy singleton dimension

I've been writing a number of functions on arrays in R that are of an image-processing type.

What this means is that the functions operate on:

The thing is, greyscale images are 2-dimensional: dim(ary) is c(n,m). By contrast, colour images are 3-dimensional: dim(ary) is c(n,m,3).

At the moment my functions all have something like (following example is contrived but demonstrates my problem):

f <- function(img)
{
    if ( length(dim(img)) == 2 )
        return( img[1:10,] )
    else
        return( img[1:10,,] )   # Note the extra comma to select all 3D slices?
}

That is, I always have to include a check to say "if it's a two-dimensional array then don't put in the extra comma that indicates "everything in the third dimension"".

Is there some way I can get around this? Since a 2-dimensional array is really a 3-dimensional array with dim(ary) = c(n,m,1), it'd be great to use the same subsetting syntax for both.

Is there some way I can tell R "if I add in one too many commas in the indexing, you should assume that is a singleton dimension" ?

(I suppose I could do the reshaping myself via dim(img) <- c(dim(img),1), but that still requires me to check that length(dim(img))==2 and I'd like know if there's away to avoid this check at the start of every function.

Upvotes: 2

Views: 263

Answers (1)

Patrick Burns
Patrick Burns

Reputation: 897

One approach would be to put a class on your objects -- either grey or colour -- and then the subscripting function for 'grey' objects can just ignore the last dimension. The [ method for the colour objects need not even exist.

Upvotes: 1

Related Questions