Off by one when generating indexes

[leonfrench]

A good test case is to give RRHO gene lists of size 100 and set it to step size of 25. The resulting matrix would only make sense to me if it was 3x3 in size. Because it doesn't make sense to overlap the full list to another sublist, nor an empty list to any other list - leaving three comparison sets (first 25, first 50 and first 75).

stepsize <- 25
n <- 100
The current code:
indexes<- expand.grid(i=seq(1,n,by=stepsize), j=seq(1,n,by=stepsize))
Produces 4x4 for this case.

I reckon this makes more sense:
indexes<- expand.grid(i=seq(stepsize,n-stepsize,by=stepsize), j=seq(stepsize,n-stepsize,by=stepsize))

This probably isn't a big problem when the step size is small and n is large.. but it may result in some edge effects.

I'm new to this code and visualization method, so please say if I'm not interpreting it right.

[johnros]

This is a good idea.
It may be a while until I update the package, so if you feel like forking and patching it would be great.
Otherwise, it may take a while.