Hi Claus,

I had hoped the clarification would pretty much steer away from the 1 to 20 k value increment.


Say:

Large value in row 25 is in column Header10

Large value in row 26 is in column Header10

We now need to find a different value for row 26 to prevent Header10 from repeating.

So, the new k value for row 26 cannot be the Large for row 26, it has to be the second Large to avoid returning Header10 again.

Does that make sense?

So we only need a Large(array,1) and a Large(array,2) if my thinking is correct.

So if the a Header repeats the one above it, then look for the Large(array,2) header and return it.

Howard