The problems reported by Daniele seem to be resolved. I guessed (right)
that somewhere in the code a '<' should have been a '<=', or something
similar. It's on my TODO list now.
--
Edzer
Daniele Iannuzzo wrote:
>
> - You (that was me) wrote:
>
> - Could you change this into 0 +/- 89 for <x,y> and 0+/-0.1 for <z>,
> and see whether things change?
>
> Thank you again, Edzer, for your help!
>
> Your intuition was good: with your advice, things deeply change. I
> send you the number of point pairs in the first five intervals with
> three different angular tolerances:
>
> <x,y> 0+/-90 <z> 0+/-0 (19,134,127,153,94);
>
> <x,y> 0+/-89 <z> 0+/-0 (8,83,56,79,46);
>
> <x,y> 0+/-89 <z> 0+/-0.1 (19,129,125,153,94);
>
> of course the little difference between the first and the third
> variogram is due to few points lying exactly on an east-west
> direction. The distance between the layers is enough to be sure that
> such a vertical tolerance couldn't add any point pairs.
>
> Do you suggest I calculate variograms with a minimum of vertical
> tolerance everyway, even if I need 'layer' variograms?
>
> Thank you again for your unreplaceable help.
>
> Daniele
