[GRASS-user] Re: r.resamp.stats max method not working as expected

[Glynn Clements]

Glynn Clements wrote:

> > Okay, I did find verification that this doesn't work in r.resamp.stats:
> > 
> > *[1] The min and max aggregates can't use weights, so -w has no effect for
> > those.*
> 
> Hmm. There's no fundamental reason why I can't add "weighted" versions
> of the min/max aggregates. Obviously, these would ignore the weights,
> but it would allow r.resamp.stats to include the source cell in the
> calculation for each destination cell which overlaps it.

Implemented in r33888 (in 7.0).

The aforementioned comment in the Wiki still stands. min/max aren't
actually weighted, so the behaviour of those aggregates with a
perfectly aligned grid is "unstable" (i.e. the exact result may depend
upon miniscule rounding errors) even when -w is used.

-- 
Glynn Clements <glynn at gclements.plus.com>