[GRASS-dev] Re: [GRASS-user] Resampling with percent cover

GRASS GRASS developer
1

on grass-user Glynn wrote:

Create a reclass of the input such that shrub=1, non-shrub=0 and
cloud=null, then resample that with "r.[resamp.]stats -n method=average ...".

does anyone feel up to adding method=trimmean (+ a trim= option) to
r.resamp.stats (see r.in.xyz) and/or percentile= (see r.univar) instead
of just fixed median,q1,q3,perc90 levels?

Hamish

2

Hamish wrote:

> Create a reclass of the input such that shrub=1, non-shrub=0 and
> cloud=null, then resample that with "r.[resamp.]stats -n method=average ...".

does anyone feel up to adding method=trimmean (+ a trim= option) to
r.resamp.stats (see r.in.xyz) and/or percentile= (see r.univar) instead
of just fixed median,q1,q3,perc90 levels?

Rather than hack this in with a global variable, I'll modify the stats
library functions to take an additional void* argument.

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

3

Glynn Clements wrote:

> > Create a reclass of the input such that shrub=1, non-shrub=0 and
> > cloud=null, then resample that with "r.[resamp.]stats -n method=average ...".
>
> does anyone feel up to adding method=trimmean (+ a trim= option) to
> r.resamp.stats (see r.in.xyz) and/or percentile= (see r.univar) instead
> of just fixed median,q1,q3,perc90 levels?

Rather than hack this in with a global variable, I'll modify the stats
library functions to take an additional void* argument.

In r36918, aggregate functions take an additional argument, and
r.resamp.stats, r.series and r.neighbors support method=quantile and a
quantile= argument.

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