[GRASS-user] V.kernel

[Richard Chirgwin]

Hamish wrote:
> Richard wrote:
>> I would like to come up with a "hard" expression of point density -
>> X per sq km - for my points map, and I'm not sure whether v.kernel
>> would do so.
> 
> If by hard you mean no smoothing, extrapolation, etc, just what the data
> points say, you might try 'v.out.ascii | r.in.xyz method=n' after setting
> up an appropriate grid size with g.region.
> 

The attraction of v.kernel may be expressed in the answer to this question: how
many 10 km grid squares exist in Australia? Stepping through square-by-square
might be slow; so I thought if I could relate the raster values directly to
original points map, it would get where I want faster.

I suppose if I use a non-lat-long location, with resolution at 1m, it's easy to
step a script by 10000 metres each iteration...

Richard

> maybe a r.neighbors run could smooth things over.
> 
> 
> I don't know much about using v.kernel beyond the basics, so no advice
> there.
> 
> 
> Hamish
> 
> 
> 
>       
> 
>