[GRASSLIST:4470] RE:Two questions: 1. r.param.scale; usage and etc... an error got into my last post!!

[Markus Neteler]

On Mon, Sep 09, 2002 at 10:47:53PM +0200, Lars Forseth wrote:
> Hi! Hav to repost due to an error that sneaked into my first post; sorry for 
> the inconvenience!
> 
> I have just - this weekend - installed GRASS5.0.0; on my SuSE 7.3 box. only 
> problem was with NVIZ and libtcl.so.0 ; "ln libtcl.so libtcl.so.0" solved 
> that. Congratulations to the GRASS team, well done :-)
> 
> So my questions;
> 1. In their book "Open source GIS" Neteler&Mitasova on page 114 gives an 
> example of the usage of r.param.scale. This gives - with Spearfish data 
> (elevation.dem) - not a very meaningfull result. Sent to a monitor the 
> resulting datasett gives me a white rectangle covering the region. If I 
> change the parameter size=7 to 3 I can see some small cells with mostly blue 
> colors. Param=feature. What is wrong; if anything? (I have checked the syntax 
> and this is  the result every time). 

I think that r.param.scale was updated *after* we wrote that section.
That means that the default values are different, maybe leading to bad
results with integer DEM. I will investigate after GRASS conference.

> (By the way a good book :-)!)
> 
> 2. Is there a way you can to a "density" analysis a la ArcView 
> SpatialAnalyst; with sites (i e point) data? How?

The s.kernel is close to be released. It seems to do it's job :-)
But: could you write in detail what SpatialAnalyst is doing when doing
"density" analysis?

Markus