[GRASSLIST:9505] Re: dem from vector files...help...

[Māris Nartišs]

Hi,

I had similar problem. My approach was:
1) Import *.shp to vects;
2) As I had points and lines, I converted lines to points with v.to.points;
3) Merge them with v.append (You will find it in GRASS Wiki page). Its
slow, but works;
4) Get DEM with v.surf.rst; Search list archive for links about RST tuning.
5) Validate it. I.e. with r.terraflow generate rivers and compare with
rivers on map ;)


Hope it helps,
Maris.

2005/12/15, Hamish <hamish_nospam at yahoo.com>:
> > I'm harris, i'm a student and I'm a new user of Grassgis 6.0. I
> > started using Grass because of an essay and i have to finish it to get
> > my degree...So i have to make a digital elevation model from some
> > vector files.The vectors are containing elevation info and was
> > digitized to arcmap.So i followed the following method : I imported
> > the ersi .shp,I had 4  shp files so i made  4 vector files.Then I
> > combined the 4 vector files to make them one. i used the v.patch
> > command. Then i think that i used v.clean to snap the common vector
> > lines but i not so sure if that was neccessary.was it? what other
> > steps i must do? Then i tried to connect from another esri shp (that
> > was containing all the vectors together from the begging) the dbf .i
> > managed to connect the tables. My next step was to make vector to
> > raster so i used v.to.rast. an my command was
> >  v.to.rast input=string [layer=1] output=string [use=att]
> >  [column=elevation (the column that contains elevation info iin my
> > case.) is this correct? When i did that i finally used v.surf.contour
> > to make my dem. So i'm waiting for hours for my dem to be done but
> > it's procceding very slowly.
> > Is that the right way to combine vectors?Cause the first time i tried
> > to make my dem throught that way i got a very bad dem...V.patch
> > combines the dbf of each vector?i think that it isn't.
> >
> > That's all, i not so sure if i did the right steps in general and
> > that's why i need your help anyway.If anyone could help me id be very
> > grateful.
>
>
> That all seems ok. (except the module is called r.surf.contour)
>
> One change I might make: don't bother with v.patch, just do v.to.rast on
> the four component vectors and then use r.patch to combine the parts.
>
> r.surf.contour will take a long time if there are large gaps without
> contours (e.g. a flood plain) as for each cell it has to look a long way
> for data. Rather inefficient algorithm, maybe r.cost's approach could help?
> Let it run over night. It will get there in the end.
>
>
>
> Hamish
>
>