[GRASS5] Re: Vector format proposal

Radim Blazek Radim.Blazek at dhv.cz
Wed May 16 01:54:45 EDT 2001


Eric G. Miller wrote:
> On Tue, May 15, 2001 at 08:48:27AM +0200, Radim Blazek wrote:
> > USER_ID is category in grass and AREA_ID is sequential number assigned
> > during build process. Link to table may not be broken. Lib is written in
> > this way.
>
> Okay,
>    Forgive me if I'm being dense. But, how do you join an external table
> to select area(s) with category "2" when there are also lines and points
> with category "2".  Do we join all then filter out the lines and points?
>
> So we have something like:
>
> 1 : 2
> 2 : 2
> 3 : 2

Yes, that was my idea and that is how grass50 works
(v.reclass has type= option for type selection).

As I mentioned in some other mail, it may be useful to have more types
(lines, points, areas) linked to one table, so I would prefer don't
distinguish types in db connection.

I see following schemes:
(according to vector proposal I will use 'field' for distinguishing 
categories)
- all elements (various types) are the same logical feaure (water for
   example: lakes, rivers, spring - area, line, point with many attributes
   for water quality), all categories are of one field (1 for example)
   and linked to the same table.
   We can work with all types at the same time or select one type.
- map contains just one type
- map contains more features each of one type (for example lakes as
   areas with field=1 and rivers - lines with field=2), each feature linked
   to its own table. The selection is done by field= option

Note: using field number is not user friendly and I expect some place
        where names will be assigned to fields so that  running module
        would look like:
v.xx input=wat output=riv1 field=rivers where="q > 10 and n > 0.03"


Radim


---------------------------------------- 
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo at geog.uni-hannover.de with
subject 'unsubscribe grass5'



More information about the grass-dev mailing list