[GRASSLIST:3128] Re: v.support and general vector question

[Michel Wurtz]

"I.T.H. Deyzel" wrote:

> I've imported a vector catchment file from .e00 using m.in.e00. The file
> is imported succesfully but I can't seem to get the attributes and
> labels to correspond. I did copy the 'catch.catnum' file as 'catch' so
> that the category file correspond to the vector's name. If I query the
> vector field I get the incorrect label information. I've tried importing
> with m.in.e00 -i option, but to no avail. Have anyone experienced such
> difficulties using m.in.e00? It might be that the .e00 encoding differs
> a bit from what m.in.e00 expects or else I'm missing something.

The labels are attached to lines or areas by v.support (wich runs
automaticaly with the -s option). The category file is just a table
that gives an attribute value to each label value.

The -i option doesn't help much in this case.
 
> I've been trying to fix the dig_att and dig_cats files manually so that
> they correspond to reality (what a tedious task!), but when running
> v.support I get errors unknown to me. Please help.

A better way is to use v.reclass after editing the categoy file in a way
it can used by v.reclass (delete header lines, and use a colon to separate
label and category value)

> * the error : PNT_TO_AREA failed: (29.150000, -22.210000) (Category 2) -
> does it indicate that the polygon is open and that is unable to label
> it?

yes, or that the label point is outside a closed line...

> * WARNING: area 20 label: 17 matched another label: 22. - is the label
> positions to close together? How can I fix this.

the two label points are in the same polygon.  (it seems that v.support is
sometimes confused and made some mistakes...).

A good way too look at what is happening is to use the output of v.reclass
for creating a site file. After that, you can display the "missed" labels on
top of your vector file.  You can also transform the dig_att file (area labels)
in a site file, so you can see if there is really more than one label in a
polygon...

In any case, m.in.e00 cannot clean a bad e00 file (they are not allways perfect)

--
Michel WURTZ - DIG - Maison de la télédétection
               500, rue J.F. Breton
               34093 MONTPELLIER Cedex 5