Rich Shepard wrote: > On Mon, 8 Mar 2010, Wolf Bergenheim wrote: > > > The 'id' should not be the string 'id', but rather a unique IDentifier > > for each point. Usually you will want to use a running number, so that > > the first is 1 the second is 2 etc... > > Here is an example of 2 lines: > > 1 242856.60 431305.07 242551.88 432149.70 > > 2 242866.60 431325.07 242550.88 432140.70 > > Wolf, > > What you show above is the same two points on both lines. I changed the > file 'nodes' to read: > > 1|242856.60|431305.07 > 2|242551.88|432149.70 > > The command line is: > > GRASS 6.4.0svn (oregonM):/usr4/grassbase > v.net.path in=abernethy_creek > out=stream_length type=line file=/usr4/grassbase/nodes --o > > Yet I still see: > ... > Graph was built > WARNING: Wrong input format: 1|242856.60|431305.07 > WARNING: Wrong input format: 2|242551.88|432149.70 > WARNING: [2] input format errors
The input must match one of the scanf() formats: "%d %lf %lf %lf %lf" "%d %d %d" In both cases, fields should be separated by whitespace. Each record specifies a pair of points. In the first case, the start/end coordinates are given; in the second case, the (integer) category numbers of the start/end points are given. AFAICT, if you want to find the shortest distance between two points, the input should be just: 1 242856.60 431305.07 242551.88 432149.70 -- Glynn Clements <gl...@gclements.plus.com> _______________________________________________ grass-user mailing list grass-user@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/grass-user