[GRASS-user] v.net.alloc having difficulty with specific dataset

[Markus Metz]

On Fri, Oct 7, 2016 at 4:08 PM, Sam Snellings <ssnellings at gmail.com> wrote:
> Fantastic, that worked perfectly. Thanks for taking a look at the data for
> me. It didn't even occur to me that the data might have a third dimension.
>
> Is there an easy way to check whether a geometry/shapefile has a
> Z-coordinate element in case I run into a similar issue in the future? I can
> just test import as 2d as you recommend (or QGIS has an export function that
> removes z-coordinates), but perhaps there is an easier way to tell?

the easiest way to tell if layer in a OGR-recognized datasource is 3D
is to use ogrinfo -so <datasource_name> <layer>. For the data in
question, it reports with gdal-2.1.1 (amongst other info)
Geometry: 3D Line String

Markus M

>
> -Sam
>
> On Thu, Oct 6, 2016 at 3:32 PM, Markus Metz <markus.metz.giswork at gmail.com>
> wrote:
>>
>> On Wed, Oct 5, 2016 at 3:05 AM, Sam Snellings <ssnellings at gmail.com>
>> wrote:
>> > Markus,
>> >
>> > Thank you for the suggestions. I ran the following commands nine times
>> > on
>> > the data, with thresholds .00000000000000000001, 0.0000000001,
>> > 0.0000001,
>> > 0.00001, 0.01, 0.1, 1, 10, and 100. Let me know if I'm not going small
>> > enough.
>> >
>> > v.clean --overwrite input=roads output=roads_snap -c tool=snap
>> > thresh=[see
>> > above]
>> > v.net --overwrite input=roads_snap points=schools output=network
>> > operation=connect thresh=1000
>> > v.net.alloc --overwrite input=network output=network_allocation
>> > ccats=1-2
>> >
>> > I did not get a complete allocation of road segments from any of them,
>> > and
>> > as far as I can tell it got stuck at the same vertices on each run.
>> >
>> > I will email you a link to the data off-list, thanks for taking a look.
>>
>> thanks for providing a link to the data! The problem is that the roads
>> are 3D with z coordinates not matching at the end of a road segment
>> and the beginning of another road segment. The solution is to import
>> roads as a 2D vector with v.in.ogr -2. 2D snapping with v.clean
>> tool=snap thresh =1 might be helpful.
>>
>> Markus M
>>
>> >
>> > -Sam
>> >
>> > On Tue, Oct 4, 2016 at 3:53 PM, Markus Metz
>> > <markus.metz.giswork at gmail.com>
>> > wrote:
>> >>
>> >> On Tue, Oct 4, 2016 at 5:09 PM, Sam Snellings <ssnellings at gmail.com>
>> >> wrote:
>> >> > I've been running some trials through v.net.alloc using town road
>> >> > networks
>> >> > as the arc layer and schools as the node layer. I have one set of
>> >> > roads/schools that breezes through v.net.alloc without issue (Town
>> >> > 1),
>> >> > while
>> >> > another one only allocates the first few arcs connected to each
>> >> > school
>> >> > and
>> >> > leaves the rest of the network unallocated (Town 2). There is
>> >> > probably a
>> >> > problem with the data from Town 2, but I can't figure out what it is.
>> >> > Does
>> >> > this issue sound familiar to anyone?
>> >> >
>> >> > I am using the GUI. I load each pair of shapefiles into GRASS and
>> >> > then
>> >> > run
>> >> > the following statements (I only import two school points to
>> >> > simplify).
>> >> > The
>> >> > two towns are pulled right from the state-wide GIS office, but the
>> >> > towns
>> >> > are
>> >> > in different states (ie - different agencies provide the data).
>> >> >
>> >> You do not need this for v.net.alloc:
>> >> > v.net input=roads output=road_nodes operation=nodes
>> >> it will create a point for every node, the first point will have
>> >> cat=1, the second cat=2 which collides with the points you want to
>> >> attach
>> >>
>> >> > v.net input=road_nodes points=schools output=network
>> >> > operation=connect
>> >> > thresh=1000
>> >>
>> >> rather use
>> >> v.net input=roads points=schools output=network operation=connect
>> >> thresh=1000
>> >>
>> >> > v.net.alloc input=network output=network_allocation ccats=1-2
>> >>
>> >> the real problem is probably a data problem: roads are not properly
>> >> connected, i.e. the end of one road does not match the beginning of
>> >> another road. Try v.clean input=roads output=roads_snap -c tool=snap
>> >> with a very small threshold. This kind of error is quite common with
>> >> shapefiles (not a fault of the shapefile format but of the software
>> >> used to create the shapefiles).
>> >>
>> >> >
>> >> > Thanks in advance for any advice, I can also email images/data if
>> >> > that
>> >> > would
>> >> > be helpful.
>> >>
>> >> If v.clean -c tool=snap does not help, (links to) original data of the
>> >> schools and road network for Town 2 would be helpful (data please only
>> >> off-list).
>> >>
>> >> Markus M
>> >
>> >
>> >
>> > _______________________________________________
>> > grass-user mailing list
>> > grass-user at lists.osgeo.org
>> > http://lists.osgeo.org/mailman/listinfo/grass-user
>
>