[GRASS-user] Network analysis

Dear Grass Users/Developpers,

I am studying the distribution of car accidents on a road network. In order to define hotspots and hotzones (succession of hotspots on a road) based on a local density of accidents computed on the network, I use the v.net.alloc function (to allocate a subnet to each accident event). For the moment, the density is thus expressed per “Thiessen” segment. To illustrate, Typical output is similar to

----x—|—x-|-x–|–x----

where
x: is the location of one or many accidents,
-: is a piece of the network,
|: is the limit of the Thiessen segments

My density of accident is simply the number of accident divided by the length of the Thiessen segment. Howerver, as you can notice from the example, the Thiessen segments (|—x-|) are not symmetric relatively to the accident location. I thus would like to compute the length on the right and on the left of each segment. I have tried to cut the output of v.net.alloc with v.edit and coordinates of accident but it do not works. Do you have an idea on how to proceed and which function should I use or implement?

Thanks in advance,
David

On Mon, Aug 17, 2009 at 9:21 AM, David Dabin<david.dabin@gmail.com> wrote:

Dear Grass Users/Developpers,

I am studying the distribution of car accidents on a road network. In order
to define hotspots and hotzones (succession of hotspots on a road) based on
a local density of accidents computed on the network, I use the v.net.alloc
function (to allocate a subnet to each accident event). For the moment, the
density is thus expressed per "Thiessen" segment. To illustrate, Typical
output is similar to

----x---|---x-|-x--|--x----

where
x: is the location of one or many accidents,
-: is a piece of the network,
|: is the limit of the Thiessen segments

My density of accident is simply the number of accident divided by the
length of the Thiessen segment. Howerver, as you can notice from the
example, the Thiessen segments (|---x-|) are not symmetric relatively to the
accident location. I thus would like to compute the length on the right and
on the left of each segment. I have tried to cut the output of v.net.alloc
with v.edit and coordinates of accident but it do not works. Do you have an
idea on how to proceed and which function should I use or implement?

I think that you need a LRS (Linear Referencing System):
http://grass.osgeo.org/grass64/manuals/html64_user/lrs.html
It offers "v.lrs.where" to find line id and real km+offset for given
points in vector map using linear referencing system.

Please keep us in the loop how it goes. And don't hesitate to improve
things or to suggest better documentation.

Best
Markus

David Dabin ha scritto:

Dear Grass Users/Developpers,

I am studying the distribution of car accidents on a road network. In order to define hotspots and hotzones (succession of hotspots on a road) based on a local density of accidents computed on the network, I use the v.net.alloc function (to allocate a subnet to each accident event). For the moment, the density is thus expressed per "Thiessen" segment. To illustrate, Typical output is similar to

----x---|---x-|-x--|--x----

where
   x: is the location of one or many accidents,
   -: is a piece of the network,
   |: is the limit of the Thiessen segments

My density of accident is simply the number of accident divided by the length of the Thiessen segment. Howerver, as you can notice from the example, the Thiessen segments (|---x-|) are not symmetric relatively to the accident location. I thus would like to compute the length on the right and on the left of each segment. I have tried to cut the output of v.net.alloc with v.edit and coordinates of accident but it do not works. Do you have an idea on how to proceed and which function should I use or implement?

Thanks in advance,
David

------------------------------------------------------------------------

_______________________________________________
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

Not responding to your question...

see also v.kernel [1] > vector density map on vector network with a 1D kernel

[1] http://grass.itc.it/grass64/manuals/html64_user/v.kernel.html

giovanni