[GRASS-user] Vector proximity Tools.

[Sampson, David]

Hey Folks,

I come back with more questions for "addressing" and geo locations
issues... 

Looking for a tool that can take a vector file (road segment network)
and point data (FSA Centroids) to relate a road segment with the closest
FSA(s). FSA's are the first three letters of a Canadian Postal Code

SI there a tool in GRASS to check this proximity with VECTOR?  Or do we
have to convert to raster and then back to vector?

Here is the background

1. using PAGC as a program that can do addressing based on blok face
road segments and address ranges (http://www.pagcgeo.org/)
2. Using the freely availbe road network file from Stats Canada
(http://www.statcan.ca/english/freepub/92-500-GIE/2005001/how.htm)
3. Since postal code position data is not freely available in Canada you
can only realy work with the first three characters (for free at least),
or the Forward Sorting Address (FSA)
4. The FSA list can be found in various places, I like the one on
Wikipedia, although a full QC has not been done
(http://en.wikipedia.org/wiki/List_of_K_Postal_Codes_of_Canada)
5. Geoconnections has a gazateer service that will give out a centroid
based on the FSA (one at a time of course)
http://www.geoconnections.org/CGDI.cfm/fuseaction/developersCorner.detai
ls/id/4517/gcs.cfm

The problem is that in the road files there may be more than one MAIN
STREET. So I'd like another criteria to match against... in the states
the tiger data is good for zip codes... but postal codes in canada,
well, cost a lot.  So I'm thinking of the following rationelle

First, I need to create a FSA centroids shape file (that will later be
freely distributed) by running a list of FSA's for a province through
the gazateer.

Then here is where I think GRASS might be able to come in.

Match each roadsegment to the closest FSA centroid(s). Maybe the top
three. Have this info dumped to the table. Export the new road file as a
shp file and use this as reference data for PAGC. This would at least
help narrow down the choices as it is unlikely to have two MAIN STREETS
in one postal code.

This is just an idea... so any input or LIKE experience would be welcome
info

Cheers
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/grass-user/attachments/20060901/97b731f4/attachment.html