Hi folks,
Scraping is proceeding apace for the postal code to edid file, despite
elections.ca blocking IP addresses :)
To speed things up, I decided it would be interesting to avoid
querying for all FSAs that are entirely contained within an electoral
district. FSAs are 'forward sortation areas', Canada Post parlance for
the first 3 characters of a postal code - A0A to Y1A.
The resulting file can be obtained here:
http://s3.amazonaws.com/danielharan/fsa_edid (small, only 28k!)
Its format is straightforward:
G6C=24034,24036
H9E=24049
219 of the 1392 FSAs only have one corresponding EDID.
Thanks to the anonymous data liberator that obtained the FSA shape
files for me, allowing me to create this derivative product.
d.
PS: for the gis geeks (hoping you can sanity-check my method), here's
the query I used:
SELECT fsa.census_fsa, fed.feduid
FROM fsa, fed
WHERE fed.pruid = fsa.pruid
AND fed.the_geom && fsa.the_geom
AND ST_Overlaps(fed.the_geom, fsa.the_geom)