[Talk-GB] Codepoint duplicates

[Matt Williams]

I've started work using Codepoint Open to perform completeness studies
of OSM postcode data. My plan is to do a rendering like [1] based on
the codepoint data but colour each cell according to whether there's
any matching OSM postcodes.

The problem I have is that there seems to be a large number of
codepoint centroids with exactly the same coordinates (e.g. CV21 1WA,
CV21 1WS, CV21 1YF plus about a few hundred more are all at
-1.2504281683345,52.3798116758619). Are these postcodes which have no
address points and so the OS have just put them all at a random point
'out of the way' or is it some error in the data?

I'm hoping I can ignore them since they mess up the Voronoi diagram.

...

Ahh, looking at the map [2] it seems that they are all located at the
Royal Mail distribution office. My question still stands though: are
these just surplus postcodes or some error? There are still other
cases of duplicates though like can be seen nearby at [3].

-- 
Matt Williams
http://milliams.com

[1] http://random.dev.openstreetmap.org/postcodes/?layers=000F0F0FBT
[2] http://www.raggedred.net/codepoint/?zoom=17&lat=52.37981&lon=-1.25043&layers=B0T
[3] http://www.raggedred.net/codepoint/?zoom=18&lat=52.37459&lon=-1.2629&layers=B0T