[Aust-NZ] NZ Towns

[Peter]

Its an almagamation of the sources i named, geonames, linz electoral, and google geocoder for the coords, with a bit of stuff patched in (when i had any doubts) from statnz interactive boundary viewer (ambiguous regions, some populations). Where spelling disagreed i tended to go with the electoral dataset http://www.linz.govt.nz/placenames/electoral-streets-places/ Just beware there was a lot of hand editing involved, hence errors are a real possibility.

QGIS: i tried both installers, both had the same error. It wasnt the Program Files thing becasue i put it in Program-Files the second time to be safe. As far as i got in the documentation suggested it was the absence of a windows library. I ran out of time and patience. Given your guys recommendation ill try it again soon. Failing that ill try it on a linux box, thatd probably, yeah, be easier... The thought of all that FOSS sql, python, etc running on windows gives me the willys.

METADATA: Yes, No. The reason being the same as why i was reliant on such half baked sources, lack of GIS knowhow.

Peter


pcreso at pcreso.com wrote:
> Hi Peter,
>
> Thanks for that, a useful dataset. Before I use it, can you give me some
> idea of where it was derived from, & what licence the content is
> available under? If there are several sources, then identifying the
> source of each record would be useful, as most licences have an
> attribution requirement.
>
> & QGIS is worth persevering with & have a look at Postgis for spatial
> data management :-)
>
> Thanks,
>
> Brent Wood
>
>
> --- On *Mon, 11/15/10, Peter /<webwiz at pl.net>/* wrote:
>
>
>     From: Peter <webwiz at pl.net>
>     Subject: Re: [Aust-NZ] NZ Towns
>     To: aust-nz at lists.osgeo.org
>     Date: Monday, November 15, 2010, 2:01 PM
>
>     Indeed, i eventually concluded that the secondary sources that are
>     freely available are all unreliable. The geographic placename
>     register from linz is two years out of date, and its regions are not
>     TLA regions but land districts. LINZ also has an up to date
>     electoral districts file, which i was able to use to get tla
>     districts from for some 6000 odd places. The geonames dataset is
>     based on the geographic placenames dataset, and has the region field
>     mostly empty. It has partial population fields, ( about a quarter
>     maybe), and approx coords. I found the google geocoder coords to be
>     quite a bit more accurate. I suspect the geonames coords are
>     someones attempts to shift the linz place label coords by half the
>     label width or something.
>
>     Anyway the only real solution is as you say is go back to the
>     primary GIS data. Attempts to bodge together old secondary sources
>     only introduces yet more noise into the dataset.
>
>     FWIW my excel spreadhsheet is here if its any use to anyone.
>     http://www.zoneblue.org/files/nz-towns.zip
>
>     Which served my immediate purpose, but i do need to get more
>     organised GIS wise.
>
>     I have no real fondness for large java or .NET frameworks which are
>     needed for the main GIS packages. As for open source, QGIS would not
>     install, Grass did (brought a bucket load of stuff with it too), so
>     i guess ill try learning it.
>
>     Thanks for your help guys.
>
>     Peter
>
>
>
>
>
>
>
>     Gavin Treadgold wrote:
>      > On 2010-11-12, at 13:16 , Robert Coup wrote:
>      >> On Fri, Nov 12, 2010 at 11:12 AM, Peter <webwiz at pl.net
>     </mc/compose?to=webwiz at pl.net>
>      >> <mailto:webwiz at pl.net </mc/compose?to=webwiz at pl.net>>> wrote:
>      >>
>      >> I have blown an inordinate amount of time trying to merge various
>      >> online sources to get a listing of towns in NZ. Ideally this list
>      >> would have town , district/city/unitary council, region, and
>      >> population or area. This would allow me filter by towns above
>      >> certain sizes for various web app uses.
>      >> GPS coords would also be good but i can get those from google
>      >> geocoder.
>      >>
>      >> Getting area or boundaries will be harder. Possibly the Fire Service
>      >> suburbs dataset (available for $0 but under onerous terms) will
>     help,
>      >> or look at meshblocks/area-units and their populations from the
>     census
>      >> database and try and derive from that?
>      >
>      > Much of the boundaries you should be able to get from Statistics New
>      > Zealands digital boundaries.
>      >
>      >
>     <http://www.stats.co.nz/browse_for_stats/people_and_communities/geographic-areas/download-digital-boundaries.aspx>
>      >
>      > Cheers Gav
>      >
>
>     _______________________________________________
>     Aust-NZ mailing list
>     Aust-NZ at lists.osgeo.org </mc/compose?to=Aust-NZ at lists.osgeo.org>
>     http://lists.osgeo.org/mailman/listinfo/aust-nz
>