[Talk-us] More on TIGER: Where it's likely safe to import

[Steve Coast]

Mike

Nice.

Could you use relative object density per little square as a proxy for
population density?

I ask because all the bright green areas near me are forests or old milk
farms. I don’t care. I want to see high density areas and attack those as a
priority since that’s where people are and where they go...

Steve

 *From:* Michal Migurski <mike at teczno.com>
*Sent:* January 3, 2013 7:03 PM
*To:* OpenStreetMap US Talk <talk-us at openstreetmap.org>
*Subject:* Re: [Talk-us] More on TIGER: Where it's likely safe to import

On Dec 17, 2012, at 12:40 PM, Michal Migurski wrote:

> On Dec 16, 2012, at 8:32 PM, Martijn van Exel wrote:
>
>> Have you looked into full history planet parsing to get a fuller
>> picture of editing history? I took a stab at full history user metrics
>> some time ago using osmjs;
>> https://github.com/mvexel/OSMQualityMetrics/blob/master/UserStats.js -
>> this produces one set of metrics for the entire .osh file you feed it
>> but it may prove useful for future work. I haven't touched this in a
>> while but it should still work :/
>
> I have downloaded a copy and given it a beginning look. I'm new to
parsing things of that magnitude; my first thought was to use the full
history file for creations/modifications/deletions on nodes and add that to
what I'm doing already for ways on the osm2pgsql tables. Does that sound
reasonable?

Over the holiday break, I've been grinding through the full history file.
I'll write more about the process and publish some raw data, but the short
version is that I've replaced the Green Means Go tiles with new versions
that incorporate edits to the underlying nodes, address some of the
feedback I've received, and show a slightly different map of the US.

New map:
        http://www.openstreetmap.us/~migurski/green-means-go/

Old map:
        http://www.openstreetmap.us/~migurski/green-means-go-2012-12-16/

Charlotte Wolter and NE2 both pointed out that Florida should see a lot
more post-import editing that I had originally shown, and in fact that's
what the map now shows:

http://www.openstreetmap.us/~migurski/green-means-go/#9/28.3213/-81.6257

http://www.openstreetmap.us/~migurski/green-means-go-2012-12-16/#9/28.3213/-81.6257

Urban fringe areas also show more edits, making them less attractive
targets for blind imports:

http://www.openstreetmap.us/~migurski/green-means-go/#10/37.7707/-122.3451

http://www.openstreetmap.us/~migurski/green-means-go-2012-12-16/#10/37.7707/-122.3451

I've also stripped away the US coastal territory based on NLCD water
designation, per SteveC's suggestion.

These massive edits to counties in Pennsylvania are interesting to me:

http://www.openstreetmap.us/~migurski/green-means-go/#8/40.918/-77.146

-mike.

----------------------------------------------------------------
michal migurski- contact info and pgp key:
sf/ca            http://mike.teczno.com/contact.html





_______________________________________________
Talk-us mailing list
Talk-us at openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20130104/fc638508/attachment.html>