Comments on: What’s easy, and what’s hard, about getting from Excel to a GeoRSS-enabled mashup

By: rajsingh.org blog » Check out John Udell

rajsingh.org blog » Check out John Udell — Sat, 11 Aug 2007 05:04:26 +0000

[...] What’s easy, and what’s hard, about getting from Excel to a GeoRSS-enabled mashup [...]

By: Excel geocoding adventures « Jon Udell

Excel geocoding adventures « Jon Udell — Fri, 10 Aug 2007 18:38:42 +0000

[...] geocoding adventures Filed under: Uncategorized — Jon Udell @ 1:38 pm As mentioned here, I’ve been working with a spreadsheet containing addresses that want to be geocoded. [...]

By: A geographic analysis of local crime data « Jon Udell

A geographic analysis of local crime data « Jon Udell — Wed, 01 Aug 2007 16:34:46 +0000

[...] If you’ve been following the continuing saga of my exploration of local crime statistics [1, 2, 3, 4], here’s an update. The police department has provided a spreadsheet containing a [...]

By: Nigel Snoad

Nigel Snoad — Tue, 24 Jul 2007 14:14:38 +0000

Jon,

I’ve been playing with this for a while, and started coding a bunch of Popfly blocks to do the data mapping and transformation. While you and I could enter regexps into the blocks to get the outputs we want, we can pretty easily build transformation editor blocks and so forth. Pipe block to block and then to the Popfly VE block.

We’re getting there, slowly getting there.

By: Paul Daniel

Paul Daniel — Fri, 20 Jul 2007 19:13:52 +0000

I took your incidents.xml file and processed it through Yahoo Pipes. I ignored the lat/long data you supplied just to see if Pipes and Yahoo’s geocoding api could make sense of the address (“Keene” AND “New Hampshire” hardcoded).
http://pipes.yahoo.com/pipes/pipe.info?_id=bEi72dM23BGmLPTj1vC6Jw
I don’t know how accurate and inclusive the results are, although the results I did check on the map looked pretty accurate.
Pipes could possibly take the process a step further back, since it accepts csv input (I’ve not tried this yet). Alternatively, it would take the published rss feed of a Google spreadsheet

By: Brian Dorsey

Brian Dorsey — Thu, 19 Jul 2007 23:39:00 +0000

GeoPY might come in handy for the actual geocoding of addresses:

http://exogen.case.edu/projects/geopy/

“geopy makes it easy for developers to locate the coordinates of addresses, cities, countries, and landmarks across the globe using third-party geocoders and other sources of data, such as wikis.”

See the four line examples on that page for geocoding using google, yahoo or virtual earth. There are also some useful functions which make it easy to find the distance in miles/km between two lat/long points.

Take care,
-Brian

By: Jeff

Jeff — Thu, 19 Jul 2007 21:14:54 +0000

Hi Jon… I also wonder how using geoRSS (or KML) this way scales for a high volume of points. Google Maps lets you “search” for a KML file (putting it in the querystring) and the features in it will be automatically be mapped. But suppose a KML/geoRSS file has thousands of elements in it or is a multi-MB file… it’s inefficient to load all that (esp. into the client) if the user’s current zoom/pan state only makes a fraction of them visible or relevant. It seems managing a high number of points/lines/regions dynamically as the user navigates will be a key concern for all but the most trivial of samples.

Anyway, your current writing on public data (and the geocoding of it) is very interesting to follow. Thanks!

By: cmarnold

cmarnold — Thu, 19 Jul 2007 19:03:24 +0000

Jon–

Take a look at the publicly available police log format from my local PD (http://www.barremasspd.com/police_log.php). The PDF publication *should* lend itself well to being transformed to XML.

-CMA

By: Dan

Dan — Thu, 19 Jul 2007 18:37:24 +0000

Commercial tools are available to do this, such as Trillium from Harte-Hanks. The software is expensive and requires frequent data updates to stay current with US and foreign addresses.