DemoDexter stores the streets of San Francisco and the connections between them.
You use the website to add data, and the API to get maps and network analysis out.
This is all very experimental. If you're interested in the project for your own city or projects, grab the source code on Github.
I added a 'library' tag to all of the streets with libraries. Then I went to /access/library to get styles for each street in Carto for TileMill.
You can see the first level of the network mapped in the browser at /embed/TAGNAME
Another Code for America fellow (Jessica Lord) told me that choice was a major factor in urban planning, so I also added a /choice/TAGNAME endpoint. This is a /choice map representing your two nearest Chipotles.
I loaded 15 years of housing data from the City of Macon into another instance of DemoDexter. If you look at individual streets, it's hard to tell the worst streets (those with demolitions) from the rest of the pack. The distribution of these streets is just a dropoff:
When statistics branch out to include connecting streets, you get more than just higher numbers. You get a better idea of a neighborhood. Most housing violations occur in neighborhoods with few demolitions, but houses that get demolished happen in neighborhoods peaking around 10-12 demolished neighbors
These aren't the greatest diagrams, so I'd be happy to share the data with people with better maths.
You can add multiple tags (continuing the example, adding 'demolished' tags for every house demolished on a street)
Then use API urls such as /count/marketstreet/bart to return tags on the street and /networkcount/16thstreet/bart to return a count of tags on connecting streets.
Go to /tags to see what tags exist, and create a new one.
The map will be empty at first. Follow directions to add streets.
The Github Readme has API documentation.
Tweet to @mapmeld once you've got a map started, or if you run into trouble.