Geocoding

This week we spent time learning what network analysis is and how to perform this in ArcGIS.  So essentially I was able to learn what the mechanism behind google maps directions are (though I'm sure it's more complicated than what we dealt with...because of computer languages etc.)

First we had to gather the data starting with TIGER lines (I love the American..er sorry...USG's love of creating acronyms for everything) from the US Census Bureau's website. Well, this was was actually a pain in the ass for me. I live in China and there's this thing they do, where they censor the whole internet including a lot of USG websites. This is not a cool thing (partly why I'm moving out of this repressive country after only a few months of living here). My VPN, for some reason, couldn't connect to the website. After much frustration, I switched to my other VPN which worked (never doubt the simplest solutions to any problem in life).

Once I had the data needed, I followed the lovely pdf instructions and created a Lake_Roads file for Lake County. I then proceeded to add a smattering of data files (new and old alike) to the Lake County database. Many of the files I added, such as the water, county boundaries, etc were from past modules. I then clipped, exported, re-projected, etc my way through them until I had a set of data that I could work with to make the best possible map. I ended up creating a new water features layer by cutting out water features under 1.5 sq km (I never realized how much water Florida has...it can really clutter a map). I wanted to include the water inside lake county, because the many lakes in Lake County had a real impact on the network of roads and I felt it best to ensure my audience is aware of the "lakeyness" of Lake County.

After my data was arranged into feature classes I created the address locator using the EMS file. The matching of unmatched addresses was a little interesting. I matched using the map for all of them, as each time I added said address to google maps they were no where near the ones on the candidate list (those that had candidates anyway). After I matched the unmatched with the assistance of google maps, I then created the Geocoded results.

I picked the three stations you see displayed in the insert and performed the route network analysis creating the nice blue line. I kept the line in both data frames. For the EMS stations, I selected the cross and had them labeled according to their station number. Using the station numbers I could list all of the addresses in the box on the lower right hand corner of the map. I decided to have the stops look the same as the rest, only I kept the color blue. You'll notice I didn't add the stops to the legend. Here I'm assuming the audience can figure out that those are stops from the numbered order of each stop (1...2...3) and from the title of the insert map "Best Route from Station 141 to 241 to 231".  I was trying to decrease clutter, aiming for minimalism. I attempted to make the EMS stations as preeminent as possible by sort of washing the county out with the color scheme while still trying to make sure the street network was visible.

Hope you like it! 

A map displaying EMS stations in Lake County, Fl.
The insert shows the best possible route for 3 of those stations.
The addresses of the stations are displayed based on the station number,
which are labeled on the map it's self.