My new maps bring all the boys to the yard

[Adam Griffin]

I'm facing off against the final sperg boss! Recently, I decided to look into how to build thematic maps using a system Google offers that can compile shapefile data/KML files along with spreadsheets. After about three hours of tinkering, I more or less figure out how to do this. I think it's pretty awesome and there are certainly lots of uses for this. Unfortunately, there are also some limitations when comparing them to the current maps available via Atlas, such as:

• the inability to separate legend to show multiple candidates; I've been picking the winning candidate in various races and building the legend's colors around that to reflect two-party system performance
• the lack of a proper % display in the legend; it simply ranks from "0" to "1" (0%-100%)
• having to click on each area/the lack of an easy solution to provide hover data for each precinct/county/state; I've read that this is possible, but requires a bit more work that I've yet to learn
• the current color scheme (reversed atlas colors but the same colors as used here) displays purple (very close jurisdictions) and dark red (heavy Republican areas) areas as quite similar visually
• data in spreadsheets typically has to be reorganized to some degree to be compatible; for instance, The Guardian's major spreadsheet for the 2012 presidential election is compiled oddly and the FIPS data doesn't correlate exactly with the KML file I have, meaning it'll take a lot of work to get it operating unless I can find another source

Anyway, feel free to take a look at the sample maps I've created thus far (most are of various GA races from 2010 & 2012) and let me know what you think. Also, anyone with organized spreadsheet data can feel free to ask for a map, assuming I can track down shapefile data for the jurisdiction.

[Flake]

I was surprised at the number of voters registered near Atlanta. I knew that would be where the most votes in the state are located, but the sudden drop in numbers in the outer counties was surprising, to say the least.

[muon2]

Excellent work. As a quick comment, you might play with the shade for CPVI 4. It looks more Pub than CPVI 5. For example compare IL 14, 16 and 18. IL-16 seems to be shaded more towards 18 when it is more D than 14. I suspect that the blue tinting isn't coming through very well there.

[Adam Griffin]

Excellent work. As a quick comment, you might play with the shade for CPVI 4. It looks more Pub than CPVI 5. For example compare IL 14, 16 and 18. IL-16 seems to be shaded more towards 18 when it is more D than 14. I suspect that the blue tinting isn't coming through very well there.

Yeah, I see that. The main issue (one of the ones I mentioned above) has to do with using the Atlas blue/red colors. Unlike the maps on here, the colors on the legend are only displayed as a reference and used to create the gradient. Most colors displayed on there fall in between two of those displayed colors in the legend. Hence the inclusion of purple for those that fall around 50/50, as opposed to being in the "40/50% R" or "40/50% D" shades that we're so used to on here.

Purple has a tendency to look darker than some shades of red and causes areas that would be in the low-50s on either side to look like stronger Republican/Democratic areas, making the areas you mentioned look out of sync. Once it gets to around 54% R / R+5, it begins to correct. I may have to consider using an entirely different color scheme to correct this.

[Miles]

Awesome!