Pretty Pictures
Sheldon Cooper had "Fun with Flags" but today I had a lot of fun with maps instead! The capabilities with R really are quite fantastic (though I will always want to manipulate the actual data in Python!) and I've been having some fun.
This is the beginning of my work for the first major project for Term 2 of DAND. (Oh, super exciting new, Udacity has announced their new Data Science Nanodegree and this is definitely my next port of call!) I will end up providing a full write up of the info, but for now, I'm a bit excited about what I was working on today.
2016 Election Contributions - Ohio
I'm going to be taking a look at what contributions in the bellwether state of Ohio tell us about the voter behaviour that could have given us better insight into the results of the presidential election.
I find the above quite interesting as it shows quite clear support for candidates for the Republican nominees for the presidential election.
Ohioans made substantially more, and larger contributions to the Republican candidates during their primary campaigns than they did for the Democratic primary campaigns.
This might have been influenced by the fact that there was an incumbant Democrat and there were much more quickly fewer options for the Democratic nominee, but what is particularly interesting is how it compares to the results of the General Election contributions.
It's also quite fun that by plotting the lat and long of points as a scatterplot you can create the general mapping of a place.
Before I zoomed in on the state, the dots created an almost solid area of what is Ohio on my screen, and with the zoom in, you can see the clusters of the major urban areas in Ohio.
Once the election results came out - and technically much earlier, as it became very clear early on that Trump was likely to win - one of the large questions was, "How did the pundits get it so wrong?" I'll need to play with things a bit more to provide have any chance of being able to state confidently, but I do wonder if this contrast in the pictures does provide some insight. That is, if you looked at the pattern of General Election contributions in Ohio, they show substantial support for Clinton (who was the Democratic candidate). Perhaps this led to an overestimation for her levels of support, and perhaps the information from the primaries got lost in the mix.