election simulation data project
This repo contains files for an electiom simulation data science project I used to teach myself Python and simulating outcomes in general. I received most of the information from the 'Desperately Seeking Silver' homework and tutorial. They are retroactive simulations/analyses for the 2012 US Presidential election, but the methods could be applied to future elections as well.
The baic simulations use the percentage of money donated to the final two candidates and demographic information to calculate the probability of a candidate winning a particular state. Use of the R package BreakoutDetection was also used to determine when there were especially large spikes in net cash flow, expenditures, or donations. While the initial simulations were implemented with combination of Python and R, a parallel implementation was also performed in Julia.
In order to create more sophisticated models, ongoing projects will use a combination of polling data, contribution information and weighting importance by date. These will be updated periodically (meaning, as I get to it).