into-the-wild

Maximum entropy models applied to voting data. This project applies a maximum entropy Ising model to voting data from the Supreme Court and the U.S. Senate. It begins by recreating a paper published by Bialik et al. (2017) on Supreme Court voting patterns. It then extends that analysis to the U.S. Senate using the pseudolikelihood and mean field theory methods described by Lee and Daniels (2018). The SCOTUS recreation code is in the "general" directory, the mean field theory code is in the "mft" directory, and the pseudolikelihood code is in the "congress" directory.