Use political polling to build models for predicting market prices on Predictit.com

Use scrape_538.py to scrape recent polling data from 538, as far back as like June? idk exactly. Adjust 'house' or 'senate' at the bottom. Also makes use of rgb_party.csv and predict_party.py to convert the poll coloring to political party.

produces the _senate_polling.csv or _house_polling.csv files.

Use scrape_predictit_all.py to automatically scrape all markets using the URL's in predictit_market_urls.csv, or, use scrape_predictit.py to plug in a single url and scrape that market. Gets the last 30 days.

produces the all_predictit_markets.csv file.

Use market_price_modeling.R to build market price predictions using a lmer model. Does some data manipulation and merges markets and polling together. Uses an estimate for polling error to draw polling from a normal distribution, and simulates market price predictions 250 times to arrive at a set of target markets for the day.

Can do these all at once, or run the daily_execute.py file which calls all 3 of the above and puts the target markets in a .csv.