bcongdon / pyvethirtyeight

:us: Python wrapper for Fivethirtyeight's Election Forecasts

PyveThirtyEight

A Python wrapper for the 2016 FiveThirtyEight Presidential Election Forecast

Installation

pip install pyvethirtyeight

Usage

from pyvethirtyeight import President, Senate

# Presidential Election Forecasts
f = President()

# Latest Forecasts (Unordered)
latest = f.latest_forecasts()
latest[0].party # 'D'

# Current Leader (returns Forecast of latest leader)
leader = f.current_leader()
print(leader.models['plus']['winprob']) # 83.99

# Current Leader by Model
leader = f.current_leader(model='now')
print(leader.models['now']['winprob']) # 90.4

# All Forecasts
all_forecasts = f.all_forecasts()

# Same methods also work with Senate predictions
s = Senate()
print(s.latest_forecasts()[0].models['polls']['winprob']) # 67.33596

The Forecast Object

A Forecast is the prediction of the FiveThirtyEight model for a particular candidate at a particular moment in time

A Forecast has keys:

- date (datetime) -> Forecast date
- party (str) -> Candidate's party initial
- candidate (str) -> Candidate's name
- models (Dict) -> Dict of model predictions associated with candidate