Reworked NBA Predictions (Version 4- in Python)
I've decided to rework my NBA Prediction code from R to Python. Mostly to see if I could do it, and also to see if I could speed it up a bit. I'll update here with current speed/accuracy results as the 2014-15 season plays out. The structure and format is pretty much the same, with the exception that it's cleaner code. I still need to comment it a bit more, but it's Git ready for now.
-
NBA_Driver.py - This is the workhorse, the script that actually gets run. It will call the webscrapers, genetic functions, and create the data/logging as it runs.
-
NBA_Tweeter.py - This is the script that tweets the top (N/2) games for the day to twitter. If you use this script, you'll have to change the OAUTH/Secret Key information to access your twitter feed.
-
GA_Funs.py - The genetic algorithm functions are kept here. (Create Population, Get Fitness, Get Schedule Stats, Mutate Individual, Create Child)
-
Webstats_Funs.py - Houses all the webscrapping functions to get the schedule, outcomes, player and team statistics, injury information, and online betting odds.
TBD
TBD
TBD