search

pmeyers279 / basketball_db

package to query basketball-reference.com and save data.
http://basketball-db.readthedocs.io/en/latest/
0 stars 2 forks source link

readme

basketball_db

Build Status

Coverage Status

Code Health

package to query basketball-reference.com and save data.

NOTE: This package is still under very heavy development, and things are liable to change on a regular basis. At some point there might be a stable release, but for now we're just trying to get as much done as we can. NOTE: We call this "basketball db" where "db" refers to "database"...we're not creating a relational database here (at least not right now). It's a well-organized directory structure and file-naming scheme. Which is sometimes just as good or better :). NOTE: We're not software developers we're physicists. Bear with us.

Installation

Requirements

Instructions

  1. Clone the repo
  2. cd to repo top level directory
  3. run pip install . or python setup.py install

Save all games for one team for 2015/2016 season

>>> from basketball_db import db_utilities
>>> db_utilities.save_team_season('MIN',2016,basedir='./')

This saves games in

%(basedir)/%(year)/%(month)/%(day)/%(TEAM1)-AT-%(TEAM2).hdf5

and before th first game is downloaded, it actually saves the team's schedule from their basketball-reference schedule page in:

`%(basedir)/SCHEDULES/%(TEAM)-%(YEAR)

where %(YEAR) is the second year in whatever season you're looking at (i.e. Minnesota, 2015/2016 would be saved in "MIN-2016").

Please run this before trying the examples. If you run this command in the examples folder it will give you the data necessary to run the example.

Load data

It saves pandas dataframes for home_box_score and away_box_score, home_shot_chart, away_shot_cart

You can read this in with:

>>> import pandas as pd
>>> fname = './2015/10/28/MIN-AT-LAL.hdf5'
>>> away_shot_chart = pd.read_hdf(fname,'away_shot_chart')
>>> print away_shot_chart

Or, now you can do it with the under-development ShotChart class.

>>> from basketball_db.shotchart import ShotChart
>>> fnames = db_utilities.query_db('MIN','20151028','20151029')
>>> home_shot_chart, away_shot_chart = ShotChart.load_game(fnames[0])
>>> plot = home_shot_chart.plot()
>>> plot.show()

Examples

python save_team.py --team MIN --YMDStart 20151101 --YMDEnd 20151201 --inifile bball.ini

Run tests

Until I get a chance to create some good documentation, the unittests are a solid place to look for how to run the code. They're in basketball_db/tests/.

  1. cd to repo top level
  2. run python setup.py test