pyAFL is a AFL (Australian Football League) data fetching libary. It scrapes data from https://afltables.com/ and converts results to structured Python objects for easier analytics.
- Install from pyPI
pip install pyAFL
OR
- Clone the repository locally:
git clone https://github.com/RamParameswaran/pyAFL.git
- Player
- Team
- Season
Instantiates the Player object. The init method finds the matching player on afltables.com based on the name argument string. name must be an exact match, in the format "Firstname Lastname" and must be a name listed at https://afltables.com/afl/stats/playersA_idx.html
Pulls player stats (career totals, and season-by-season summary) and presents as a Python object. This function returns a PlayerStats object with attributes:
- season_stats_total (Pandas dataframe)
- season_stats_average (Pandas dataframe)
- season_results (list of Pandas dataframes)
Example
>>> from pyAFL.players.models import Player
>>> player = Player("Nick Riewoldt")
>>> player.url
"https://afltables.com/afl/stats/players/N/Nick_Riewoldt.html"
>>> stats = player.get_player_stats()
>>> stats.season_stats_total
Year Team # GM ... MI 1% BO GA
0 2001 St Kilda 12 6.00 ... 2.00 2.00 2.00 NaN
1 2002 St Kilda 12 22.00 ... 28.00 44.00 13.00 NaN
2 2003 St Kilda 12 22.00 ... 39.00 38.00 11.00 5.00
3 2004 St Kilda 12 25.00 ... 89.00 46.00 7.00 27.00
>>> stats.season_stats_average
Year Team # GM ... MI 1% BO GA
0 2001 St Kilda 12 6.00 ... 0.33 0.33 0.33 NaN
1 2002 St Kilda 12 22.00 ... 1.27 2.00 0.59 NaN
2 2003 St Kilda 12 22.00 ... 1.77 1.73 0.50 0.23
3 2004 St Kilda 12 25.00 ... 3.56 1.84 0.28 1.08
>>> stats.season_results
[ St Kilda - 2001 ...8 columns], St Kilda - 2002 ...8 columns], St Kilda - 2003 ...8 columns], St Kilda - 2004 ...8 columns], St Kilda - 2005 ...8 columns], St Kilda - 2006 ...8 columns], ...]
Instantiates a Team object. pyAFL automatically instantiates Team objects for all current and former AFL teams. They can be imported by their team abbreviation:
>>> from pyAFL.teams import (ADE, BRI, BRB, CAR, COL, ESS, FIZ, FRE, GEE, GC, GWS, HAW, MEL, NOR, POR, RIC, STK, SYD, UNI, WCE, WBD)
>>> # or
>>> from pyAFL.teams import ALL_TEAMS, CURRENT_TEAMS
>>> ADE
<Team: Adelaide>
>>> CURRENT_TEAMS
[<Team: Adelaide>, <Team: Brisbane Lions>, <Team: Carlton>, <Team: Collingwood>, <Team: Essendon>, <Team: Fremantle>, <Team: Geelong>, <Team: Gold Coast>, <Team: Greater Western Sydney>, <Team: Hawthorn>, <Team: Melbourne>, <Team: North Melbourne>, <Team: Port Adelaide>, <Team: Richmond>, ...]
Returns a list of all historical players for this team. The return is a list of pyAFL Player objects. These Player objects can be queried to get player stats using the Player classmethods noted above.
>>> from pyAFL.teams import ADE
>>> # Let's get a list of all players who have every played for Adelaide (i.e. all players from https://afltables.com/afl/stats/teams/adelaide.html)
>>> ADE.players
[<Player: Mcleod, Andrew>, <Player: Edwards, Tyson>, <Player: Ricciuto, Mark>, <Player: Hart, Ben>, <Player: Smart, Nigel>, <Player: Goodwin, Simon>, <Player: Bickley, Mark>, <Player: Thompson, Scott>, ...]
Returns a Pandas DataFrame with all historical game results for this team. The DataFrame has a datetime index.
>>> from pyAFL.teams import ADE
>>> # Let's get all historical game data (i.e. all the data from https://afltables.com/afl/teams/adelaide/allgames.html)
>>> ADE.games
Rnd T ... Crowd Date
Date ...
1991-03-22 19:40:00 R1 H ... 44902.0 Fri 22-Mar-1991 7:40 PM
1991-03-31 14:10:00 R2 H ... 43850.0 Sun 31-Mar-1991 2:10 PM
... ... .. ... ... ...
2020-09-13 13:05:00 R17 A ... 2735.0 Sun 13-Sep-2020 1:05 PM
2020-09-19 16:40:00 R18 H ... 17710.0 Sat 19-Sep-2020 4:40 PM
[688 rows x 13 columns]
>>> # The DataFrame can be sliced by Date in human-readable format!
>>> ADE.games.loc['2016-01-01':'2019-12-31']
Rnd T ... Crowd Date
Date ...
2016-03-26 19:25:00 R1 A ... 25485.0 Sat 26-Mar-2016 7:25 PM
2016-04-02 13:15:00 R2 H ... 50555.0 Sat 02-Apr-2016 1:15 PM
... ... .. ... ... ...
2019-08-17 16:05:00 R22 H ... 48175.0 Sat 17-Aug-2019 4:05 PM
2019-08-25 13:10:00 R23 A ... 9560.0 Sun 25-Aug-2019 1:10 PM
[93 rows x 13 columns]
>>> # Let's see what columns are contained in the DataFrame
>>> ADE.games.columns
Index(['Rnd', 'T', 'Opponent', 'Scoring', 'For', 'Scoring', 'Against', 'Result', 'Margin', 'W-D-L', 'Venue', 'Crowd', 'Date'], dtype='object')
Retrieves the season stats for the specified year, including the individual player stats for all Players who played a game during the year. This function returns a Pandas DataFrame.
>>> from pyAFL.teams import ADE
>>> # Who player for Adelaide in 2019, and how did they perform? (See https://afltables.com/afl/stats/2019.html)
>>> ADE.season_stats(2019)
# Player GM ... GA %P SU
0 10 Boak, Travis 21 ... 14.0 79.8 NaN
1 11 Rockliff, Tom 18 ... 8.0 77.3 NaN
2 33 Byrne-Jones, Darcy 22 ... 7.0 84.0 NaN
3 43 Houston, Dan 21 ... 3.0 82.1 NaN
... ... ... ...
34 12 McKenzie, Trent 1 ... NaN 94.0 NaN
35 30 Atley, Joe 1 ... NaN 68.0 NaN
36 31 Johnson, Aidyn 1 ... 1.0 77.0 NaN
[38 rows x 28 columns]
Instantiates the Season object. This is a simple way to access match scores and results, leaving detailed team and player statistics to the Team and Player classes.
The init method finds the season on afltables.com based on the year argument integer.
Pulls season stats (ladders, matches, match summaries) and presents as a Python object. This function returns a SeasonStats object with attributes:
- season_ladders (list): abbreviated ladders (dataframes) at end of each round
- season_matches (list): Match object for each match in the season
- match_summary (Pandas dataframe): summary of all matches in the season
- final_ladder (Pandas dataframe): the final season ladder (empty list if season is unfinished)
>>> from pyAFL.seasons.models import Season
# Instantiate Season object for 2021 (See https://afltables.com/afl/seas/2021.html)
>>> season_2021 = Season(2021)
>>> stats_2021 = season_2021.get_season_stats()
>>> stats_2021.match_summary
Date Round Game number Venue Home team ... Home team score detail Away team score detail Winning team Margin Year stage
2021-03-18 18:25:00 1 1 M.C.G. Richmond ... [3, 3, 8, 5, 10, 8, 15, 15] [3, 2, 6, 6, 8, 12, 11, 14] Richmond 25 Regular season
2021-03-19 18:50:00 1 2 M.C.G. Collingwood ... [1, 2, 4, 6, 5, 7, 7, 11] [3, 2, 6, 5, 9, 7, 10, 9] Western Bulldogs 16 Regular season
2021-03-20 12:45:00 1 3 M.C.G. Melbourne ... [4, 4, 7, 6, 9, 8, 11, 14] [0, 4, 3, 8, 6, 9, 8, 10] Melbourne 22 Regular season
2021-03-20 15:35:00 1 4 Adelaide Oval Adelaide ... [4, 4, 11, 7, 13, 9, 15, 13] [2, 3, 5, 5, 10, 8, 13, 13] Adelaide 12 Regular season
2021-03-20 18:25:00 1 5 Docklands Essendon ... [2, 4, 10, 6, 11, 10, 13, 13] [3, 1, 4, 3, 12, 3, 14, 8] Hawthorn 1 Regular season
...
The unit tests can be run by running pytest from the project directory, like so;
pytest
There is a lot to do so contributions are really appreciated! This is a great project for early stage developers to work with.
To begin it is recommended starting with issues labelled good first issue.
How to get started:
- Fork the pyAFL repo.
- Create a new branch in you current repo from the 'master' branch with issue label.
- 'Check out' the code with Git or GitHub Desktop
- Check contributing.md
- Push commits and create a Pull Request (PR) to pyAFL