okNBA is coded in Python and was created to compare ESPN Fantaasy Basketball League's against eachother. Based on the user's predefined categories (FA%, 3's, GP, ...categories), okNBA calcutates the relative strength of their team in relation to their opponents. It accesses your ESPN Fantasy League's information and fetch's all relevent fantasty league data, free agent data being included.
Pandas to organize the player/team statistics and matplotlib to visually chart said data.
It's for people that don't know a thing about basketball and don't want to end up last place in their fantasy pool
beautifulsoup4 | pytz | random | requests | datetime | numpy | matplotlib | pandas | statistics | unicodedata | pickle
- espn_api (https://github.com/cwendt94/espn-api)
- most data scraped from here: https://basketball-reference.com
Navigate to ClassData_Center.py and input your ESPN credentials. Then, open BETA.ipynb and run/install dependencies that you do not have. BETA will:
- create a data scructure that will hold all active players and their corresponding stats
- rank order every stat for every player. There are 9 stats to consider
- sign into your espn_fantasy league
- get roster data for every fantasy team
- rank order fantasy teams according to the metrics/categories
- make analytic comparisons of players and teams
scraper()
- Scrapes basketball reference.com
- Returns a PD datafrane
class Player
- Stores player data
- Makes a dictionary entry in AllPlayersData
- Various getters/setters
- Various data organizers
class AllPlayersData()
- Main dictionary that holds every players NBA stats
- Key: a Player() object
- Value: alist representing stats and data
- functions to handle the normalization of player strings (TJ vs. T.J.)
- combines multiple data sets and manages some exceptions, handles duplicate entries (mid season trade problem)
- sorting functions, key functions
- imports double double data from CSV
class Run()
- coordinates the operations between sub classes, initiates the creation of the AllplayersData object
- scrapes web data, creats pandas, manually deals with some exceptions, checks for data object duplicates
- rank orders every single player
class Data_Center()
- The Data_Center is the main storage of all objects used and referenced in the program.
- It creates and stores AllPlayerData which Player data can be accessed
- It creates stores private leagueID, URLs, various cookies and keys
- It creates stores Private League Data
- It creates stores NBA Team Data and their corresponding schedules
Uses the requests library, to navigate ESPN and submits the appropriate cookies and keys to access said private league. Retrieve and update the private league object from ‘espn_api’, and store this object as a self variable in the Data_Center
Functions that compare private teams and player matchups, with corresponding charts
Obtain your ESPN cookies! ESPN_S2 and SWID cookie are values needed to authenticate your private fantasy sports leagues. Go find them!
- a dynamic and responsive UI. Integrate a charts library.
- categories are currently hardcoded and need to be updated as Fantasy leagues are changed. you can find them hardcoded in espn_scrapper.py
- find an accurate singular data source, so data normalization can be off-loaded
- currently, there is no dependency architecture
Have fun testing and improving it! 😎