swar/nba_api

nba_api is a Python package that gives developers a way to pull data from NBA.com without having to reverse-engineer the site's data endpoints themselves. The NBA's website exposes various data feeds that power its stats pages, and this library wraps those feeds so you can call them with simple Python code instead of manually constructing HTTP requests. There are two main categories of data. The stats section covers historical records: career stats for a player, game logs, team records, box scores, and more. The live section covers in-progress games and today's scoreboard. Both sections return data that you can work with as raw JSON, as a Python dictionary, or as a pandas DataFrame if you have that library installed. Installing the package takes one command: pip install nba_api. It requires Python 3.10 or newer and depends on the requests and numpy libraries. Once installed, you import whichever endpoint class you need, create an instance of it with the relevant parameters (such as a player ID), and then call a method to get your data back in the format you prefer. The README includes working code examples showing how to fetch a specific player's career stats and how to get the current day's scoreboard. The package also includes static data files for players and teams. These are bundled locally so that common lookups, like finding a player's ID by name, do not require a network request every time. The project is open source under the MIT license. It includes community support via Slack and Stack Overflow. One note from the README: NBA.com has its own Terms of Use that govern how you may use data retrieved through the site, so checking those before building a public-facing product is worth doing.