DOCUMENTATION / LOCAL / OVERVIEW
Overview
One scorer, three ways to use it. Start here to understand the data model and choose your integration.
From raw data to explainable output.
Adapters describe the input schema and transforms. Weight profiles define priorities. Score profiles name the outputs and define their scales. The scorer evaluates the model and returns structured results.
| Way in | Layer | Start here |
|---|---|---|
| Local | App: CLI and StatLine OS | Install the app extra → |
| DevKit | Core: the public Python library | Explore the Python SDK → |
| API | Gateway: HTTP and authentication | Understand access → |
Python EXAMPLE
from statline import load_dataset, score
rows = load_dataset('EBA_Elevate302/eba_players_gp3_plus.csv')
results = score('eba.players', rows, mode='batch', weights='pri')
ranked = sorted(zip(rows, results),
key=lambda pair: (-pair[1]['pri'], -pair[1]['pri_raw']))
for row, result in ranked[:3]:
print(row['PLAYER'], result['pri'])These guides describe source v4.0.0rc6. Live API reference data is identified separately, and the published stable package may be a different version.
Preview v4.0.0rc6 · Generated 2026-10-30T02:37:44.398745+00:00 · Python >=3.10,<3.15 · Generated product manifest