How djsherwood/circe works
This is a compact regression library for named numeric tables with discrete circular features. Its principal workflow is fitting a forest on training rows, then predicting from a table with the same columns and order. The implementation also supports ordinary threshold splits for non-circular features, including drift features such as cycle_number. The sampled source shows input validation and tree split behavior, but omits part of the forest-fitting implementation, so the graph stays at the public estimator boundary.
Main entry points: ring_forest.py Ring forest estimator
Components 3
Other components
- User
- Ring forest estimator
ring_forest.py - Predictions
Connections 2
- User → Ring forest estimator: fits and predicts
README.md - Ring forest estimator → Predictions: returns predictions
README.md