Skip to content
forked from Affirm/splinator

Splinator: spline-based regression models for calibration

License

Notifications You must be signed in to change notification settings

jxudata/splinator

 
 

Repository files navigation

Splinator 📈

Spline-based Regression Models for Probability Calibration

scikit-learn compatible

Installation

pip install splinator

Algorithm

Coming (Link to medium blog and arxiv PDF)

Releases

alpha version in active development. the stable release is expected to arrive by the end of 2022

Development

The dependencies are managed by poetry

To run tests, run poetry run pytest splinator/tests

Example Usage

from splinator.estimators import LinearSplineLogisticRegression
import numpy as np

# random synthetic dataset
n_samples = 100
rng = np.random.RandomState(0)
X = rng.normal(loc=100, size=(n_samples, 2))
y = np.random.randint(2, size=n_samples)

lslr = LinearSplineLogisticRegression(n_knots=10)
lslr.fit(X, y)

About

Splinator: spline-based regression models for calibration

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%