schulzch / pava

An implementation of the Pool-Adjacent-Violators algorithm (PAVA) for isotonic regression.

PAVA

An implementation of the Pool-Adjacent-Violators algorithm (PAVA) [1,2] for isotonic regression.

References

[1] Barlow, R.E., D.J. Bartholomew, J.M. Bremner and H.D. Brunk. "Statistical Inference under Order Restrictions". Wiley, 1972.

[2] M. Martin Diaz and B. Salvador González, "The validity of the "Pool-Adjacent-Violator" algorithm". Stat. Probab. Lett., vol. 6, no. 3, pp. 143–145, 1988.

License

Licensed under the Apache 2.0 License. Copyright © 2019 Christoph Schulz.