package prc

  1. Overview
  2. Docs
Utilities for precision-recall curves

Install 

dune-project
 Dependency

github.com Readme Edit opam file Versions (2)

Authors

  1. P
    Philippe Veber

Maintainers

  1. P
    philippe.veber@gmail.com

Sources

prc-0.1.0.tbz
sha256=898322ef7f722e37f332362a24c1ad5f1bf8f8bb396cdb549fb44d59fc4f769a
sha512=c85fc65530195913b20c7b0e1763938f3d45187ca5510a37514f5dec5b4b903a9e40b7621ce8509cf023f1ccb1045fc09b362d716b2e6340cbd5960919a65efc

doc/prc/Prc/Binormal_model/index.html

Module Prc.Binormal_modelSource

Binormal model

A Gaussian mixture model for which the precision-recall curve can be computed explicitly (see 1)

Sourcetype t = {
  1. mu_pos : float;
  2. sigma_pos : float;
  3. mu_neg : float;
  4. sigma_neg : float;
  5. alpha : float;
}
Sourceval make : ?mu_pos:float -> ?sigma_pos:float -> ?mu_neg:float -> ?sigma_neg:float -> float -> t

make ?mu_pos ?sigma_pos ?mu_neg ?sigma_neg alpha is a representation of the mixture of two Gaussian distributions of means mu_pos and mu_neg resp. and standard deviation sigma_pos and sigma_neg resp., while alpha is the proportion of positive samples in the mixture.

Sourceval simulation : Gsl.Rng.t -> n:int -> t -> dataset
Sourceval curve : ?n:int -> t -> (float * float) array
Sourceval estimate : dataset -> t
Sourceval auc : t -> float