Quadrature, i.e. numerical integration.

This module collects both classic and Bayesian quadrature rules used for numerical integration of functions.

Bayesian quadrature methods integrate a function by iteratively building a probabilistic model and using its predictions to adaptively choose points to evaluate the integrand.


quad(fun, bounds[, type]) One-dimensional numerical integration.
nquad(fun, domain[, type]) N-dimensional numerical integration.
bayesquad(fun, fun0, bounds[, nevals, type]) One-dimensional Bayesian quadrature.
nbayesquad(fun, fun0, domain[, nevals, type]) N-dimensional Bayesian quadrature.


Quadrature() An abstract base class for quadrature methods.
PolynomialQuadrature(nodes, weights, bounds) Quadrature rule based on polynomial functions.
BayesianQuadrature() An abstract base class for Bayesian quadrature methods.
VanillaBayesianQuadrature() Vanilla Bayesian quadrature in 1D.
WSABIBayesianQuadrature() Warped Sequential Active Bayesian Integration (WSABI).
ClenshawCurtis(npts_per_dim, ndim, bounds) Clenshaw-Curtis quadrature rule.

Class Inheritance Diagram

Inheritance diagram of probnum.quad.Quadrature, probnum.quad.PolynomialQuadrature, probnum.quad.BayesianQuadrature, probnum.quad.bayesian.bayesquadrature.VanillaBayesianQuadrature, probnum.quad.bayesian.bayesquadrature.WSABIBayesianQuadrature, probnum.quad.polynomial.clenshawcurtis.ClenshawCurtis