ACVMF

class pyapprox.control_variate_monte_carlo.ACVMF(cov, costs)[source]

Bases: object

Methods Summary

__call__(values)

Call self as a function.

allocate_samples(target_cost, **kwargs)

generate_data(nhf_samples, nsample_ratios, …)

get_covariance()

get_nsamples(nhf_samples, nsample_ratios)

get_rsquared(nsample_ratios)

get_variance(nhf_samples, nsample_ratios)

jacobian(x)

objective(x)

variance_reduction(nsample_ratios)

This is not the variance reduction relative to the equivalent Monte Carlo estimator.

Methods Documentation

__call__(values)[source]

Call self as a function.

allocate_samples(target_cost, **kwargs)[source]
generate_data(nhf_samples, nsample_ratios, generate_samples, model_ensemble)[source]
get_covariance()[source]
get_nsamples(nhf_samples, nsample_ratios)[source]
get_rsquared(nsample_ratios)[source]
get_variance(nhf_samples, nsample_ratios)[source]
jacobian(x)[source]
objective(x)[source]
variance_reduction(nsample_ratios)[source]

This is not the variance reduction relative to the equivalent Monte Carlo estimator. A variance reduction can be smaller than one and still correspond to a multi-fidelity estimator that has a larger variance than the single fidelity Monte Carlo that uses the equivalent number of high-fidelity samples