warmup_admkr: Burn-in period

In bbemkr: Bayesian bandwidth estimation for multivariate kernel regression with Gaussian error

Description

By minimising the cost value, the function estimates the bandwidths of the regressors and kernel-form error density for the burn-in period

Usage

warmup_admkr(x, inicost, mutsizp, errorsizp, warm = 100, prob = 0.234, 
    errorprob = 0.44, data_x, data_y) 

Arguments

x	Log of square bandwidths
inicost	Cost value
mutsizp	Step size of random-walk Metropolis algorithm for the regressors
errorsizp	Step size of random-walk Metropolis algorithm for the kernel-form error density
warm	Number of burn-in iterations
prob	Optimal covergence rate of random-walk Metropolis algorithm for the regressors
errorprob	Optimal covergence rate of random-walk Metropolis algorithm for the kernel-form error density
data_x	Regressors
data_y	Response variable

Value

x	Log of square bandwidths
cost	Cost value
mutsizp	Step size of random-walk Metropolis algorithm for the regressors
errorsizp	Step size of random-walk Metropolis algorithm for the kernel-form error density

Author(s)

Han Lin Shang

See Also

mcmcrecord_admkr, logdensity_admkr, loglikelihood_admkr, logpriors_admkr

bbemkr documentation built on May 1, 2019, 10:53 p.m.