summary.bdw

This function produces result summaries from a MCMC object of class 'bdw'

A Bayesian regression model for discrete response, where the conditional distribution is modelled via a discrete Weibull distribution. This package provides an implementation of Metropolis-Hastings and Reversible-Jumps algorithms to draw samples from the posterior. It covers a wide range of regularizations through any two parameter prior. Examples are Laplace (Lasso), Gaussian (ridge), Uniform, Cauchy and customized priors like a mixture of priors. An extensive visual toolbox is included to check the validity of the results as well as several measures of goodness-of-fit.

Hamed Haselimashhadi

BDWreg

Bayesian Inference for Discrete Weibull Regression

summary.bdw function

<dl> <dt>object</dt>
<dd>The object containing the MCMC results of class 'bdw'.</dd> <dt>est</dt>
<dd>The statistic that is used to estimate parameters from marginal densities. The default is 'mode'.</dd> <dt>prob</dt>
<dd>A numerical value in (0 , 1). Corresponding probability for Highest Posterior Density (HPD) interval. If either RJ=TRUE or penalized=TRUE, coefficients are marked as zero if corresponding prob% HPD intervals contain zero.</dd> <dt>samp</dt>
<dd>Logical flag. If TRUE, analyse a sample instead of whole MCMC chain to save time. 
 * enable if object is created by 'bdw.mc' function.</dd><dt>...</dt>
<dd>Advance users. Extra parameters to pass to optim() function</dd></dl>

Arguments

Hamed Haselimashhadi &lt;hamedhaseli@gmail.com&gt;

Author

Summary for a MCMC object of class 'bdw' — summary.bdw

<dl>

 <dt>object</dt>
<dd>The object containing the MCMC results of class 'bdw'.</dd>

 <dt>est</dt>
<dd>The statistic that is used to estimate parameters from marginal densities. The default is 'mode'.</dd>

 <dt>prob</dt>
<dd>A numerical value in (0 , 1). Corresponding probability for Highest Posterior Density (HPD) interval. If either RJ=TRUE or penalized=TRUE, coefficients are marked as zero if corresponding prob% HPD intervals contain zero.</dd>

 <dt>samp</dt>
<dd>Logical flag. If TRUE, analyse a sample instead of whole MCMC chain to save time. 
 * enable if object is created by 'bdw.mc' function.</dd>

<dt>...</dt>
<dd>Advance users. Extra parameters to pass to optim() function</dd>

</dl>

summary.bdw: Summary for a MCMC object of class 'bdw'

Description

Usage

Arguments

Author

See Also

Examples