blasso.s3 {monomvn}R Documentation

Summarizing Bayesian Lasso Output

Description

Summarizing, printing, and plotting the contents of a "blasso"-class object containing samples from the posterior distribution of a Bayesian lasso model

Usage

## S3 method for class 'blasso':
print(x, ...)
## S3 method for class 'blasso':
summary(object, burnin = 0, ...)
## S3 method for class 'blasso':
plot(x, which=c("coef", "s2", "lambda2", "tau2i", "m", "pi"),
     burnin = 0, ... )
## S3 method for class 'summary.blasso':
print(x, ...)

Arguments

object a "blasso"-class object that must be named object for the generic methods summary.blasso
x a "blasso"-class object that must be named x for the generic printing and plotting methods print.summary.blasso and plot.blasso
burnin number of burn-in rounds to discard before reporting summaries and making plots. Must be non-negative and less than x$T
which indicates the parameter whose characteristics should be plotted; does not apply to the summary
... passed to print.blasso, or plot.default

Details

print.blasso prints the call followed by a brief summary of the MCMC run and a suggestion to try the summary and plot commands.

plot.blasso uses an appropriate plot command on the list entries of the "blasso"-class object thus visually summarizing the samples from the posterior distribution of each parameter in the model depending on the which argument supplied.

summary.blasso uses the summary command on the list entries of the "blasso"-class object thus summarizing the samples from the posterior distribution of each parameter in the model.

print.summary.monomvn calls print.blasso on the object and then prints the result of summary.blasso

Value

summary.blasso returns a "summary.blasso"-class object, which is a list containing (a subset of) the items below. The other functions do not return values.

B a copy of the input argument thin
T total number of MCMC samples to be collected from x$T
thin number of MCMC samples to skip before a sample is collected (via thinning) from x$T
coef a joint summary of x$mu and the columns of x$beta
s2 a summary of x$s2
lambda2 a summary of x$lambda2 when lasso or ridge regression is active
tau2 a summary of the columns of x$tau2i when lasso is active
bn0 the estimated posterior probability that the individual components of the regression coefficients beta is nonzero
pi the estimated Binomial proportion in the prior for the model order when 2-vector input is provided for mprior

Author(s)

Robert B. Gramacy bobby@statslab.cam.ac.uk

References

http://www.statslab.cam.ac.uk/~bobby/monomvn.html

See Also

blasso


[Package monomvn version 1.6-1 Index]