getFunctionSamples: Extract function values samples from posterior coefficients...

In hypergsplines: Bayesian model selection with penalised splines and hyper-g prior

Description

Extract function values samples from posterior coefficients samples

Usage

  getFunctionSamples(x, covName, samples, modelData)

Arguments

x	the vector of abscissa values (on the original scale!)
covName	string with the name of the covariate
samples	the samples object (either from getSamples or the samples element from glmGetSamples)
modelData	the corresponding model data object

Value

the function values samples as a matrix.

Author(s)

Daniel Sabanes Bove [e-mail]

Examples

## get some data
attach(longley)

## get model data
md <- modelData(y=Employed,
                X=cbind(GNP, Armed.Forces))

## get posterior samples for a specific model configuration
samples <- getSamples(config=c(2, 1),
                      nSamples=1000L,
                      modelData=md)

## and then get function samples:
res <- getFunctionSamples(x=
                          seq(from=min(GNP),
                              to=max(GNP),
                              length=100L),
                          covName="GNP",
                          samples=samples,
                          modelData=md)
str(res)


## get model data with cubic splines
md <- modelData(y=Employed,
                X=cbind(GNP, Armed.Forces),
                nKnots=10L,
                splineType="cubic")

## get posterior samples for a specific model configuration
samples <- getSamples(config=c(2, 1),
                      nSamples=1000L,
                      modelData=md)

## and then get function samples:
res <- getFunctionSamples(x=
                          seq(from=min(GNP),
                              to=max(GNP),
                              length=100L),
                          covName="GNP",
                          samples=samples,
                          modelData=md)
str(res)

hypergsplines documentation built on May 2, 2019, 6:14 p.m.