summaryGspline2 {glmmAK}R Documentation

Summary for a bivariate G-spline (penalized Gaussian mixture)

Description

This function is primarily designed to work out the MCMC output from functions cumlogitRE and logpoissonRE in which a distribution of the bivariate random effect was specified as a G-spline. It computes posterior pointwise mean and quantiles for a G-spline density based on the MCMC output.

Besides the summary for the joint bivariate G-spline density it also directly computes summaries for both marginal G-splines.

Usage

summaryGspline2(x1, x2, mu1, mu2, sigma1, sigma2,
   standard=TRUE, intcpt, scale,
   probs, values=FALSE,
   dir=getwd(), wfile="weight.sim", indfile="knotInd.sim",
   header=TRUE, logw=FALSE, is.indfile=TRUE, 
   skip=0, nwrite)

Arguments

x1 grid of values for the first margin at which we evaluate the G-spline density
x2 grid of values for the second margin at which we evaluate the G-spline density
mu1 a vector with G-spline knots (means of basis G-splines) for the first margin
mu2 a vector with G-spline knots (means of basis G-splines) for the second margin
sigma1 basis standard deviation(s) for the first margin. If a single number is supplied then it is assumed that all basis G-splines in the first margin have the same standard deviation. Alternatively a vector of the same length as mu1 can be given in which case the basis G-splines in the first margin do not necessarily have the same standard deviations
sigma2 basis standard deviation(s) for the second margin. If a single number is supplied then it is assumed that all basis G-splines in the second margin have the same standard deviation. Alternatively a vector of the same length as mu2 can be given in which case the basis G-splines in the second margin do not necessarily have the same standard deviations
standard if TRUE then the standardized (zero-mean, unit-variance) G-splines are computed and summarized
intcpt a two-column matrix with sampled intercept values. If not supplied it is assumed that all intercepts are equal to zero.
It does not have to be supplied if standard=TRUE.
scale a two-column matrix with sampled intercept values of the G-spline scale (in most of my papers denoted by tau) If not supplied it is assumed that all scale values are equal to one.
It does not have to be supplied if standard=TRUE.
probs probabilities for which the (pointwise) sample quantiles of the G-spline should be computed.
If not given only average (and values) of the G-spline are computed
values if TRUE also values of the G-spline at each (MCMC) iteration are returned.
If FALSE only sample mean (and quantiles) of the G-spline are returned
dir character giving the directory where the file with (sampled) G-spline (log-)weights is stored
wfile character giving the name of the file with (sampled) G-spline (log-)weights
indfile character giving the name of the file where it is indicated which G-spline components correspond to non-zero weights.
It does not have to be supplied if is.indfile=FALSE.
header logical indicating whether the files wfile, indfile contain a header
logw logical indicating whether the file wfile contains logarithms of the weights
is.indfile logical.
If TRUE then wfile contains only the non-zero weights and the G-spline is reconstructed using indfile.
If FALSE then wfile must contain on each row weights of all components and indfile is ignored.
skip number of data rows that should be skipped at the beginning of the files wfile, indfile
nwrite frequency with which is the user informed about the progress of computation (every nwriteth iteration count of iterations change)

Value

A list with the following components (component values, values1, values2 are present only when the argument values was TRUE). The description below applies to the situation when probs=0.5.

summary a list with the components
x1 grid of values for the first margin at which we evaluate the G-spline density
x2 grid of values for the second margin at which we evaluate the G-spline density
Mean a matrix with length(x1) rows and length(x2) columns giving
the pointwise posterior mean of the joint G-spline density
50% a matrix with length(x1) rows and length(x2) columns
giving the pointwise posterior 50% quantile of the joint G-spline density.
There is a matrix of this type for each probs value.
summary1 a data.frame with the following columns
x a grid of values at which the first marginal G-spline density is evaluated
Mean pointwise posterior mean of the first marginal G-spline density
50% pointwise posterior 50% quantile of the first marginal G-spline density.
There is one column of this type for each probs value
summary2 a data.frame for the second marginal G-spline having the same structure as summary1.
values a matrix with one column for each (x1, x2) value and number of rows equal to the length of the MCMC. In each row, there is a joint G-spline density evaluated at one MCMC iteration.
values1 a matrix with one column for each x1 value and number of rows equal to the length of the MCMC. In each row, there is the first marginal G-spline density evaluated at one MCMC iteration.
values1 a matrix with one column for each x1 value and number of rows equal to the length of the MCMC. In each row, there is the second marginal G-spline density evaluated at one MCMC iteration.

Author(s)

Arnošt Komárek arnost.komarek[AT]mff.cuni.cz

Examples

### See ex-Toenail.pdf, ex-Toenail.R
### and ex-Epileptic.pdf, ex-Epileptic.R
### available in the documentation
### to the package

[Package glmmAK version 1.2 Index]