plot.gppack: Plots several GP's simultaneously

In nsgp: Non-Stationary Gaussian Process Regression

Description

Plots the GP's corresponding to the control and case data, as well as the null model. Visualizes the log likelihood ratios between the null and individual models. Several boolean parameters for modifying the plot. By default plots the data, posterior mean and 95% interval for CASE and CONTROL.

Usage

## S3 method for class 'gppack'
plot(x, y = NULL, plotdata = TRUE, plotmeans = TRUE,
  plotcovs = TRUE, plotnoises = FALSE, plotnull = FALSE,
  plotratios = "emll", thr = 1, samples = 0, sigma = 2, title = NULL,
  legend = FALSE, plotgradient = TRUE, ...)

Arguments

x	the gppack-object
y	placeholder variable
plotdata	plot the data (default)
plotmeans	plot the GP mean (default)
plotcovs	plot the GP covariances (default)
plotnoises	plot the observational noise (default)
plotnull	plots also the null model
plotratios	plots the ratios, choices are emll, mll, npc, pc
thr	ratio threshold
samples	plot N samples from the GP
sigma	variance level to plot
title	plot title
legend	plot legend
plotgradient	use gradient graphics
...	...

Details

The threshold thr is the logarithmic likelihood ratio between null and control+case models. The default value 1 hence corresponds to a likelihood ratio of 2.72.

Examples

# read toy data
data(toydata)

## Not run: can take several minutes
# perform two-sample regression
res = gpr2sample(toydata$ctrl$x, toydata$ctrl$y, toydata$case$x, toydata$case$y, seq(0,22,0.1))

# pre-computed model for toydata
data(toygps)
res = toygps

# basic plot
plot(res)

# plot also the null model, don't plot data, means or noise
plot(res, plotnull=TRUE, plotdata=FALSE, plotmeans=FALSE, plotnoise=FALSE)
## End(Not run)

nsgp documentation built on May 2, 2019, 9:19 a.m.