assist (version 3.1.3)

plot.bCI: Bayesian Confidence Interval Plot of a Smoothing Spline Fit

Description

Create trellis plots of a nonparametric function fit together with its (approximate) 95% Bayesian confidence intervals from a ssr/slm/snr/snm object.

Usage

# S3 method for bCI
plot(x, x.val=NULL, type.name=NULL, …)

Arguments

x

an object of class "bCI" containing point evaluation of the unknown function and/or corresponding posterior standard devaitions.

x.val

an optional vector representing values of argument based on which the function is to evaluate.

type.name

an optional character vector specifying the names of fits.

…

options suitable for xyplot.

Details

This function is to visualize a spline fit by use of trellis graphic facility with Bayesian confidence intervals superposed. Multi-panel plots, based on xyplot, are suitable for SS ANOVA decomposition of a spline estimate.

See Also

predict.ssr, intervals.slm, intervals.snr, intervals.snm

Examples

# NOT RUN {
x<- seq(0, 1, len=100)
y<- 2*sin(2*pi*x)+rnorm(x)*0.5

fit<- ssr(y~x, cubic(x))
p.fit<- predict(fit)
# }
# NOT RUN {
plot(p.fit)
# }
# NOT RUN {
plot(p.fit,type.name="fit")
# }