plot.asp: Plots fitted curves or their derivatives

In AdaptFitOS: Adaptive Semiparametric Additive Regression with Simultaneous Confidence Bands and Specification Tests

Plots fitted curves or their derivatives

Description

Plots fitted curves or their derivatives together with simultaneous confidence bands.

Usage

## S3 method for class 'asp'
plot(x, select=NULL, drv=0, bands=TRUE, level=0.95, grid=50, pages=0, 
				plot=TRUE, ylim=NULL, xlab=NULL, ylab=NULL,
     			scb.lwd=1, scb.lty="dotted", shade=FALSE, shade.col=grey(0.85),
     			residuals=FALSE, residuals.col="steelblue", 
     			bayes=FALSE, rug=TRUE,...)

Arguments

x	an asp object created by asp or aspHetero
select	vector specifying which curves in an additive model should be plotted. If NULL, all curves are plotted.
drv	the derivative order. Defaults to 0, i.e. the estimated curves themselves are plotted. First and second derivatives are supported. Does not apply to objects created by scbM.
bands	TRUE in order to include simultaneous confidence bands.
grid	number of points used for the plot, default value 50.
plot	if FALSE no plot is given
ylim	vector or list of vecotrs of limits on y axes. If NULL limits are automatically chosen. If multiple curves are plotted and a two-dimensional vector is given, y axes for all curves will be equal. A list with length equal to the number of smooth curves in the model can be given to specify different y-axes for each smooth.
pages	The number of pages over which to spread the output as in package mgcv. For example, if pages=1 then all terms will be plotted on one page in an automatic fashion. If pages=0 (default) all graphics settings are left as they are.

level	the level of confidence (does not apply to objects created by scbM).
xlab	label for the x axis. A list with length equal to the number of smooth curves in the model can be given to specify different labels for each smooth.
ylab	label for the y axis. A list with length equal to the number of smooth curves in the model can be given to specify different labels for each smooth.
scb.lwd	line width for simultaneous confidence bands
scb.lty	line type for simultaneous confidence bands. Use scb.lty="blank", if you only want to plot the shades.
shade	set to TRUE to produce shaded regions as simultaneous confidence bands for smooths
shade.col	define the color used for shading confidence bands
residuals	if TRUE, partial residuals are added to the plot
residuals.col	color of partial residuals
rug	adds a rug representation (1-d plot) of the data to the plot.
bayes	FALSE for simultaneous confidence bands with (approximate) frequentist coverage probability, TRUE for (approximate) Bayesian coverage probability. See Krivobokova et al. (2010) for details.
...	further arguments to be passed to plot()

Details

plot.asp() first calls scbM and then plot.scbm() to plot an asp object. If plotting takes long (because of a large data set) and you want to plot multiple times with different settings, use scbM and then plot the resulting scbm object with plot.scbm(). Estimated fits are centred to have zero mean. The simultaneous confidence bands have (approximate) frequentist coverage probabilities with automatic bias correction (see references).

Value

grid.x	list of the grid values used
fitted	list of the fitted values on the grid
lcb	list of the lower bounds of the confidence bands
ucb	list of the upper bounds of the confidence bands
drv	the derivative order
Stdev.fit	the standard deviations on the grid
ylim	list of ylim used for plotting
residuals	the partial residuals.

References

Krivobokova, T., Kneib, T., and Claeskens, G. (2010)
Simultaneous confidence bands for penalized spline estimators. Journal of the American Statistical Association, 105(490):852-863.

Wiesenfarth, M., Krivobokova, T., Klasen, S., Sperlich, S. (2012).
Direct Simultaneous Inference in Additive Models and its Application to Model Undernutrition. Journal of the American Statistical Association, 107(500): 1286-1296.

See Also

plot.spm in package SemiPar

Examples

# see asp2()

AdaptFitOS documentation built on July 21, 2022, 5:10 p.m.