Comparison Plots for Analysis of Variance Models

DESCRIPTION:

Creates a set plots useful for comparing fitted analysis of variance models stored in a class fit.models object belonging to virtual class aovfm. The plot options are (2) Normal QQ-Plot of Residuals, (3) Estimated Kernel Density of Residuals, (4) Residuals vs Fitted Values, (5) Sqrt of abs(Residuals) vs Fitted Values, (6) Response vs Fitted Values, (7) Residual-Fit Spread, (8) Standardized Residuals vs Index (Time), (9) Overlaid Normal QQ-Plot of Residuals, and (10) Overlaid Estimated Density of Residuals.

USAGE:

plot.aovfm(x, which.plots="all", vertical.outlier=.99, smooths=F,
           rugplot=F, id.n=3, envelope=T, half.normal=F,
           robustQQline=T, mc.samples=100, level=0.95, seed=289, ...)

REQUIRED ARGUMENTS:

x
a fit.models object belonging to virtual class aovfm.

OPTIONAL ARGUMENTS:

which.plots
either "ask", "all", or an integer vector specifying which plots to draw. If which.plots is an integer vector, use the plot numbers given in the description above (or in the "ask" menu).
vertical.outlier
p-value used to calculate the standard normal quantile used as the outlier threshold for residuals.
smooths
if TRUE, smooth curves are approximated to the scatterplots using loess.smooth and added to the appropriate plots.
rugplot
if TRUE, a univariate histogram or rugplot is displayed along the base of each plot, showing the occurrence of each x-value; ties are broken by jittering.
id.n
number of outliers identified in plots.
envelope
if TRUE, a simulation envelope is added to the QQ-plot.
half.normal
if TRUE, half normal QQ-plots will be used.
robustQQline
if TRUE, a robust fit is added to the QQ-plot.
mc.samples
number of samples used to compute the simulation envelope.
level
confidence level for the simulation envelope.
seed
an integer between 0 and 1023. The seed value used for random number generation in the QQ-plot simulation envelope.

VALUE:

x is invisibly returned.

SIDE EFFECTS:

The selected plots are drawn on a graphics device.

REFERENCES:

Atkinson, A. C. (1985). Plots, Transformations and Regression. New York: Oxford University Press.

SEE ALSO:

, .

EXAMPLES:

lawson.aov <- aov(Loss ~ ., data=lawson.dat)
lawson.aovRob <- aovRob(Loss ~ ., data=lawson.dat)
# lawson.fm is a fit.models object with virtual class aovfm:
lawson.fm <- fit.models(lawson.aovRob, lawson.aov)
plot(lawson.fm)