int.plot: Generic Interaction Plots of Response with Interval Bars

In byandell/pda: Practial Data Analysis for Designed Experiments

Description

The routines lsd.plot simply calls int.plot, which is a method. ci.plot is a method. An interaction plot is created. By default, lines for each value of the group are drawn in differentline styles ('lsd.plot' and 'int.plot') and with plot symbols so that they may be more easily distinguished. This corresponds to plot type="b", which can be overridden. RA Fisher's LSD bar (lsd.plot), confidence intervals (ci.plot) or test ellipses may be added with the bar.plot option.

Usage

lsd.plot(object, ...)
ci.plot(object, ...)
int.plot(object, ...)

Arguments

object	object to be plotted
...	optional specification of other parameters to be passed to methods, plotting routines and/or to the bar-drawing routines lsd.bar and se.bar. There are some parameters, such as width and edge which can change plotting area.

Value

list containing sd, rdf, level and sample size, and either lsd (lsd.plot) or width (ci.plot).

See Also

xyplot, margin.plot, lsd.bar, ci.width, int.plot.lm, pvalue.ellipse.

Examples

# the first use mimics interaction.plot
data(ToothGrowth)
ToothGrowth$dose = ordered(ToothGrowth$dose)
attach(ToothGrowth)
lsd.plot(dose, len, supp)
detach()
# or you could get more fancy by using a model fit
# here is an interaction plot based on linear model fit
ToothGrowth$dose = ordered(ToothGrowth$dose)
tooth.fit = aov(len~dose*supp, ToothGrowth)
lsd.plot( tooth.fit, ToothGrowth, factors = c("supp","dose"))

byandell/pda documentation built on May 13, 2019, 9:27 a.m.