mmcisomeans: Functions used by mmcplot.

In HH: Statistical Analysis and Data Display: Heiberger and Holland

Functions used by mmcplot.

Description

Functions used by mmcplot.

Usage

mmcisomeans(mmc, col=c("black","red"), lwd=c(1,1), lty=c(2,1),
            type = "mca", xlim = NULL, ylim = NULL, ...,
            axis.right=2.2,
            ylab=paste(
              mmc$none$ylabel, "means",
              " | ",
              mmc$none$focus, "level"),
            ylab.right=NULL,
            xlab="contrast value",
            contrast.label=TRUE,
            means.height=TRUE)

mmcmatch(mmc, col=c("black","red"), lwd=c(1,1), lty=c(2,1),
         type = "mca", xlim = NULL, ylim = NULL, ...,
         axis.right=2.2,
         ylab=NULL,
         ylab.right=NULL,
         xlab="contrast value",
         contrast.label=TRUE,
         xlim.match=(type != "none"))

mmcboth(mmc, col=c("black","red"), lwd=c(1,1), lty=c(2,1),
        type = "mca", h = c(0.7, 0.3), xlim = NULL, ylim = NULL, ...,
        ylab.right=NULL, MMCname="MMC", Tiebreakername="Tiebreaker")

Arguments

mmc	mmc object or other object as indicated by method.
type	One of c("mca", "lmat", "linfct", "none"). For the default "mca", an MMC plot is drawn of the pairwise contrasts. For "lmat" or "linfct", an MMC plot is drawn of the contrasts specified to mmc in the lmat or linfct argument. For "none", a confidence interval plot for the group means is drawn.
h	h argument for resizePanels.
xlim, ylim, xlab, ylab, ylab.right	Standard xyplot arguments.
col, lwd, lty	Standard xyplot arguments applied to the line segments representing the contrasts.
...	Other arguments, to be forwarded to methods.
axis.right	Value used internally for par.settings=list(layout.widths=list(axis.right=axis.right)). The user may need to set this in two circumstances. First, if the contrast names overflow the right edge of the plotting window, then use a larger value than the default. Second, if there is a ylab.right and it is too far away from the figure, then use a smaller value than the default.
contrast.label	Logical. The default TRUE means place the word contrasts at the bottom of the right axis under the tick labels. FALSE means don't place anything there.
MMCname, Tiebreakername	Panel names when mmcplot is used with style="both".
xlim.match	Logical. If TRUE, use xlim based on the contrasts in the mca component. If FALSE, use xlim based on the values of the estimates in the current component.
means.height	Logical, with default value TRUE. When TRUE, then display the values of the group means as the left axis tick labels.

Value

A "trellis" object.

Author(s)

Richard M. Heiberger <[e-mail]>

References

See mmc for the references.

See Also

mmc for the discussion of the MMC. mmcplot for the user calls that get executed by the functions documented here.

Examples

## Not run: 
  ## these examples exercise all optional arguments
data(catalystm)
catalystm1.aov <- aov(concent ~ catalyst, data=catalystm)

catalystm.glht <-
   glht(catalystm1.aov, linfct = mcp(catalyst = "Tukey"))
confint(catalystm.glht)

plot(catalystm.glht)  ## this is the multcomp:::plot.glht

mmcplot(catalystm.glht)  ## mmcplot.glht sends its argument to HH:::as.multicomp.glht with
## the default arguments (estimate.sign = 1, order.contrasts = TRUE) unless overridden:
##
mmcplot(catalystm.glht, order.contrasts=FALSE, estimate.sign=0, main="B'")

catalystm.lmat <- cbind("AB-D" =c(1, 1, 0,-2),
                        "A-B"  =c(1,-1, 0, 0),
                        "ABD-C"=c(1, 1,-3, 1))
dimnames(catalystm.lmat)[[1]] <- levels(catalystm$catalyst)
catalystm.mmc <-
   mmc(catalystm1.aov,
       linfct = mcp(catalyst = "Tukey"),
       focus.lmat=catalystm.lmat)

mmcplot(catalystm.mmc, type="mca",  style="confint")
mmcplot(catalystm.mmc, type="lmat", style="confint")
mmcplot(catalystm.mmc, type="none", style="confint")
mmcplot(catalystm.mmc, type="none", style="confint", xlim.match=FALSE,
        main="xlim.match=FALSE is default for none confint")
mmcplot(catalystm.mmc, type="none", style="confint", xlim.match=TRUE, main="out of bounds")

mmcplot(catalystm.mmc$mca,  style="confint")
mmcplot(catalystm.mmc$lmat, style="confint")
mmcplot(catalystm.mmc$none, style="confint")

plot(catalystm.mmc) ## HH:::plot.mmc.multicomp method

mmcplot(catalystm.mmc)

mmcplot(catalystm.mmc)
mmcplot(catalystm.mmc, style="isomeans")
mmcplot(catalystm.mmc, style="confint")
mmcplot(catalystm.mmc, style="both")

mmcplot(catalystm.mmc, style="isomeans", type="mca")
mmcplot(catalystm.mmc, style="isomeans", type="lmat")
mmcplot(catalystm.mmc, style="isomeans", type="linfct")
mmcplot(catalystm.mmc, style="isomeans", type="none")
mmcplot(catalystm.mmc, style="isomeans", type="none", xlim.match=FALSE)

mmcplot(catalystm.mmc, style="confint",  type="mca")
mmcplot(catalystm.mmc, style="confint",  type="lmat")
mmcplot(catalystm.mmc, style="confint",  type="linfct")
mmcplot(catalystm.mmc, style="confint",  type="none")
mmcplot(catalystm.mmc, style="confint",  type="none", xlim.match=FALSE)

mmcplot(catalystm.mmc, style="both",     type="mca")
mmcplot(catalystm.mmc, style="both",     type="lmat")
mmcplot(catalystm.mmc, style="both",     type="linfct")
mmcplot(catalystm.mmc, style="both",     type="none")
mmcplot(catalystm.mmc, style="both",     type="none", xlim.match=FALSE)

mmcplot(catalystm.mmc$mca)
mmcplot(catalystm.mmc$mca$glht)
mmcplot(catalystm.mmc$none)
mmcplot(catalystm.mmc$none$glht)
mmcplot(catalystm.mmc$lmat)
mmcplot(catalystm.mmc$lmat$glht)

mmcplot(catalystm.mmc, type="none")
mmcplot(catalystm.mmc, type="none", xlim.match=FALSE)
mmcplot(catalystm.mmc$none)

## End(Not run)

HH documentation built on May 29, 2024, 6:24 a.m.