mmc {HH} | R Documentation |
Constructs a "mmc.multicomp"
object from the formula and
other arguments. The object must be explicitly plotted.
glht.mmc(model, ...) ## R ## S3 method for class 'glht': glht.mmc(model, ...) ## S3 method for class 'lm': glht.mmc(model, ## lm object linfct=NULL, focus= if (is.null(linfct)) { if (length(model$contrasts)==1) names(model$contrasts) else stop("focus or linfct must be specified.") } else { if (is.null(names(linfct))) stop("focus must be specified.") else names(linfct) }, focus.lmat, ylabel=deparse(terms(model)[[2]]), lmat=if (missing(focus.lmat)) { t(linfct) } else { lmatContrast(t(none.glht$linfct), focus.lmat) }, lmat.rows=lmatRows(model, focus), lmat.scale.abs2=TRUE, estimate.sign=1, order.contrasts=TRUE, level=.95, calpha=NULL, alternative = c("two.sided", "less", "greater"), ... ) multicomp.mmc(x, ## S-Plus focus=dimnames(attr(x$terms,"factors"))[[2]][1], comparisons="mca", lmat, lmat.rows=lmatRows(x, focus), lmat.scale.abs2=TRUE, ry, plot=TRUE, crit.point, iso.name=TRUE, estimate.sign=1, x.offset=0, order.contrasts=TRUE, main, main2, focus.lmat, ...) ## S3 method for class 'mmc.multicomp': x[..., drop = TRUE]
model |
"aov" object in "lm" method. |
ylabel |
name of the response variable. |
lmat |
contrast matrix as in the S-Plus multicomp .
The convention for lmat in R is to use
the transpose of the linfct component produced by
glht . Required for user-specified contrasts. |
lmat.rows |
rows in lmat for the focus factor. |
focus |
define the factor to compute contrasts of.
See mcp in R. |
focus.lmat |
R only. Contrast matrix used in the user-specified
comparisons of the focus factor. This is the matrix the user
constructs. This matrix is multiplied by the lmat from the none
component to create the lmat for the user-specified contrasts.
Display the hibrido.lmat and maiz2.lmat in the maiz
example below to see what is happening.
|
linfct |
In R, see glht . |
... |
other arguments. alternative and
base are frequently used with glht . |
comparisons |
argument to multicomp |
lmat.scale.abs2 |
logical, scale the contrasts in the columns of
lmat to make the sum of the absolute values of each column equal 2. |
estimate.sign |
numeric. If 0 , leave contrasts in the
default lexicographic direction. If positive, force all contrasts to positive,
reversing their names if needed (if contrast A-B is negative, reverse it
to B-A). If negative, the force all contrasts to positive. |
order.contrasts |
sort the contrasts in the (mca , none ,
lmat ) components by height on the MMC plot. This will place the
contrasts in the multicomp plots in the same order as in the MMC plot. |
alternative |
Direction of alternative hypothesis.
See confint in R. S-Plus multicomp
uses the argument bounds for this concept.
|
level |
Confidence level. Defaults to 0.95. |
crit.point, calpha |
critical value for the tests. The value from the
specified multicomp method is used for the user-specified
contrasts when lmat is specified. This argument is called
crit.point with multicomp in S-Plus and calpha
when used with glht and confint in R.
In R, with a large number of levels for the focus factor, calpha
should
be specified. See notes below for discussion of the timing issues
and the examples for an illustration how to use calpha . |
plot |
logical, display the plot if TRUE . |
ry, iso.name, x.offset, main, main2 |
arguments to
plot.mmc.multicomp . |
x, drop |
See "[" . |
By default, if lmat
is not specified, we plot the isomeans grid
and the pairwise comparisons for the focus
factor. By default,
we plot the specified contrasts if the lmat
is specified.
Each contrast is plotted at a height which is the weighted average of
the means being compared. The weights are scaled to the sum of their
absolute values equals 2.
We get the right contrasts automatically if the aov is oneway. If we specify an lmat for oneway it must have a leading row of 0.
For any more complex design, we must study the lmat
from the mca
component of the result to see how to construct the lmat
(with the
extra rows as needed) and how to specify the lmat.rows
corresponding to the rows for the focus factor.
glht.mmc
in R works from either an "glht"
object or an
"aov"
object. multicomp.mmc
in S-Plus works from an
"aov"
object.
An "mmc.multicomp"
object contains either the first two or all
three of the "multicomp"
components mca
, none
,
lmat
described here. Each "multicomp"
component in
R also contains a "glht"
object.
mca |
Object containing the pairwise comparisons. |
none |
Object comparing each mean to 0. |
lmat |
Object for the contrasts specified in
the lmat argument. |
"[.mmc.multicomp"
is a subscript method.
The multiple comparisons calculations in R and S-Plus use
completely different functions.
MMC plots in R are constructed by glht.mmc
based on glht
.
MMC plots in S-Plus are constructed by
multicomp.mmc
based on the S-Plus multicomp
.
The MMC plot is the same in both systems. The details of getting the
plot differ.
Function glht.mmc
calls glht
and
confint.glht
. With a large number of levels
for the focus factor, the confint
function is exceedingly slow
(80 minutes for 30 levels on 1.5GHz Windows XP). Therefore,
always specify calpha
to reduce the time to under a second for
the same example.
plot.mmc.multicomp
chooses sensible defaults for its many
arguments. They will often need manual adjustment. The examples show
several types of adjustments. We have changed the centering and scaling
to avoid overprinting of label information. By default the significant
contrasts are shown in a more intense color than the nonsignificant
contrasts.
We have an option to reduce the color intensity of the isomeans grid.
Richard M. Heiberger <rmh@temple.edu>
Heiberger, Richard M. and Holland, Burt (2004b). Statistical Analysis and Data Display: An Intermediate Course with Examples in S-Plus, R, and SAS. Springer Texts in Statistics. Springer. ISBN 0-387-40270-5.
Heiberger, R.~M. and Holland, B. (2006). "Mean–mean multiple comparison displays for families of linear contrasts." Journal of Computational and Graphical Statistics, 15:937–955.
Hsu, J. and Peruggia, M. (1994). "Graphical representations of Tukey's multiple comparison method." Journal of Computational and Graphical Statistics, 3:143–161.
as.multicomp
, plot.mmc.multicomp
## Use glht.mmc with R. ## Use multicomp.mmc with S-Plus. ## data and ANOVA ## catalystm example catalystm <- read.table(hh("datasets/catalystm.dat"), header=FALSE, col.names=c("catalyst","concent")) catalystm$catalyst <- factor(catalystm$catalyst, labels=c("A","B","C","D")) if.R(r= bwplot(concent ~ catalyst, data=catalystm, scales=list(cex=1.5), ylab=list("concentration", cex=1.5), xlab=list("catalyst",cex=1.5)) ,s= t(bwplot(catalyst ~ concent, data=catalystm, scales=list(cex=1.5), xlab=list("concentration", cex=1.5), ylab=list("catalyst",cex=1.5))) ) catalystm1.aov <- aov(concent ~ catalyst, data=catalystm) summary(catalystm1.aov) catalystm.mca <- if.R(r=glht(catalystm1.aov, linfct = mcp(catalyst = "Tukey")), s=multicomp(catalystm1.aov, plot=FALSE)) ## plot(catalystm.mca) if.R(s=catalystm.mca, r=confint(catalystm.mca)) ## pairwise comparisons old.omd <- par(omd=c(0,.95,0,1)) catalystm.mmc <- if.R(r=glht.mmc(catalystm1.aov, linfct = mcp(catalyst = "Tukey")), s=multicomp.mmc(catalystm1.aov, plot=FALSE)) catalystm.mmc if.R(s=plot(catalystm.mmc, x.offset=1), r=plot(catalystm.mmc, ry=c(50,58), x.offset=1.8)) ## tiebreaker if.R(r=tmp.omd <- par(omd=c(par()$omd[1:2],.5,1)), s={}) plot.matchMMC(catalystm.mmc$mca, xlabel.print=FALSE) if.R(r=par(tmp.omd), s={}) ## user-specified contrasts 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$catalyst) catalystm.mmc <- if.R(r=glht.mmc(catalystm1.aov, linfct = mcp(catalyst = "Tukey"), focus.lmat=catalystm.lmat), s=multicomp.mmc(catalystm1.aov, focus.lmat=catalystm.lmat, plot=FALSE)) catalystm.mmc if.R(s=plot(catalystm.mmc, x.offset=1), r=plot(catalystm.mmc, ry=c(50,58), x.offset=1.8)) ## tiebreaker if.R(r=tmp.omd <- par(omd=c(par()$omd[1:2],.65,1)), s={}) plot.matchMMC(catalystm.mmc$lmat, xlabel.print=FALSE, col.signif='blue') if.R(r=par(tmp.omd), s={}) par(old.omd) ## Dunnett's test ## weightloss example weightloss <- read.table(hh("datasets/weightloss.dat"), header=TRUE) weightloss <- data.frame(loss=unlist(weightloss), group=rep(names(weightloss), rep(10,5))) if.R(r= bwplot(loss ~ group, data=weightloss, scales=list(cex=1.5), ylab=list("Weight Loss", cex=1.5), xlab=list("group",cex=1.5)) ,s= t(bwplot(group ~ loss, data=weightloss, scales=list(cex=1.5), xlab=list("Weight Loss", cex=1.5), ylab=list("group",cex=1.5))) ) weightloss.aov <- aov(loss ~ group, data=weightloss) summary(weightloss.aov) if.R(r={ group.count <- table(weightloss$group) },s={}) tmp.dunnett <- if.R(r= glht(weightloss.aov, linfct=mcp(group=contrMat(group.count, base=4)), alternative="greater") ,s= multicomp(weightloss.aov, method="dunnett", comparisons="mcc", bounds="lower", control=4, valid.check=FALSE) ) if.R(r=tmp.omd <- par(omd=c(0,1,.5,1)), s={}) plot(tmp.dunnett) if.R(r=par(tmp.omd), s={}) tmp.dunnett.mmc <- if.R(r= glht.mmc(weightloss.aov, linfct=mcp(group=contrMat(group.count, base=4)), alternative="greater") ,s= multicomp.mmc(weightloss.aov, method="dunnett", comparisons="mcc", bounds="lower", control=4, valid.check=FALSE, plot=FALSE) ) tmp.dunnett.mmc plot(tmp.dunnett.mmc) ## two-way ANOVA ## display example display <- read.table(hh("datasets/display.dat"), header=TRUE) display$panel <- factor(display$panel) position(display$panel) <- (1:3)+.5 if.R(r={ contrasts(display$panel) <- "contr.treatment" } ,s={}) display$emergenc <- factor(display$emergenc) if.R(s={old.omd <- par(omd=c(.08,1,.05,1))}, r={}) interaction2wt(time ~ emergenc * panel, data=display) if.R(s={par(old.omd)}, r={}) displayf.aov <- aov(time ~ emergenc * panel, data=display) anova(displayf.aov) ## multiple comparisons ## MMC plot displayf.mmc <- if.R(r={ glht.mmc(displayf.aov, linfct=mcp(panel="Tukey"), interaction.average=TRUE) }, s=multicomp.mmc(displayf.aov, "panel", plot=FALSE)) if.R(s= plot(displayf.mmc) ,r= plot(displayf.mmc, x.offset=1.7, ry=c(17.8,25.8)) ) panel.lmat <- cbind("3-12"=c(-1,-1,2), "1-2"=c( 1,-1,0)) dimnames(panel.lmat)[[1]] <- levels(display$panel) displayf.mmc <- if.R(r={ glht.mmc(displayf.aov, linfct=mcp(panel="Tukey"), interaction.average=TRUE, focus.lmat=panel.lmat) }, s=multicomp.mmc(displayf.aov, "panel", focus.lmat=panel.lmat, plot=FALSE)) if.R(s= plot(displayf.mmc) ,r= plot(displayf.mmc, x.offset=1.7, ry=c(17.8,25.8)) ) ## split plot design with tiebreaker plot ## ## This example is based on the query by Tomas Goicoa to R-news ## http://article.gmane.org/gmane.comp.lang.r.general/76275/match=goicoa ## It is a split plot similar to the one in HH Section 14.2 based on ## Yates 1937 example. I am using the Goicoa example here because its ## MMC plot requires a tiebreaker plot. maiz <- read.table(hh("datasets/maiz.dat"), header=TRUE) maiz$hibrido <- factor(maiz$hibrido, levels=c("P3747","P3732","Mol17","A632","LH74")) maiz$nitrogeno <- factor(maiz$nitrogeno) position(maiz$nitrogeno) <- c(1.2, 2.4, 3.6, 4.8) ## inherits(maiz$nitrogeno, "ordered") position(maiz$bloque) <- c(2.25, 3.75) ## inherits(maiz$bloque, "ordered") if.R(s={old.omd <- par(omd=c(.1,1,.05,1))}, r={}) interaction2wt(yield ~ hibrido+nitrogeno+bloque, data=maiz) interaction2wt(yield ~ hibrido+nitrogeno, data=maiz) if.R(s={par(old.omd)}, r={}) maiz.aov <- aov(yield ~ nitrogeno*hibrido + Error(bloque/nitrogeno), data=maiz) summary(maiz.aov) summary(maiz.aov, split=list(hibrido=list(P3732=1, Mol17=2, A632=3, LH74=4))) ## multicomp(maiz.aov, focus="hibrido") ## can't use 'aovlist' objects ## glht(maiz.aov, linfct=mcp(hibrido="Tukey")) ## can't use 'aovlist' objects sapply(maiz[-1], contrasts) if.R(r={ ## R glht.mmc requires treatment contrasts contrasts(maiz$nitrogeno) <- "contr.treatment" contrasts(maiz$bloque) <- "contr.treatment" sapply(maiz[-1], contrasts) }, s={}) ## Both R glht() and S-Plus multicomp() require aov, not aovlist maiz2.aov <- aov(terms(yield ~ bloque*nitrogeno + hibrido/nitrogeno, keep.order=TRUE), data=maiz) summary(maiz2.aov) ## There are many ties in the group means. ## These are easily seen in the MMC plot, where the two clusters ## c("P3747", "P3732", "LH74") and c("Mol17", "A632") ## are evident from the top three contrasts including zero and the ## bottom contrast including zero. The significant contrasts are the ## ones comparing hybrids in the top group of three to ones in the ## bottom group of two. ## We have two graphical responses to the ties. ## 1. We constructed the tiebreaker plot. ## 2. We construct a set of orthogonal contrasts to illustrate ## the clusters. ## pairwise contrasts with tiebreakers. if.R(s={ maiz2.mmc <- multicomp.mmc(maiz2.aov, focus="hibrido", plot=FALSE) old.omd <- par(omd=c(.05,.85,0,1)) plot(maiz2.mmc, ry=c(145,170), x.offset=4) par(omd=c(.05,.85,0,1)) plot.matchMMC(maiz2.mmc$mca) par(old.omd) },r={ maiz2.mmc <- glht.mmc(maiz2.aov, linfct=mcp(hibrido="Tukey"), interaction.average=TRUE) old.omd <- par(omd=c(.05,.85,.35,1)) ## x1 x2 y1 y2 plot(maiz2.mmc) par(omd=c(.05,.85,0,.5), new=TRUE) plot.matchMMC(maiz2.mmc$mca, cex.axis=.7) par(old.omd) }) ## orthogonal contrasts ## user-specified contrasts hibrido.lmat <- cbind("PPL-MA" =c(2, 2,-3,-3, 2), "PP-L" =c(1, 1, 0, 0,-2), "P47-P32"=c(1,-1, 0, 0, 0), "M-A" =c(0, 0, 1,-1, 0)) dimnames(hibrido.lmat)[[1]] <- levels(maiz$hibrido) hibrido.lmat maiz2.mmc <- if.R(s=multicomp.mmc(maiz2.aov, focus="hibrido", focus.lmat=hibrido.lmat, plot=FALSE), r=glht.mmc(maiz2.aov, linfct=mcp(hibrido="Tukey"), interaction.average=TRUE, focus.lmat=hibrido.lmat) ) if.R(s={ old.omd <- par(omd=c(.05,.85,0,1)) plot(maiz2.mmc, ry=c(145,170), x.offset=4) par(omd=c(.05,.85,0,1)) plot.matchMMC(maiz2.mmc$lmat, col.signif='blue') par(old.omd) },r={ old.omd <- par(omd=c(.05,.85,.35,1)) ## x1 x2 y1 y2 plot(maiz2.mmc) par(omd=c(.05,.85,0,.45), new=TRUE) plot.matchMMC(maiz2.mmc$lmat, cex.axis=.7, col.signif='blue') par(old.omd) })