Create a table of comparisons based on the cgOneFactorFit object. Pairwise or custom specified contrasts are estimated and tested. A cgOneFactorComparisonsTable class object is created.
# S4 method for cgOneFactorFit
comparisonsTable(fit, type="pairwisereflect",
alpha=0.05, addpct=FALSE, display="print", ...)Creates an object of class cgOneFactorComparisonsTable, with the
following slots:
ols.comprsThe table of comparisons based on the
olsfit component of the cgOneFactorFit,
unless model="rronly" is specified. In that case the slot
value is NULL. Will not be appropriate in
the case where a valid aftfit component is present in the
cgOneFactorFit object. See below for the data frame structure
of the table.
rr.comprsThe table of comparisons based on the
rrfit component of the cgOneFactorFit
object, if a valid resistant & robust fit object is present.
If rrfit is a simple character value of
"No fit was selected.", or model="olsonly" was
specified, then the value is NULL. See below for the data frame structure
of the table.
aft.comprsThe table of comparisons based on the
aftfit component of the cgOneFactorFit
object if a valid accelerated failure time fit object is present.
If aftfit is a simple character value of
"No fit was selected.", then the value is NULL.
See below for the data frame structure
of the table.
uv.comprsThe table of comparisons based on the
uvfit component of the cgOneFactorFit
object if a valid unequal variances fit object is present.
The error degrees of freedom for each comparison estimate and
test is individually estimated
with a Satterthwaite approximation. See below for the data frame structure
of the table.
settingsA list of settings carried from the
cgOneFactorFit fit object, and the addition
of some specified arguments in the method call above: alpha,
mcadjust, type, and addpct. These are used
for the print.cgOneFactorComparisonsTable method,
invoked for example when
display="print".
The data frame structure of the comparisons table in a *.comprs
slot consists of row.names that specify the comparison of the
form A vs. B, and these columns:
estimateThe difference in group means in the
comparison: A vs. B. If settings$endptscale=="log" in the
fit object, this will be back-transformed to a percent
difference scale.
seThe estimated standard error of the difference
estimate. If settings$endptscale=="log" in the
fit object, this estimate will be based on the Delta
method, and will particularly begin to be a poor approximation when the
standard error in the logscale exceeds 0.50.
lowerciThe lower 100 * (1-alpha) % confidence limit of the
difference estimate. With the default alpha=0.05,
this is 95%. If settings$endptscale=="log" in the
fit object, the confidence limit is first computed in the
logarithmic scale of analysis, and then back-transformed to a percent
difference scale.
upperciThe upper 100 * (1-alpha) % confidence limit of the
difference estimate. With the default alpha=0.05,
this is 95%. If settings$endptscale=="log" in the
fit object, the confidence limit is first computed in the
logarithmic scale of analysis, and then back-transformed to a percent
difference scale.
pvalThe computed p-value from the test of the difference estimate.
meanA or geomeanAThe estimated mean for the
left hand side "A" of the A vs. B comparison.
If settings$endptscale=="log" in the
fit object, this is a back-transform to the original scale,
and therefore is a geometric mean, and will be labelled
geomeanA.
Otherwise it is the arithmetic mean and labelled meanA.
seAThe estimated standard error of the meanA
estimate. If settings$endptscale=="log" in the
fit object, this estimate will be based on the Delta
method, and will particularly begin to be a poor approximation when the
standard error in the logscale exceeds 0.50.
meanB or geomeanBThe estimated mean for the
right hand side "B" of the A vs. B comparison.
If settings$endptscale=="log" in the
fit object, this is a back-transform to the original scale,
and therefore is a geometric mean, and will be labelled
geomeanB.
Otherwise it is the arithmetic mean and labelled meanB.
seBThe estimated standard error of the meanB
estimate. If settings$endptscale=="log" in the
fit object, this estimate will be based on the Delta
method, and will particularly begin to be a poor approximation when the
standard error in the logscale exceeds 0.50.
An additional column addpct of percent differences is added if
endptscale=="original" and addpct=TRUE,
as a descriptive supplement to the original scale
differences that are formally estimated.
An object of class cgOneFactorFit.
Can be one of four values:
"pairwisereflect"The default value, it calculates and lists all possible pairwise comparison permutations, as each pair order is included. In other words, Groups A vs. B and B vs. A will be included.
"pairwise"Calculates and lists all possible
pairwise comparison combinations. Once a pair such as Groups
A vs. B is specified, then the reflective B vs. A is not
included. So the number of comparisons is half that produced
by "pairwisereflect". The ordering of group levels in the
fit object is used to determine which
ordering is included and which is not. If all orderings are
of interest, such as for settings$endptscale=="log" in the fit
objects, use the "pairwisereflect" value above.
"allgroupstocontrol"Takes the value of
settings$refgrp in the cg fit
object, deems it the "control" group, and constructs
pairwise comparisons of all other groups to it. This setting is
required when the refgrp argument is specified in the
call (see Additional Arguments section below.)
"custom"Indicates the a custom matrix of
comparisons will be constructed, and that matrix needs to be
specified in the contrastmatrix argument.
Significance level, by default set to 0.05.
Only relevant if settings$endptscale=="original" in the
fit object. An column of percent differences is added for the
comparisons, as a descriptive supplement to the original scale
differences that are formally estimated.
One of three valid values:
"print"The default value; It calls a print method for the
created cgOneFactorComparisonsTable object, which is
a formatted text output of the table(s).
"none"Supresses any printing. Useful, for example, when just assignment of the resulting object is desired.
"show"Calls the default showDefault method, which
will just print out the cgOneFactorComparisonsTable components.
Additional arguments.
mcadjust Do a multiple comparisons adjustment, based on the simultaneous
inference capabilities of the multcomp package. See Details
below. The default value is FALSE.
If mcadjust=TRUE is specified, there will be a delay,
usually just for a few seconds, due to computing time of the
critical point in order to conduct the adjusted comparisons.
contrastmatrix Only relevant if type="custom" is
specified. In that case, a numeric matrix with the number of
rows equal to the number of comparisons of interest. The number
of columns must be equal to the number of group means. Each row in
the matrix is assumed to represent a contrast of coefficients
amongst the groups that defines the comparison of interest.
refgrp If left at the default value of NULL, it will
be set to the settings$refgrp value in the cg fit
object. When set, it is deemed the "reference", or "control" group, so that
pairwise comparisons of all other groups to it will be
constructed when type="allgroupstocontrol". Please note the
type="allgroupstocontrol" setting is
REQUIRED when the refgrp argument is specified in the
call with a valid non-NULL value.
model For cgOneFactorFit fit objects that have
classical least squares lm or resistant & robust
rlm fits, the following argument values are possible:
"both"Comparison tables based on both the ordinary
classical least squares and resistant & robust fits are
created. This is the default when both fits are present in
the cgOneFactorFit object specified in the fit
argument. If the resistant & robust fit is not available,
this value is not relevant.
"olsonly"Only a comparison table based on the ordinary
classical least squares olsfit fit slot is performed.
"rronly"Only a comparison table based on the resistant and
robust rrfit fit slot is performed.
For other possible cgOneFactorFit fit components such as
accelerated failure time or unequal variance models, the model
argument is not relevant, and the appropriate comparisons table will
be calculated for these model types.
Bill Pikounis [aut, cre, cph], John Oleynick [aut], Eva Ye [ctb]
When mcadjust=TRUE, a status message of "Some time may be
needed as the critical point"
"from the multcomp::summary.glht function
call is calculated" is displayed at the console. This computed critical point
is used for all subsequent p-value and confidence interval
calculations.
The multcomp package provides a unified way to calculate critical points based on the comparisons of interest in a "family". Thus a user does not need to worry about choosing amongst the myriad names of multiple comparison procedures.
Hothorn, T., Bretz, F., Westfall, P., Heiberger, R.M., and
Schuetzenmeister, A. (2010). The multcomp package.
Hothorn, T., Bretz, F., and Westfall, P. (2008). "Simultaneous Inference in General Parametric Models", Biometrical Journal, 50, 3, 346-363.
data(canine)
canine.data <- prepareCGOneFactorData(canine, format="groupcolumns",
analysisname="Canine",
endptname="Prostate Volume",
endptunits=expression(plain(cm)^3),
digits=1, logscale=TRUE, refgrp="CC")
canine.fit <- fit(canine.data)
canine.comps0 <- comparisonsTable(canine.fit)
canine.comps1 <- comparisonsTable(canine.fit, mcadjust=TRUE,
type="allgroupstocontrol", refgrp="CC")
data(gmcsfcens)
gmcsfcens.data <- prepareCGOneFactorData(gmcsfcens, format="groupcolumns",
analysisname="cytokine",
endptname="GM-CSF (pg/ml)",
logscale=TRUE)
gmcsfcens.fit <- fit(gmcsfcens.data, type="aft")
gmcsfcens.comps <- comparisonsTable(gmcsfcens.fit)
Run the code above in your browser using DataLab