compareMutoss: Functions for comparing outputs of different procedures.

In mutoss: Unified multiple testing procedures

Description

Functions for comparing outputs of different procedures.

Usage

compareMutoss(...)
mu.compare.adjusted(comparison.list, identify.check=F)
mu.compare.critical(comparison.list, identify.check=F)
mu.compare.summary(comparison.list)

Arguments

...	An arbitrary number of Motoss class objects.
comparison.list	The output of the compareMutoss function.
identify.check	Logical parameter specifying if hypotheses should be identified on the output plots.

Details

These functions are used to compare the results of different multiple comparisons procedures stored as Mutoss class objects. compareMutoss takes as input an arbitrary number of Mutoss objects and arranges them in a simple list objects (non S4). mu.compare.adjuted, mu.compare.critical and mu.compare.summary take the output of the compareMutoss and plots or summerize the results textually or graphically.

Value

compareMutoss	Returns a list with the following components: types: Character vector of error types corrsponding to each procedure. rates: Numeric vector of error rates used for each procedure. pi.nulls: Numeric vector of estimates of the proportion of true null hypothesis if avilable. raw.pValues: The raw p-values used for each procedure. adjusted.pvals: Data frame with columns holding procedure specific adjusted p values. criticalValue: Data frame with columns holding the critical values corresponding to each procedure and error rate. rejections: Data frame with columns holding logical vectors of rejected hypotheses (TRUE for rejected).
mu.compare.adjusted	Creates a plot with the adjusted p-values for each procedure.
mu.compare.critical	Creates a plot with the critical values for each procedure and error rate.
mu.compare.summary	Creates a short textual summary for comparing results of different procedures.

Author(s)

Jonathan Rosenblatt

Examples

# TODO: EXAMPLE PROBLEMS
## Not run: 
\dontrun{Creating several Mutoss class objects}
mu.test.obj.1 <- mutoss.apply(new(Class="Mutoss", pValues=runif(10)), f=bonferroni, label="Bonferroni Correction", alpha=0.05, silent=T)#' 
mu.test.obj.2 <- mutoss.apply(new(Class="Mutoss", pValues=runif(10)), f=holm, label="Holm's step-down-procedure", alpha=0.05, silent=T)#' 
mu.test.obj.3 <- mutoss.apply(new(Class="Mutoss", pValues=runif(10)), f=aorc, label="Asymtotically optimal rejection curve", alpha=0.05, startIDX_SUD = 1, silent=T)
\dontrun{Trying to coercing a non-Mutoss object}
compareMutoss(1)
\dontrun{ Coercing several objects into a list}
compare.1<- compareMutoss(mu.test.obj.1, mu.test.obj.2)
compare.2<- compareMutoss(mu.test.obj.1, mu.test.obj.2, mu.test.obj.3)
\dontrun{Plotting the adjusted pvalues. Identification available.}
mu.compare.adjusted(compare.1, T)
mu.compare.adjusted(compare.2, T)
\dontrun{Plotting the critical values. Identification available.}
mu.compare.critical(compare.1, T)
mu.compare.critical(compare.2, T)
\dontrun{Showing a textual sumary}
mu.compare.summary(compare.1)
mu.compare.summary(compare.2)

## End(Not run)

mutoss documentation built on May 2, 2019, 5:56 p.m.