simulation: Simulation studies

In mutoss: Unified Multiple Testing Procedures

Simulation studies

Description

This function generates data according to a specified function and parameters and then applies specified procedures to the generated data. The generated data is stored in $data and the results are stored in $results. In order to recognize which results and generated data belong together every result contains an element $data.set.number. The data generating function must return a list which contains a list $procInput. Every element in $procInput will be used as an input parameter for the procedures provided by listOfProcedures.

Usage

simulation(replications, DataGen, listOfProcedures, discardProcInput=FALSE)

Arguments

replications	The number of replications. This means how many simulation runs will be performed.
DataGen	A list that contains the function and parameters for generating data which will be analyzed be the procedures in listOfProcedures. It must contain the elements $funName (character) $fun (function)
listOfProcedures	A list of lists which contains the procedures with their parameters for the simulation.
discardProcInput	A list of lists which contains the procedures with their parameters

Value

A list with 2 elements. $data contains the objects generated by DataGen. $results contains the objects generated by the procedures augmented by the data set number and the parameter constellation.

Author(s)

MarselScheer

Examples

# this function generates pValues 
myGen <- function(n, n0) {
  list(procInput=list(pValues = c(runif(n-n0, 0, 0.01), runif(n0))), 
  groundTruth = c(rep(FALSE, times=n-n0), rep(TRUE, times=n0)))
}

sim <- simulation(replications = 10, list(funName="myGen", fun=myGen, n=200, n0=c(50,100)), 
list(list(funName="BH", 
  fun=function(pValues, alpha) BH(pValues, alpha, silent=TRUE), alpha=c(0.25, 0.5)),
list(funName="holm", fun=holm, alpha=c(0.25, 0.5),silent=TRUE)))

# the following happend:
# Call myGen(200,50) and append result to sim$data
# Apply bonferroni and holm each with alpha 0.25 and 0.5 to this data set
# Append the results to sim$restults
# Repeat this 10 times.
# Call myGen(200, 100) and append restult to sim$data
# Apply bonferroni and holm each with alpha 0.25 and 0.5 to this data set
# Append the results to sim$restults
# Repeat this 10 times.

length(sim$data)
length(sim$results)

# we now reproduce the 6th item in results
print(sim$results[[6]]$data.set.number)
print(sim$results[[6]]$parameters)

all(BH(sim$data[[2]]$procInput$pValues, 0.5, silent=TRUE)$adjPValues == sim$results[[6]]$adjPValues)

#
# Just calculating some statistics and making some plots
NumberOfType1Error <- function(data, result) sum(data$groundTruth * result$rejected)
result.all <- gatherStatistics(sim, list(NumOfType1Err = NumberOfType1Error))
result <- gatherStatistics(sim, list(NumOfType1Err = NumberOfType1Error), 
  list(median=median, mean=mean, sd=sd))
print(result)
require(lattice)
histogram(~NumOfType1Err | method*alpha, data = result.all$statisticDF)
barchart(NumOfType1Err.median + NumOfType1Err.mean ~ method | alpha, data = result$statisticDF)

mutoss documentation built on March 31, 2023, 8:46 p.m.