parallelizeMCsimulation {TempStable}

Function to parallelize the Monte Carlo Simulation

Description

Since the Monte Carlo Simulation is very computationally intensive, it may be worthwhile to split it across all available processor cores. To do this, simply pass all the parameters from the TemperedEstim_Simulation() function to this function in the same way.

Usage

parallelizeMCsimulation(
  ParameterMatrix,
  MCparam = 10000,
  SampleSizes = c(200),
  saveOutput = FALSE,
  cores = 2,
  SeedOptions = NULL,
  iterationDisplayToFileSystem = FALSE,
  ...
)

Arguments

ParameterMatrix

The matrix is to be composed of vectors, row by row. Each vector must fit the pattern of theta of the TemperedType. Compared to the function TemperedEstim_Simulation(), the matrix here may contain only one parameter vector.

MCparam

Number of Monte Carlo simulation for each couple of parameter, default=100; integer

SampleSizes

Sample sizes to be used to simulate the data. By default, we use 200 (small sample size). Vector of integer. Compared to the function TemperedEstim_Simulation(), the vector here may contain only one integer.

saveOutput

Logical flag: In the function TemperedEstim_Simulation() the argument can be true. Then an external csv file is created. Here the argument must be false. The output of the values works in this function exclusively via the return of the function.

cores

size of cluster for parallelization. Positive Integer.

SeedOptions

is an argument what can be used in TemperedEstim_Simulation() but must be NULL here.

iterationDisplayToFileSystem

creates a text file in your file system that displays the current iteration of the simulation.

...

The function works only if all necessary arguments from the function TemperedEstim_Simulation() are passed. See description and details.

Details

In this function exactly the arguments must be passed, which are also needed for the function TemperedEstim_Simulation(). However, a few functions of TemperedEstim_Simulation() are not possible here. The restrictions are described in more detail for the individual arguments.

In addition to the arguments of function TemperedEstim_Simulation(), the argument "cores" can be assigned an integer value. This value determines how many different processes are to be parallelized. If value is NULL, R tries to read out how many cores the processor has and passes this value to "cores".

During the simulation, the progress of the simulation can be viewed in a file in the workspace named "IterationControlForParallelization.txt".

Value

The return object is a list of 2. Results of the simulation are listed in $outputMat.