Monte Carlo Iteration

Overview

These commands collectively control the ability to perform a set of simulation runs in a single execution of the program, where each run starts with a different random number seed. This is often called “Monte Carlo iteration.’

This concept can be thought of as a loop of the form:

runNumber = initial_run_number
while (runNumber <= final_run_number)
{
   SetSeed(random_seed[runNumber])
   execute simulation
   runNumber = runNumber + run_number_increment
}

Many WSF simulations support Monte Carlo integration and already implement the above concept in their code.

Commands

final_run_number <positive-integer>

Set the run number of the final run of the set.

Default: 1

generate_random_seeds <positive-integer>

This command causes a set of random number seeds to be generated for performing a set of simulation runs. The value is used to specify the initial seed to be used for generating the random number seeds. A vector of random number seeds will be generated when the first run is performed, with the number of seeds generated equal to the value of final_run_number. The random number generator for a particular run will be initialized at the start of using the random number seed associated with the current run number.

The default is to not generate random seeds, in which case the random number generator will not be initialized at the start of each run.

initial_run_number <positive-integer>

Set the run number of the initial run of the set. This command, along with final_run_number is useful for either repeating a specific run or it can be used with the run_number_increment to partition a Monte Carlo set among multiple computers (see run_number_increment for an example).

Default: 1

run_number_increment <positive-integer>

Set the increment for the runs to be executed. This command is useful for subdividing a large number of Monte Carlo iterations among multiple computers.

Default: 1

For example, if you wanted to run 1000 iterations using 3 computers:

final_run_number 1000
run_number_increment 3
initial_run_number <computer-number (1, 2, 3 )>

  • The first computer will execute iterations 1, 4, 7, …, 1000

  • The second computer will execute iterations 2, 5, 8, …, 998

  • The third computer will execute iterations 3, 6, 9, …, 999