Controller

class Controller.Controller

Bases: object

The MonteCarloController class is used to run a monte carlo simulation. It is used to execute multiple runs of a simulation with varying initial parameters. Data from each run is retained in order to analyze differences in the simulation runs and the parameters used.

addDispersion(disp)

Add a dispersion to the simulation. :param disp: Dispersion

The dispersion to add to the simulation.

addRetentionPolicy(policy)

Add a retention policy to the simulation. :param disp: RetentionPolicy

The retention policy to add to the simulation. This defines variables to be logged and saved

executeCallbacks(rng=None, retentionPolicies=[])

Execute retention policy callbacks after running a monteCarlo sim. rng: A list of simulations to execute callbacks on retentionPolicies: the retention policies to execute

executeSimulations()

Execute simulations in parallel Args: None :returns:

int[]

A list of the indices of all failed simulation runs.

Return type

failed

generateICSims(caseList)

Generator function to clone a baseSimulation for IC run :param baseSimulation: SimulationParams

A base simulation to clone.

Parameters

numSims – int[] The desired runs to generate.

Returns

generator<SimulationParams>

A generator that yields that number of cloned simulations

generateSims(simNumList)

Generator function to clone a baseSimulation :param baseSimulation: SimulationParams

A base simulation to clone.

Parameters

numSims – int[] The desired runs to generate.

Returns

generator<SimulationParams>

A generator that yields that number of cloned simulations

getParameters(caseNumber)

Get the parameters used for a particular run of the montecarlo :param caseNumber: int

The number of the run to get the parameters used for.

Returns

A dictionary of the parameters of the simulation For example: {

”keyForSim”: parameterValue, ‘TaskList[0].TaskModels[0].RNGSeed’: 1674764759

}

getRetainedData(case)

Get the data that was retained for a run, or list of runs. :param cases: int The desired case to get data from.

Returns

The retained data for that run is returned.

getRetainedDatas(cases)

Get the data that was retained for a list of runs. :param cases: int[] The desired cases to get data from.

Returns

A generator is returned, which will yield, in-order, the retained data for each of these cases

static load(runDirectory)

Load a previously completed MonteCarlo simulation :param The path to the MonteCarlo.data file that contains the archived MonteCarlo run:

reRunCases(caseList)

Rerun some cases from a MonteCarlo run. Does not run in parallel :param caseList: int[]

The list of runs to repeat, a list of numbers.

Returns

int[]

The list of failed runs.

Return type

failures

runInitialConditions(caseList)

Run initial conditions given in a file :param caseList: int[]

The list of runs to repeat, a list of numbers.

Returns

int[]

The list of failed runs.

Return type

failures

setArchiveDir(dirName)

Set-up archives for this MonteCarlo run :param dirName: string

The name of the directory to archive runs in. None, if no archive desired.

setConfigureFunction(newModule)

Set an execution function that executes a simulation instance. :param executionFunction: (sim: SimulationBaseClass) => None

A function with one parameter, a simulation instance. The function will be called after the creationFunction and configurationFunction in each simulation run. It must execute the simulation. Its return value is not used.

setDispMagnitudeFile(magnitudes)

Save .txt with the magnitude of each dispersion in % or sigma away from mean :param magnitudes: bool

Whether to save extra files for analysis.

setExecutionCount(newCount)

Set the number of runs for the MonteCarlo simulation :param newCount: int

The number of runs to use for the simulation

setExecutionFunction(newModule)

Set an execution function that executes a simulation instance. :param executionFunction: (sim: SimulationBaseClass) => None

A function with one parameter, a simulation instance. The function will be called after the creationFunction and configurationFunction in each simulation run. It must execute the simulation. Its return value is not used.

setICDir(dirName)

Set-up archives containing IC data :param dirName: string

The name of the directory to archive runs in. None, if no archive desired.

setICRunFlag(bool)

Set the number of threads to use for the monte carlo simulation :param threads: int

Number of threads to execute the montecarlo run on.

setShouldDisperseSeeds(seedDisp)

Disperse the RNG seeds of each run in the MonteCarlo :param seedDisp: bool

Whether to disperse the RNG seeds in each run of the simulation

setSimulationFunction(newObject)

Set the function that creates the simulation instance. :param creationFunction: () => SimulationBaseClass

A function with no parameters, that returns a simulation instance.

setThreadCount(threads)

Set the number of threads to use for the monte carlo simulation :param threads: int

Number of threads to execute the montecarlo run on.

setVarCast(varCast)

Set the variable type to downcast the data to :param varCast: ‘float’, ‘integer’, ‘signed’, ‘unsigned’ (see pandas.to_numeric documentation) :return:

setVerbose(verbose)

Use verbose output for this MonteCarlo run :param verbose: bool

Whether to print verbose information during this MonteCarlo sim.

class Controller.SimulationExecutor

Bases: object

This class is used to execute a simulation in a worker thread. To use, create an instance of this class, and then call the instance with the simulation parameters to run them in.

executor = SimulationExecutor() simParams = SimulationParameters() successFlag = executor(simParams)

This class can be used to execute a simulation on a different thread, by using this class as the processes target.

static disperseSeeds(simInstance)

disperses the RNG seeds of all the tasks in the sim, and returns a statement that contains the seeds :param simInstance: SimulationBaseClass

A basilisk simulation to set random seeds on

Returns

string

A dictionary with the random seeds that should be applied to the sim: Example: “” {

’.TaskList[0].TaskModels[1]’: 1934586, ‘.TaskList[0].TaskModels[2]’: 3450093, ‘.TaskList[1].TaskModels[0]’: 2221934, ‘.TaskList[2].TaskModels[0]’: 1123244

Return type

statement

static populateSeeds(simInstance, modifications)

only populate the RNG seeds of all the tasks in the sim :param simInstance: SimulationBaseClass

A basilisk simulation to set random seeds on

Parameters

modifications – A dictionary containing RNGSeeds to be populate for the sim, among other sim modifications.

class Controller.SimulationParameters(creationFunction, executionFunction, configureFunction, retentionPolicies, dispersions, shouldDisperseSeeds, shouldArchiveParameters, filename, icfilename, index=None, verbose=False, modifications={})

Bases: object

This class represents the run parameters for a simulation, with information including

• a function that creates the simulation

• a function that executes the simulation

• the dispersions to use on that simulation

• parameters describing the data to be retained for a simulation

• whether randomized seeds should be applied to the simulation

• whether data should be archived