predeterminedruns_driver.py

Baseclass for design-of-experiments Drivers that have pre-determined parameter sets.

class openmdao.drivers.predeterminedruns_driver.PredeterminedRunsDriver(num_par_doe=1, load_balance=False)[source]

Bases: openmdao.core.driver.Driver

Baseclass for design-of-experiments Drivers that have pre-determined parameter sets.

Args:

num_par_doe : int, optional

The number of DOE cases to run concurrently. Defaults to 1.

load_balance : bool, Optional

If True and running under MPI, use rank 0 as master and load balance cases among all of the other ranks. Default is False. If multiprocessing is being used instead of MPI, then cases are always load balanced.
get_req_procs()[source]
Returns:

tuple

A tuple of the form (min_procs, max_procs), indicating the min and max processors usable by this Driver.
run(problem)[source]

Build a runlist and execute the Problem for each set of generated parameters.

openmdao.drivers.predeterminedruns_driver.worker(problem, response_vars, case_queue, response_queue, worker_id)[source]

This is used to run parallel DOEs using multprocessing. It takes a case off of the case_queue, runs it, then puts responses on the response_queue.