mooflow.moo_algos.Mocmaes

class mooflow.moo_algos.Mocmaes(config)

Class to setup and run the optimization.

Parameters

configdict

with the following fields:

“num_parameter”

number of optimimzation parameters;

“num_fitness”

number of fitness functions;

“pop_size”

size of the population (for MOCMAES not so big)

“fitness_weights”

list with length num_fitness, -1 for minimization, 1 for maximization;

“target_folder”:

base folder in which the model folder resides;

“modelname”

the name of the model folder

“lower_boundary”

lower box constraint (better not change), default 0;

“upper_boundary”

upper box constraint (better not change), default 1;

“num_threads”

number of threads to use. should exceed the number of physical cores, default: 2;

“init_population”

set the parameters for the inidividuals of the initial population, not tested. numpy array with dimension pop x parameter, default False;

“seed”

set the seed for the random number generator;

__init__(config)

Initialize self. See help(type(self)) for accurate signature.

Methods

__init__(config)	Initialize self.
run()	run the multi-objective optimization with MOCMAES
setup_modflow_parallel([new_basefolder])	Duplicate the modflow model for parallel evaluation.