Problem

Class for defining problem.

Problem

Problem(
    model,
    weather_input,
    /,
    model_inputs=(),
    outputs=(),
    *,
    evaluation_dir=None,
    has_templates=False,
    noise_sample_kwargs=None,
    clean_patterns=_OutputManager._DEFAULT_CLEAN_PATTERNS,
    removes_subdirs=False,
)

Define a parametrics/optimisation problem.

PARAMETER	DESCRIPTION
model	Model file path. TYPE: str or path-like object
weather_input	Weather input variable. TYPE: WeatherModifier
model_inputs	Model input variables. TYPE: iterable of ModelModifier DEFAULT: ()
outputs	Output variables. TYPE: iterable of Collector DEFAULT: ()
evaluation_dir	Evaluation directory path, default is a directory named 'evaluation' in the same folder as the model file. TYPE: str or path-like object DEFAULT: None
has_templates	Whether the model has HVAC templates. TYPE: bool DEFAULT: `False`
noise_sample_kwargs	Settings for sampling uncertain variables, including Key Description Value type 'mode' Sampling mode {'elementwise', 'cartesian', 'auto'} 'size' Sample size int 'method' Sampling method {'random', 'latin hypercube'} 'seed' Random seed int If 'mode'='elementwise', 'size' and 'method' are mandatory. If 'mode'='cartesian', all model inputs must be non-continuous variables, If 'mode'='auto', the sampling mode is set to 'cartesian' if all variables are non-continuous, otherwise 'elementwise'. TYPE: dict DEFAULT: None
clean_patterns	Patterns to clean simulation files. This is ignored if removes_subdirs is set to True. TYPE: str or iterable of str DEFAULT: `{'*.audit', '*.end', 'sqlite.err'}`
removes_subdirs	Whether to remove the subdirectories in the evaluation directory. TYPE: bool DEFAULT: `False`

run_random

run_random(size, /, *, mode='auto', seed=None)

Run parametrics via a random sample.

PARAMETER	DESCRIPTION
size	Sample size. TYPE: int
mode	Sampling mode. TYPE: ('elementwise', 'cartesian', 'auto') DEFAULT: 'elementwise'
seed	Random seed. TYPE: int DEFAULT: None

run_latin_hypercube

run_latin_hypercube(size, /, *, seed=None)

Run parametrics via a latin hypercube sample.

PARAMETER	DESCRIPTION
size	Sample size. TYPE: int
seed	Random seed. TYPE: int DEFAULT: None

run_exhaustive

run_exhaustive()

Run parametrics via the exhaustive sample.

run_nsga2

run_nsga2(
    population_size,
    termination,
    /,
    *,
    p_crossover=1.0,
    p_mutation=0.2,
    init_population_size=0,
    saves_history=True,
    checkpoint_interval=0,
    seed=None,
)

Run optimisation via the NSGA2 algorithm.

PARAMETER	DESCRIPTION
population_size	Population size. TYPE: int
termination	Termination criterion, see https://pymoo.org/interface/termination.html. TYPE: Termination
p_crossover	Crossover probability. TYPE: float DEFAULT: `1.0`
p_mutation	Mutation probability. TYPE: float DEFAULT: `0.2`
init_population_size	Initial population size. This allows setting a different (usually larger) population size for the first generation. Any non-positive value falls back to the value of population_size. TYPE: int DEFAULT: `0`
saves_history	Whether to save the iteration history, see https://pymoo.org/interface/minimize.html. TYPE: bool DEFAULT: `True`
checkpoint_interval	Frequency of saving a checkpoint. Any non-positive value disables the checkpoint function. TYPE: int DEFAULT: `0`
seed	Random seed. TYPE: int DEFAULT: None

RETURNS	DESCRIPTION
Result	Pymoo result object, see https://pymoo.org/interface/result.html.

run_nsga3

run_nsga3(
    population_size,
    termination,
    /,
    reference_directions=None,
    *,
    p_crossover=1.0,
    p_mutation=0.2,
    init_population_size=0,
    saves_history=True,
    checkpoint_interval=0,
    seed=None,
)

Run optimisation via the NSGA3 algorithm.

PARAMETER	DESCRIPTION
population_size	Population size. TYPE: int
termination	Termination criterion, see https://pymoo.org/interface/termination.html. TYPE: Termination
reference_directions	Reference directions, see https://pymoo.org/misc/reference_directions.html. The Riesz s-Energy method is used to generate reference directions as per the input count and the population size. TYPE: ndarray DEFAULT: None
p_crossover	Crossover probability. TYPE: float DEFAULT: `1.0`
p_mutation	Mutation probability. TYPE: float DEFAULT: `0.2`
init_population_size	Initial population size. This allows setting a different (usually larger) population size for the first generation. Any non-positive value falls back to the value of population_size. TYPE: int DEFAULT: `0`
saves_history	Whether to save the iteration history, see https://pymoo.org/interface/minimize.html. TYPE: bool DEFAULT: `True`
checkpoint_interval	Frequency of saving a checkpoint. Any non-positive value disables the checkpoint function. TYPE: int DEFAULT: `0`
seed	Random seed. TYPE: int DEFAULT: None

RETURNS	DESCRIPTION
Result	Pymoo result object, see https://pymoo.org/interface/result.html.

resume staticmethod

resume(checkpoint_file, /, *, termination=None, checkpoint_interval=0)

Resume optimisation using a checkpoint.

PARAMETER	DESCRIPTION
checkpoint_file	Checkpoint file path. TYPE: str or path-like object
termination	Termination criterion. The one in the checkpoint will be reused if not specified. TYPE: Termination DEFAULT: None
checkpoint_interval	Frequency of saving a checkpoint. Any non-positive value disables the checkpoint function. TYPE: int DEFAULT: `0`

RETURNS	DESCRIPTION
Result	Pymoo result object, see https://pymoo.org/interface/result.html.