selector.generators.surrogates.surrogates

This module contains surrogate management functions.

Classes

SurrogateManager(scenario[, seed, logger])

Managing surrogates and related functions.

class selector.generators.surrogates.surrogates.SurrogateManager(scenario, seed=False, logger=None)

Bases: object

Managing surrogates and related functions.

Parameters:

• scenario (selector.scenario.Scenario) – AC scenario.

• seed (int) – Random seed.

• logger (logging.Logger) – Logger from main loop. Default is None, so no Debug infos.

ei(surrogate, suggestions, next_instance_set)

Compute expected improvement.

Parameters:

• surrogate (selector.pool.Surrogates) – Key to the surrogate model.

• suggestions (list of selector.pool.Configuration) – Suggested configurations.

• next_instance_set (list of str) – List of next instances to be run.

Returns:

ei: Expected improvements for suggestions.

Return type:

ndarray or list

pi(surrogate, suggestions, cot, results, next_instance_set)

Compute probability of improvement.

Parameters:

• surrogate (selector.pool.Surrogates) – Key to the surrogate model.

• suggestions (list of selector.pool.Configuration) – Suggested configurations.

• cot (float) – Timelimit set in AC scenario.

• results (dict) – Performances of the configuration on the instance set of the tournament.

• next_instance_set (list) – List of next instances to run the tournament on.

Returns:

pi: Probabilities of improvement.

Return type:

ndarray

predict(surrogate, configs, cot, next_instance_set)

Get prediction for mean and variance concerning the points quality.

Parameters:

• surrogate (selector.pool.Surrogates) – Key to the surrogate model.

• configs (list of selector.pool.Configuration) – Suggested configurations.

• cot (float) – Timelimit set in AC scenario.

• next_instance_set (list) – List of next instances to run the tournament on.

Returns:

[{“Configuartion ID”: {“qual”: mean predicted performance, “var”: variance, “gen”: selector.pool.Generator}}]

Return type:

list of dict of dict

suggest(suggestor, scenario, n_samples, data, results, next_instance_set)

Suggest points based on surrogate.

Parameters:

• suggestor (selector.pool.Surrogates) – Key to the surrogate model.

• scenario (selector.scenario.Scenario) – AC scenario.

• seed (int) – Random seed.

• logger (logging.Logger) – Logger from main loop. Default is None, so no Debug infos.

Returns:

Suggested configurations.

Return type:

list of selector.pool.Configuration

update_surr(surrogate, history, configs, results, terminations, ac_runtime=None)

Update surrogate model with runhistory.

Parameters:

• surrogate (selector.pool.Surrogates) – Key to the surrogate model.

• history (list of selector.pool.Tournament) – Tournament history.

• configs (list of selector.pool.Configuration) – Configurations that participated in the tournament.

• results (dict) – Results of the tournament.

• terminations (dict) – Information about terminations of runs that occurred.

• ac_runtime (int) – Total AC runtime in seconds so far.