ExplorationDiagnostics

class optimas.diagnostics.ExplorationDiagnostics(source)

Utilities for analyzing the output of an exploration.

Parameters:

sourcestr

Path to the exploration directory or to an individual .npy history file, or an Exploration instance.

Methods

build_gp_model(parameter[, minimize, ...])	Build a GP model of the specified parameter.
delete_evaluation_dir(trial_index)	Delete the directory with the output of the given evaluation.
get_best_evaluation([objective])	Get the evaluation with the best objective value.
get_best_evaluation_dir_path([objective])	Get the path to the directory of the best evaluation.
get_best_evaluations([objective, top])	Get a dataframe with the best evaluations.
get_evaluation_dir_path(trial_index)	Get the path to the directory of the given evaluation.
get_objective_trace([objective, ...])	Get the cumulative maximum or minimum of the objective.
get_pareto_front_evaluations([objectives])	Get data of evaluations in the Pareto front.
plot_history([parnames, xname, select, ...])	Print selected parameters versus evaluation index.
plot_objective([objective, ...])	Plot the values that where reached during the optimization.
plot_pareto_front([objectives, ...])	Plot Pareto front of two optimization objectives.
plot_worker_timeline([fidelity_parameter, ...])	Plot the timeline of worker utilization.
print_best_evaluations([objective, top])	Print top evaluations according to the given objective.
print_evaluation(trial_index)	Print the parameters of the given evaluation.

Attributes

analyzed_parameters	Get the analyzed parameters of the exploration.
exploration_dir_path	Get the exploration dir path.
history	Return a pandas DataFrame with the exploration history.
objectives	Get the objectives of the exploration.
varying_parameters	Get the varying parameters of the exploration.