core.optimization module¶
Module containing the optimization surrogate.
-
core.optimization.ref_points¶ Reference points for benchmark problems.
Type: dict
-
class
core.optimization.Optimization(model)¶ Bases:
objectThe class to define the optimization problem.
-
model¶ The model object.
Type: core.Model
-
iterations¶ Number of optimization iterations.
Type: int
-
converged¶ Convergence status.
Type: bool
-
algorithm¶ Optization algorithm object.
-
termination¶ Termination object.
-
n_const¶ NUmber of constraints.
Type: int
-
ref_point¶ Reference point for hypervolume calculation.
Type: np.array
-
direct¶ Whether a direct optimization is performed.
Type: bool
-
range_in¶ Input parameter allowable ranges.
Type: np.array
-
function¶ Function used to evaluate the candidates.
-
problem¶ Problem object.
Type: datamod.problems.Custom
-
surrogate¶ Surrogate object.
Type: core.Surrogate
-
res¶ Results object.
Type: pymoo.model.result.Result
-
optimization_stats¶ Optimization benchmark statistics.
Type: dict
-
optimum_model¶ Candidates evaluated by the original model.
Type: np.array
-
optimum_surrogate¶ Candidates evaluated by the surrogate.
Type: np.array
-
error_measure¶ Maximum of the error metrics.
Type: float
-
error¶ Benchmark percent error.
Type: np.array
-
benchmark()¶ Determine the benchmark optimization accuracy.
-
optimize()¶ Wrapper function to perform optimization.
-
plot_results()¶ Plot the optimized candidates.
-
report()¶ Report the optimal solutions and their verification accuracy.
-
set_problem(surrogate)¶ Wrapper function to set the problem.
Parameters: surrogate (core.Surrogate) – Surrogate object. Notes
Direct optimization does not normalize.
-
verify()¶ Wrapper function to verify the optimized solutions.
-