pareto module¶
A module to define a Pareto front.
- class gemseo.algos.pareto.ParetoFront(distance_from_utopia, f_anchors, f_anti_utopia, f_optima, f_utopia, f_utopia_neighbors, x_anchors, x_optima, x_utopia_neighbors)[source]
Bases:
object
A Pareto front.
The design and objective vectors are noted
x
andf
respectively.- Parameters:
distance_from_utopia (float) –
f_anchors (RealArray) –
f_anti_utopia (RealArray) –
f_optima (RealArray) –
f_utopia (RealArray) –
f_utopia_neighbors (RealArray) –
x_anchors (RealArray) –
x_optima (RealArray) –
x_utopia_neighbors (RealArray) –
- classmethod from_optimization_problem(problem)[source]
Create a
ParetoFront
from anOptimizationProblem
.- Parameters:
problem – The optimization problem.
- Returns:
The Pareto front.
- Return type:
- f_anchors: RealArray
The values of the objectives of all anchor points.
At those points, each objective is minimized one at a time.
Its shape is
(n_anchors, f_dimension)
.
- f_anti_utopia: RealArray
The anti-utopia point, i.e. the maximum objective vector.
Its shape is
(f_dimension,)
.
- f_optima: RealArray
The objective values of the Pareto optima.
Its shape is
(n_optima, f_dimension)
.
- f_utopia: RealArray
The utopia point, i.e. the minimum objective vector.
In most Pareto fronts, there is no design value for which the objective is equal to the utopia.
Its shape is
(f_dimension,)
.
- f_utopia_neighbors: RealArray
The objectives value of the closest point(s) to the
f_utopia
.The distance separating them from
f_utopia
isdistance_from_utopia
.Its shape is
(n_neighbors, f_dimension)
.
- x_anchors: RealArray
The values of the design variables values of all anchor points.
At those points, each objective is minimized one at a time.
Its shape is
(n_anchors, x_dimension)
.
- x_optima: RealArray
The values of the design variables of the Pareto optima.
Its shape is
(n_optima, x_dimension)
.
- x_utopia_neighbors: RealArray
The design variables value of the closest point(s) to the
f_utopia
.The distance separating them from
f_utopia
isdistance_from_utopia
.Its shape is
(n_neighbors, x_dimension)
.