mo.pareto
mo.pareto
Functions
| Name | Description |
|---|---|
| is_pareto_efficient | Find the Pareto-efficient points from a set of points. |
| mo_pareto_optx_plot | Visualizes the Pareto-optimal points in the input space for each pair of inputs |
| mo_xy_contour | Generates contour plots of every combination of two input variables x_i and x_j |
| mo_xy_surface | Generates surface plots of every combination of two input variables x_i and x_j |
is_pareto_efficient
mo.pareto.is_pareto_efficient(costs, minimize=True)Find the Pareto-efficient points from a set of points.
A point is Pareto-efficient if no other point exists that is better in all objectives. This function assumes that lower values are preferred for each objective when minimize=True, and higher values are preferred when minimize=False.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
| costs | np.ndarray | An (N,M) array-like object of points, where N is the number of points and M is the number of objectives. | required |
| minimize | bool | If True, the function finds Pareto-efficient points assuming lower values are better. If False, it assumes higher values are better. Defaults to True. | True |
Returns
| Name | Type | Description |
|---|---|---|
| np.ndarray | np.ndarray: A boolean mask of length N, where True indicates that the corresponding point is Pareto-efficient. |
Examples:
>>> import numpy as np
>>> from spotoptim.mo.pareto import is_pareto_efficient
>>> points = np.array([[1, 2], [2, 1], [1.5, 1.5], [3, 3]])
>>> pareto_mask = is_pareto_efficient(points, minimize=True)
>>> print(pareto_mask)
[ True True True False]mo_pareto_optx_plot
mo.pareto.mo_pareto_optx_plot(
X,
Y,
minimize=True,
feature_names=None,
target_names=None,
**kwargs,
)Visualizes the Pareto-optimal points in the input space for each pair of inputs x_i and x_j (with i < j) and each objective f_k.
Plots are placed on a grid where rows correspond to input pairs and columns correspond to objectives.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
| X | np.ndarray | An (N,D) array of input points, where N is the number of points and D is the number of variables (dimensions). | required |
| Y | np.ndarray | An (N,M) array of objective values, where N is the number of points and M is the number of objectives. | required |
| minimize | bool | If True, assumes minimization of objectives. Defaults to True. | True |
| feature_names | list | List of names for the input variables. Defaults to None. | None |
| target_names | list | List of names for the objectives. Defaults to None. | None |
| **kwargs | Any | Additional arguments passed to plt.subplots (e.g., figsize). | {} |
Returns
| Name | Type | Description |
|---|---|---|
| None | None |
Examples
>>> from spotoptim.mo.pareto import mo_pareto_optx_plot
>>> X = np.array([[1, 2], [3, 4], [5, 6]])
>>> Y = np.array([[1, 2], [3, 4], [5, 6]])
>>> mo_pareto_optx_plot(X, Y)mo_xy_contour
mo.pareto.mo_xy_contour(
models,
bounds,
target_names=None,
feature_names=None,
resolution=50,
feature_pairs=None,
**kwargs,
)Generates contour plots of every combination of two input variables x_i and x_j (where i < j) and for each of the multiple objectives f_k.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
| models | list | List of trained models (one per objective). | required |
| bounds | list | List of tuples (min, max) for each input variable. | required |
| target_names | list | List of names for the objectives. Defaults to None. | None |
| feature_names | list | List of names for the input variables. Defaults to None. | None |
| resolution | int | Grid resolution for the contour plot. Defaults to 50. | 50 |
| feature_pairs | list | List of tuples (i, j) specifying which feature pairs to plot. If None, all combinations are plotted. Defaults to None. | None |
| **kwargs | Any | Additional keyword arguments passed to plt.subplots (e.g., figsize). | {} |
Returns
| Name | Type | Description |
|---|---|---|
| None | None |
Examples
>>> from sklearn.ensemble import RandomForestRegressor
>>> from spotoptim.mo.pareto import mo_xy_contour
>>> import numpy as np
>>> # Train dummy models
>>> X = np.random.rand(10, 2)
>>> y1 = X[:, 0] + X[:, 1]
>>> y2 = X[:, 0] * X[:, 1]
>>> m1 = RandomForestRegressor().fit(X, y1)
>>> m2 = RandomForestRegressor().fit(X, y2)
>>> # Plot
>>> mo_xy_contour([m1, m2], bounds=[(0, 1), (0, 1)], target_names=["Sum", "Prod"])mo_xy_surface
mo.pareto.mo_xy_surface(
models,
bounds,
target_names=None,
feature_names=None,
resolution=50,
feature_pairs=None,
**kwargs,
)Generates surface plots of every combination of two input variables x_i and x_j (where i < j) and for each of the multiple objectives f_k.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
| models | list | List of trained models (one per objective). | required |
| bounds | list | List of tuples (min, max) for each input variable. | required |
| target_names | list | List of names for the objectives. Defaults to None. | None |
| feature_names | list | List of names for the input variables. Defaults to None. | None |
| resolution | int | Grid resolution for the surface plot. Defaults to 50. | 50 |
| feature_pairs | list | List of tuples (i, j) specifying which feature pairs to plot. If None, all combinations are plotted. Defaults to None. | None |
| **kwargs | Any | Additional keyword arguments passed to plt.subplots (e.g., figsize). | {} |
Returns
| Name | Type | Description |
|---|---|---|
| None | None |
Examples
>>> from sklearn.ensemble import RandomForestRegressor
>>> from spotoptim.mo.pareto import mo_xy_surface
>>> import numpy as np
>>> # Train dummy models
>>> X = np.random.rand(10, 2)
>>> y1 = X[:, 0] + X[:, 1]
>>> y2 = X[:, 0] * X[:, 1]
>>> m1 = RandomForestRegressor().fit(X, y1)
>>> m2 = RandomForestRegressor().fit(X, y2)
>>> # Plot
>>> mo_xy_surface([m1, m2], bounds=[(0, 1), (0, 1)], target_names=["Sum", "Prod"])