Source code for botorch.utils.multi_objective.box_decompositions.box_decomposition_list

#!/usr/bin/env python3
# Copyright (c) Facebook, Inc. and its affiliates.
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.

r"""Box decomposition container."""

from __future__ import annotations

from typing import List, Union

import torch
from botorch.exceptions.errors import BotorchTensorDimensionError
from botorch.utils.multi_objective.box_decompositions.box_decomposition import (
from torch import Tensor
from torch.nn import ModuleList, Module

[docs]class BoxDecompositionList(Module): r"""A list of box decompositions.""" def __init__(self, *box_decompositions: BoxDecomposition) -> None: r"""Initialize the box decomposition list. Args: *box_decompositions: An variable number of box decompositions Example: >>> bd1 = FastNondominatedPartitioning(ref_point, Y=Y1) >>> bd2 = FastNondominatedPartitioning(ref_point, Y=Y2) >>> bd = BoxDecompositionList(bd1, bd2) """ super().__init__() self.box_decompositions = ModuleList(box_decompositions) @property def pareto_Y(self) -> List[Tensor]: r"""This returns the non-dominated set. Note: Internally, we store the negative pareto set (minimization). Returns: A list where the ith element is the `n_pareto_i x m`-dim tensor of pareto optimal outcomes for each box_decomposition `i`. """ return [p.pareto_Y for p in self.box_decompositions] @property def ref_point(self) -> Tensor: r"""Get the reference point. Note: Internally, we store the negative reference point (minimization). Returns: A `n_box_decompositions x m`-dim tensor of outcomes. """ return torch.stack([p.ref_point for p in self.box_decompositions], dim=0)
[docs] def get_hypercell_bounds(self) -> Tensor: r"""Get the bounds of each hypercell in the decomposition. Returns: A `2 x n_box_decompositions x num_cells x num_outcomes`-dim tensor containing the lower and upper vertices bounding each hypercell. """ bounds_list = [] max_num_cells = 0 for p in self.box_decompositions: bounds = p.get_hypercell_bounds() max_num_cells = max(max_num_cells, bounds.shape[-2]) bounds_list.append(bounds) # pad the decomposition with empty cells so that all # decompositions have the same number of cells for i, bounds in enumerate(bounds_list): num_missing = max_num_cells - bounds.shape[-2] if num_missing > 0: padding = torch.zeros( 2, num_missing, bounds.shape[-1], dtype=bounds.dtype, device=bounds.device, ) bounds_list[i] = [ bounds, padding, ], dim=-2, ) return torch.stack(bounds_list, dim=-3)
[docs] def update(self, Y: Union[List[Tensor], Tensor]) -> None: r"""Update the partitioning. Args: Y: A `n_box_decompositions x n x num_outcomes`-dim tensor or a list where the ith element contains the new points for box_decomposition `i`. """ if ( torch.is_tensor(Y) and Y.ndim != 3 and Y.shape[0] != len(self.box_decompositions) ) or (isinstance(Y, List) and len(Y) != len(self.box_decompositions)): raise BotorchTensorDimensionError( "BoxDecompositionList.update requires either a batched tensor Y, " "with one batch per box decomposition or a list of tensors with " "one element per box decomposition." ) for i, p in enumerate(self.box_decompositions): p.update(Y[i])
[docs] def compute_hypervolume(self) -> Tensor: r"""Compute hypervolume that is dominated by the Pareto Froniter. Returns: A `(batch_shape)`-dim tensor containing the hypervolume dominated by each Pareto frontier. """ return torch.stack( [p.compute_hypervolume() for p in self.box_decompositions], dim=0 )