`pyepo.data.dataset`

optDataset class based on PyTorch Dataset

Module Contents

Classes

`optDataset`	This class is Torch Dataset for optimization problems.
`optDatasetKNN`	This class is Torch Dataset for optimization problems, when using the robust kNN-loss.

class pyepo.data.dataset.optDataset(model, feats, costs)

Bases: torch.utils.data.Dataset

This class is Torch Dataset for optimization problems.

model

Optimization models

Type:: optModel

feats

Data features

Type:: np.ndarray

costs

Cost vectors

Type:: np.ndarray

sols

Optimal solutions

Type:: np.ndarray

objs

Optimal objective values

Type:: np.ndarray

_getSols(): A method to get optimal solutions for all cost vectors

_solve(cost)

A method to solve optimization problem to get an optimal solution with given cost

Parameters:: cost (np.ndarray) – cost of objective function
Returns:: optimal solution (np.ndarray) and objective value (float)
Return type:: tuple

__len__()

A method to get data size

Returns:: the number of optimization problems
Return type:: int

__getitem__(index)

A method to retrieve data

Parameters:: index (int) – data index
Returns:: data features (torch.tensor), costs (torch.tensor), optimal solutions (torch.tensor) and objective values (torch.tensor)
Return type:: tuple

class pyepo.data.dataset.optDatasetKNN(model, feats, costs, k=10, weight=0.5)

Bases: optDataset

This class is Torch Dataset for optimization problems, when using the robust kNN-loss.

Reference: <https://arxiv.org/abs/2310.04328>