# License: BSD-3-Clause
try:
import torch
DLBaseDataset = torch.utils.data.Dataset
deepmodule_installed = True
except ImportError:
deepmodule_installed = False
deepmodule_error = "Module 'deep' needs to be installed. See https://imml.readthedocs.io/stable/main/installation.html#optional-dependencies"
DLBaseDataset = object
[docs]
class MRGCNDataset(DLBaseDataset):
r"""
This class provides a `torch.utils.data.Dataset` implementation for handling multi-modal datasets with `MRGCN`.
Parameters
----------
Xs : list of array-likes objects
- Xs length: n_mods
- Xs[i] shape: (n_samples, n_features_i)
A list of different modalities.
transform : list of callable, defult=None
A list of functions or transformations to apply to each sample in the dataset.
Example
--------
>>> import numpy as np
>>> import torch
>>> from imml.load import MRGCNDataset
>>> Xs = [torch.from_numpy(np.random.default_rng(42).random((20, 10))) for i in range(3)]
>>> train_data = MRGCNDataset(Xs=Xs)
"""
def __init__(self, Xs: list, transform = None):
if not deepmodule_installed:
raise ImportError(deepmodule_error)
if not isinstance(Xs, list):
raise ValueError(f"Invalid Xs. It must be a list of array-likes objects. A {type(Xs)} was passed.")
if len(Xs) < 2:
raise ValueError(f"Invalid Xs. It must have at least two modalities. Got {len(Xs)} modalities.")
if any(len(X) == 0 for X in Xs):
raise ValueError("Invalid Xs. All elements must have at least one sample.")
if len(set(len(X) for X in Xs)) > 1:
raise ValueError("Invalid Xs. All elements must have the same number of samples.")
self.Xs = Xs
self.transform = transform
def __len__(self):
return len(self.Xs[0])
def __getitem__(self, idx):
if self.transform is not None:
Xs = [self.transform[idx](X[idx]) for X in self.Xs]
else:
Xs = [X[idx] for X in self.Xs]
Xs = tuple(Xs)
return Xs