Deep learning for NeuroImaging in Python.

Note

This page is a reference documentation. It only explains the class signature, and not how to use it. Please refer to the gallery for the big picture.

class nidl.transforms.MultiViewsTransform(transforms: Callable | Sequence[Callable], n_views: int | None = None, **kwargs)[source]¶

Bases: Transform

Multi-views transformation.

It generates several “views” of the same input data, i.e. it applies transformations (usually stochastic) multiple times to the input.

Parameters:

transforms : Callable or Sequence of Callable

Transformation or sequence of transformations to be applied. If a single transform is given, it generates n_views of the same input using the same transformation applied n_views times. If a sequence is given, it applies this sequence of transforms to the input in the same order.

n_views : int or None, default=None

Number of views to generate if transforms is a Transform. If n_views != 1 and transforms is a sequence, a ValueError is raised. If None, it is set to 1 if transforms is a Transform and ignored otherwise.

kwargs : dict

Additional keyword arguments given to Transform.

Returns:

data : list of array or torch.Tensor

List of transformed data.

Notes

The data are not parsed by this transformation. It should be handled elsewhere.

apply_transform(x: Any, **kwargs) → list[ndarray | Tensor][source]¶

Apply the transformation on the data parsed by parse_data(). This should be implemented in all subclasses.

Parameters:

data_parsed : Any

Input data with type and shape already checked.

*args : Any

Additional positional arguments.

**kwargs : dict

Additional keyword arguments.

Returns:

data : Any

The transformed data.

parse_data(data: Any)[source]¶: Data are not parsed here.