torchcvnn.transforms.RealImaginary

class torchcvnn.transforms.RealImaginary(dtype: str | None = None)

Transform a complex-valued tensor into its real and imaginary components.

This transform separates a complex-valued tensor into its real and imaginary parts, stacking them along a new channel dimension. The output tensor has twice the number of channels as the input.

Returns:

Real-valued tensor containing real and imaginary parts,

with shape (2*C, H, W) where C is the original number of channels.

Return type:

np.ndarray | torch.Tensor

__init__(dtype: str | None = None) → None

Methods

__init__([dtype])