tomodrgn.fft.iht3_center_torch#

iht3_center_torch(possibly_batched_vols: Tensor) Tensor[source]#

Batch-vectorized 3-D inverse DHT of even-sized volume(s) as torch tensor Note: torch maintains default input dtype torch.float32 to output dtype torch.float32

Parameters:

possibly_batched_vols – torch.tensor of volumes DHT, shape ([N], D, D, D)

Returns:

torch.tensor of volumes, shape ([N], D, D, D)