mrpro.operators.models.EPG.initial_state

mrpro.operators.models.EPG.initial_state(shape: Sequence[int], n_states: int = 20, device: device | str = 'cpu', dtype: dtype = torch.complex64) → Tensor[source]

Generate initial EPG states.

Parameters:

• shape (Sequence[int]) – Shape of the state tensor, excluding the last two dimensions. Should match the parameters shape.

• n_states (int, default: 20) – Number of EPG configuration states.

• device (device | str, default: 'cpu') – Device to create the state tensor on.

• dtype (dtype, default: torch.complex64) – Data type of the state tensor.

Returns:

Initial EPG state tensor. Shape *shape, 3 (f_plus, f_minus, z), n