netket.experimental.nn.rnn.ensure_prev_neighbors#
- netket.experimental.nn.rnn.ensure_prev_neighbors(*, reorder_idx=None, inv_reorder_idx=None, prev_neighbors=None, graph=None, check=False)[source]#
Deduce the missing arguments between reorder_idx, inv_reorder_idx, and inv_reorder_idx from the specified arguments.
See
netket.experimental.models.RNN
for details about the reordering indices.If no information on neighbors or graph is provided, assumes a 1D ordering.
- Parameters:
reorder_idx (
Optional
[HashableArray
]) – indices to transform the inputs from unordered to ordered. Seenetket.models.AbstractARNN.reorder()
for detailsinv_reorder_idx (
Optional
[HashableArray
]) – indices to transform the inputs from ordered to unordered. Seenetket.models.AbstractARNN.reorder()
for details.prev_neighbors (
Optional
[HashableArray
]) – previous neighbors of each site.graph (
AbstractGraph
) – graph of the physical system, to deduce neighbors.check (
bool
) – check the validity of the provided values.
- Return type: