1 RriResult

rri_metrics.RriResult(
    rri,
    pm_dist_before,
    pm_dist_after,
    pm_acc_before,
    pm_comp_before,
    pm_acc_after,
    pm_comp_after,
    fscore_tau=None,
)

Batch of per-candidate RRI outcomes and distance diagnostics.

Shapes follow the candidate batch dimension C produced by the caller. Scalars such as the reference-only distances are broadcast to (C,) so downstream code can remain shape-agnostic.

1.1 Attributes

Name	Description
rri	Tensor[“C”] Relative reconstruction improvement (d_before - d_after) / d_before.
pm_dist_before	Tensor[“C”] Bidirectional Chamfer-style distance between P_t and the GT mesh.
pm_dist_after	Tensor[“C”] Bidirectional distance between P_t ∪ P_q and the GT mesh.
pm_acc_before	Tensor[“C”] Point→mesh (accuracy) distance for P_t (broadcast).
pm_comp_before	Tensor[“C”] Mesh→point (completeness) distance for P_t (broadcast).
pm_acc_after	Tensor[“C”] Point→mesh distance for P_t ∪ P_q.
pm_comp_after	Tensor[“C”] Mesh→point distance for P_t ∪ P_q.
fscore_tau	Optional F-score values at configured distance thresholds.

1.2 Methods

Name	Description
to_serializable	Serialize this result into a cache-friendly CPU payload.
from_serializable	Reconstruct one result from a serialized payload.
to	Move all tensors in this result to the specified device.

1.2.1 to_serializable

rri_metrics.RriResult.to_serializable()

Serialize this result into a cache-friendly CPU payload.

1.2.2 from_serializable

rri_metrics.RriResult.from_serializable(payload, *, device)

Reconstruct one result from a serialized payload.

1.2.2.1 Parameters

Name	Type	Description	Default
payload	dict[str, Any]	Serialized payload produced by to_serializable.	required
device	torch.device	Destination device for tensors.	required

1.2.2.2 Returns

Name	Type	Description
	'RriResult'	Reconstructed RRI result.

1.2.3 to

rri_metrics.RriResult.to(device)

Move all tensors in this result to the specified device.