1 CounterfactualCandidateEvaluation

pose_generation.CounterfactualCandidateEvaluation(
    scores,
    score_label='score',
    metrics=None,
    metric_vectors=None,
    candidate_point_clouds_world=None,
    candidate_point_cloud_lengths=None,
    target_eval_current_points_world=None,
    target_eval_candidate_points_world=None,
    target_eval_candidate_point_lengths=None,
    target_eval_crop_policy=None,
    target_eval_voxel_size_m=None,
    target_eval_max_points=None,
)

Structured per-valid-candidate rollout scores and optional diagnostics.

1.1 Attributes

Name	Description
metric_vectors	Return present metric vectors by stable metric name.

1.2 Methods

Name	Description
validate	Normalize score vectors and ensure they align with valid candidates.

1.2.1 validate

pose_generation.CounterfactualCandidateEvaluation.validate(
    num_valid,
    device,
    dtype,
)

Normalize score vectors and ensure they align with valid candidates.