scSLAT.metrics.edge_score

scSLAT.metrics.edge_score(edges, matching, score=[1, -1], punish_distance=False, punish_scale=1)

Calculate edge score in two aligned graph (higher means better)

Parameters:

• edges (typing.List[torch.Tensor]) – list of edge of every dataset

• matching (torch.Tensor) – matching result

• score (typing.Optional[typing.List[float]]) – score of [match_edge, mismatch_edge]

• punish_distance (typing.Optional[bool]) – if punish on the distance of mismatch cell

• punish_scale (typing.Optional[float]) – punish scale if punish_distance is True

• Reference –

• ---------- –

• Alignment" (Joel Douglas et al. "Metrics for Evaluating Network) –

Return type:

float