agrag.ingestion.reports.consolidation_report.ConsolidationReport

Bases: BaseModel

Report from Graph.consolidate().

Attributes:

• would_match (list[MatchDecision]) – Confirmed non-exact matches found, whether applied or not.
• applied (bool) – Whether the matches were applied.
• failures (list[StageFailure]) – Failures reading the candidates of an entity, which that entity then skips, and failures writing a match graph or rebuilding resolved entities. Only the candidate read failures appear when apply is False.
• ambiguous_count (int) – LLM answers that were uncertain, missing, or malformed. These pairs never merge.
• failed_llm_requests (int) – LLM verification requests that raised, summed over all labels. Their pairs are undecided and not merged.
• cap_truncated_pairs (int) – Ambiguous pairs over the per-label cap (max_llm_pairs), summed over all labels. They never reached the LLM.
• llm_unavailable_pairs (int) – Ambiguous pairs that never reached the LLM because no LLM settings were found, summed over all labels.

ambiguous_count

ambiguous_count: int = 0

applied

applied: bool = False

cap_truncated_pairs

cap_truncated_pairs: int = 0

failed_llm_requests

failed_llm_requests: int = 0

failures

failures: list[StageFailure] = Field(default_factory=list)

llm_unavailable_pairs

llm_unavailable_pairs: int = 0

would_match

would_match: list[MatchDecision] = Field(default_factory=list)