latent.flows.conversation_trajectory_flow

conversation_trajectory_flow — Trajectory distance analysis (FR-3.7).

Functions

conversation_trajectory_flow

conversation_trajectory_flow(conversations: list, score_key: str = 'score', reference_pattern: list[float] | None = None, confidence_level: float = 0.95, n_resamples: int = 10000, seed: int | None = None) -> dict[str, Any]

Compute trajectory distances between conversations.

Args: conversations: List of Conversation objects with scores in metadata. score_key: Key in turn metadata to extract scores from. reference_pattern: Optional ideal trajectory to compare against. confidence_level: CI confidence level. n_resamples: Bootstrap resamples. seed: Random seed.

Returns: Dict with: trajectories, distances, report, markdown.