src.llm_interpreter.confidence_scorer

Confidence scoring for model diagnostics.

Module Contents

class src.llm_interpreter.confidence_scorer.ConfidenceScorer

Score confidence in model diagnostics.

static score_convergence(r_hat: float, ess: float) → Tuple[str, float]

Score convergence confidence.

Returns:

(confidence_level, score) where level is ‘high’, ‘medium’, or ‘low’.

static score_model_fit(r_squared: float, mape: float) → Tuple[str, float]

Score model fit confidence.

Returns:

(confidence_level, score).

static get_confidence_emoji(level: str) → str

Get emoji representation of confidence level.

static get_action_recommendation(level: str, metric_type: str) → str

Get action recommendation based on confidence level.