/**
 * The statistics the default evals stand on (DES-DEFAULT-EVALS-001 §6, the same rules as DES-RUNS-REVIEW-001 §4).
 * Pure and deterministic: the same cases give the same numbers, byte for byte.
 *
 *  - Rates: the Wilson score interval (95 %), honest at small n and always inside [0, 1].
 *  - Below {@link MIN_CASES} a figure is "too few cases": the value is still shown, no trend is called.
 *  - Calibration (ECE over 10 bins, Brier) needs {@link MIN_CALIBRATION} confident answers.
 *
 * One statistics module: the Wilson interval and the comparison rule are the run review's (`insights/stats.ts`);
 * this module only shapes them into the default evals' wire (`Rate`) and adds calibration.
 */
import type { Calibration, Rate, TrendWord } from './types.js';
export declare const MIN_CASES = 5;
export declare const MIN_CALIBRATION = 20;
export declare function round4(v: number | null): number | null;
/** Wilson score interval for `k` of `n` (95 %). Value and bounds are `null` when n is 0. */
export declare function rate(k: number, n: number, min?: number): Rate;
/** Calibration of a confidence against being right: ECE over 10 equal-width bins, and the Brier score. */
export declare function calibration(points: ReadonlyArray<{
    confidence: number;
    correct: boolean;
}>): Calibration;
/**
 * This run's rate against the last run's point value, by the run review's rule (`compare`): notable only when both
 * reach {@link MIN_CASES} and this run's interval EXCLUDES the previous value. Higher agreement is better.
 */
export declare function trendOf(now: Rate, previous: {
    value: number | null;
    n: number;
} | null): TrendWord;
//# sourceMappingURL=stats.d.ts.map