calibration: count same-day closes honestly (#34)

The cold-start surface showed "N/20 closed issues estimated", implying you're
just (20−N) closes away. But calibrationSamples silently drops closed+estimated
issues that closed in 0 working days (same-day closes) — real closes that
structurally can't calibrate. On this repo that's 10 of 24 closes hidden: the
note read 14/20 as if 6 away, when a third of the history will never count.

- core: `calibrationCoverage(issues, timelines, asOf)` → { candidates, usable,
  excludedSameDay }, counting the silently-excluded same-day closes. Pure, tested.
- surface it: CalibrationData gains `excludedSameDay`; backlogCalibration returns
  the coverage; the Runway note and the Calibration screen now say "… · N same-day
  closes can't calibrate" so the thin sample is explained, not just reported.

Verified on christian/commitea: closed=24, usable=14, excludedSameDay=10.
131 core green (incl. new coverage test); core + desktop typecheck; 14 fixture e2e.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Croissant Le Doux
2026-07-09 15:18:16 -04:00
parent 62a521e7eb
commit 89c873b368
8 changed files with 88 additions and 5 deletions

View File

@@ -5,6 +5,7 @@ import type { LifecycleEvent } from '../lifecycle/lifecycle-v0.js'
import type { GiteaIssue } from '../gitea/types.js'
import {
CALIBRATION_BUCKET_FLOOR,
calibrationCoverage,
calibrationSamples,
type CalibrationSample,
COLD_START_THRESHOLD,
@@ -118,4 +119,26 @@ describe('calibrationSamples', () => {
const noEst = issue({ number: 9, labels: [] })
expect(calibrationSamples([open, noEst], { ...events(8), ...events(9) }, asOf)).toEqual([])
})
it('coverage counts same-day closes as excluded candidates, not as "more closes needed"', () => {
// usable: commit Wed 01-07 → close Mon 01-12 = 3 working days
const usable = issue({ number: 7, labels: ['est/2d'] })
// same-day close: commit and close on the same day = 0 working days → excluded
const sameDay = issue({ number: 10, labels: ['est/2d'], createdAt: '2026-01-12T08:00:00Z' })
const sameDayEvents = {
10: [
{ type: 'commit', at: '2026-01-12T09:00:00Z' } as LifecycleEvent,
{ type: 'close', at: '2026-01-12T17:00:00Z' } as LifecycleEvent,
],
}
const open = issue({ number: 8, state: 'open', labels: ['est/2d'], closedAt: null })
const noEst = issue({ number: 9, labels: [] })
const cov = calibrationCoverage([usable, sameDay, open, noEst], { ...events(7), ...sameDayEvents }, asOf)
expect(cov.candidates).toBe(2) // closed + estimated only (usable + sameDay)
expect(cov.usable).toBe(1)
expect(cov.excludedSameDay).toBe(1)
// the honest denominator: usable matches the model's n
expect(cov.usable).toBe(calibrationSamples([usable, sameDay, open, noEst], { ...events(7), ...sameDayEvents }, asOf).length)
})
})

View File

@@ -125,3 +125,40 @@ export function calibrationSamples(
}
return out
}
/** How the closed+estimated backlog splits into usable samples vs. what can't calibrate. */
export interface CalibrationCoverage {
/** Closed issues carrying an estimate — the calibration candidates. */
candidates: number
/** Candidates that yielded a usable actual (> 0 working days) → become samples. */
usable: number
/**
* Candidates excluded because the issue closed with 0 working days (same-day
* close) or no resolvable actual — real closes that structurally can't
* calibrate. Counting them keeps `usable/threshold` honest: it's not "N more
* closes away" if some of your closes will never count.
*/
excludedSameDay: number
}
/**
* Coverage of the calibration candidates — how many closed+estimated issues are
* usable vs. silently unusable (same-day / 0-day closes). {@link calibrationSamples}
* drops the latter; this counts them so the UI can say *why* the sample is thin.
*/
export function calibrationCoverage(
issues: GiteaIssue[],
timelines: Record<number, LifecycleEvent[]>,
asOf: Date,
): CalibrationCoverage {
let candidates = 0
let usable = 0
for (const issue of issues) {
if (issue.state !== 'closed') continue
if (issue.facts.estimateDays == null) continue
candidates++
const inf = inferLifecycle(issue, timelines[issue.number] ?? [], asOf)
if (inf.actualWorkingDays != null && inf.actualWorkingDays > 0) usable++
}
return { candidates, usable, excludedSameDay: candidates - usable }
}

View File

@@ -73,6 +73,7 @@ export type {
export {
CALIBRATION_BUCKET_FLOOR,
calibrationCoverage,
calibrationSamples,
COLD_START_THRESHOLD,
fitCalibration,
@@ -80,6 +81,7 @@ export {
} from './calibration/calibration-v0.js'
export type {
BucketFit,
CalibrationCoverage,
CalibrationModel,
CalibrationSample,
PersonBias,