Fixed-cohort patch-event relative-change aggregates
Question
Compute fixed-cohort, event-indexed median relative-price-change aggregates for admitted patch-notice events.
Verdict
287 cohort-event rows cleared the minimum-sample and promotion gate out of 182 admitted events x their matched cohorts (9 cohorts, partition tradeable-v1-b3bfe7781e). Each row is one event's cohort-wide MEDIAN relative price change; no per-item value, extreme or cumulative statistic is ever computed or published. A supporting dataset for the optional patch-events route, not a hypothesis-test verdict.
In plain language (1 term)
This site's own explanation of the words above. The verdict itself is published exactly as the research index records it and is not rewritten here.
- cohort
- A group of items fixed in advance, so the grouping cannot be adjusted after the result is seen.
- Population
- 9 cohorts, partition tradeable-v1-b3bfe7781e; 182 admitted events.
- Resolution
- Per-patch-event cohort medians (event-indexed, not fixed-interval).
- Cutoff
- Development span only.
- Baseline
- n/a -- a supporting dataset for the optional patch-events route, not a hypothesis-test verdict.
Limitations
287 of the possible cohort-event rows cleared the minimum-sample and promotion gate; no per-item, extreme or cumulative statistic is ever computed or published.
Notes
Landed in 0608207, 13 minutes before this index's own commit on the same branch -- a completed, tested, already-merged terminal result at the time this index was written, not concurrent in-flight work. An earlier draft of this index excluded it under that mistaken premise; corrected on review.
Source
- Citation
- po-research:patch-event-aggregates:7fd43754181e614f
- Write-up
- docs/research/patch-event-aggregates.md
- Data
- docs/research/patch-event-aggregates.json
- Artifact
research-index.jsonv1 — sha256:cb3d976779a5cbcb4188039a8d26d5e558a2d844030a638c2f16fe5a8eb0a2fe- Build data cutoff
- 2026-08-12
- Issues
- #135, #132, #131