baseline(#40): add h0-not-a-boundary region-subset trap (global-30x30 q5) - #87
Merged
Conversation
… q5) Grow-on-fix for the geo-agent #322/#325 fix (region subset via WHERE h0 IN (…) renders whole res-0 base cells instead of the state). New answer-mode question 'How much irrecoverable carbon (2024) is stored in California?' guards traps h0-not-a-boundary + region-subset-finest-shared-res. Operator-verified via duckdb-geo (2026-07-15, v0.8.7): - CA correct 3.999e8 Mg C (~400 Mt); h0-only trap 2.543e9 (6.4x too high) - FL correct 4.658e8 Mg C (~466 Mt); h0-only trap 1.934e9 (single base cell) Gold + authoritative SQL in gold/global-30x30.md; golden.json regenerated.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Grow-on-fix addition to the #40 standing baseline, guarding the geo-agent #322/#325 fix.
Trap
Subsetting a global value dataset (carbon) to a state via
WHERE h0 IN (…)renders whole res-0 base cells (~4.35M km² each), not the state —h0is the hive partition key, not a boundary. Correct: clip via the census state hex mask at the finest shared resolution (h8).New question (answer-mode, global-30x30 q5)
How much irrecoverable carbon (2024 layer) is stored in California?
h0-not-a-boundary,region-subset-finest-shared-res.Operator-verified via duckdb-geo 2026-07-15 (v0.8.7); authoritative SQL in
gold/global-30x30.md.golden.jsonregenerated viabuild_golden.py(30 questions).The tool-mechanics halves of this work (#321 sql-alias, #326 schema-title-leak) are covered by deterministic pytest in mcp-data-server; this captures the behavioral region-subset trap that only shows up through a model.