Evaluation harness: Belgium jurisdiction with Axiom, EUROMOD-engine, and EUROMOD/SILC sources - #164
Merged
Merged
Conversation
- Generic precomputed-checkpoint adapter (microcosm_be.evaluation_source_checkpoint.v1): joins chronicle_record_ids to snapshot facts, composite rows as semantic aggregates (CPS ASEC 80-84 convention), estimate_basis carried into result provenance; the Yale checkpoint now uses the shared materializer instead of its own. - Tested calibration-exposure rule: direct_calibration_target only for release-target facts with calibration_measure/dataset_weights basis; related_calibration_family for engine-simulated values of calibrated concepts; out_of_sample otherwise. - Snapshot jurisdiction derivation: BE only from package source_id belgium, Eurostat geo=BE, or NIS geography vintages (760-row bundle -> 726 BE + 34 unknown comparators). - run_full_chronicle_evaluation.py: --jurisdictions (default US, unchanged), a per- jurisdiction source-plan registry, --precomputed-checkpoint, --microcosm-release-dir. - docs/cross-dataset-api.md: jurisdiction model replaces the US-only statement. - 403 tests green incl. a trimmed BE end-to-end run with all three sources. Authored by a sol lane; reviewed by fable.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
This was referenced Aug 23, 2026
Merged
Merged
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Belgium becomes a first-class jurisdiction of the cross-dataset evaluation (the
/microcosm/datasetsbundle), with EUROMOD entering as sources scored against Chronicle actuals rather than as a bespoke panel (the panel added in #163 is being removed in a sibling PR).Sources (precomputed checkpoints, schema
microcosm_be.evaluation_source_checkpoint.v1, every cell with anestimate_basis):microcosm_be_v04_axiom— Microcosm-BE v0.4 population × Axiom rules engine (56/66 surface rows)microcosm_be_v04_euromod— Microcosm-BE v0.4 population × EUROMOD BE_2025 engine (55/66)euromod_be2025_jrc_silc— EUROMOD BE_2025 on its own EU-SILC input, JRC country report EUROMOD column (4/66 today; the remaining columns are a Chronicle package follow-up)Harness changes:
unknown)--jurisdictions(default US, behaviour unchanged), per-jurisdiction source registry,--precomputed-checkpoint,--microcosm-release-dir403 tests green. The BE bundle produced by this branch (run
evaluation-f28ca06a…, snapshotchronicle-82b574e3…, all partition hashes verified,jurisdictions: ["BE"]) is published separately and wired viaCROSS_DATASET_ARTIFACT_BASE_URL.🤖 Generated with Claude Code