policyengine 4.4.2__tar.gz → 4.4.3__tar.gz

{policyengine-4.4.2 → policyengine-4.4.3}/CHANGELOG.md

@@ -1,3 +1,11 @@
+## [4.4.3] - 2026-05-11
+
+### Changed
+
+- Derive US program-statistics entity from variable metadata instead of duplicating it in the program list.
+- Update the bundled US release to policyengine-us 1.690.7 and refresh data provenance metadata.
+
+
 ## [4.4.2] - 2026-05-09
 
 ### Fixed

{policyengine-4.4.2 → policyengine-4.4.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: policyengine
-Version: 4.4.2
+Version: 4.4.3
 Summary: A package to conduct policy analysis using PolicyEngine tax-benefit models.
 Author-email: PolicyEngine <hello@policyengine.org>
 License:                     GNU AFFERO GENERAL PUBLIC LICENSE
@@ -686,11 +686,11 @@ Requires-Dist: plotly>=5.0.0; extra == "plotting"
 Provides-Extra: graph
 Requires-Dist: networkx>=3.0; extra == "graph"
 Provides-Extra: uk
-Requires-Dist: policyengine_core>=3.26.0; extra == "uk"
+Requires-Dist: policyengine_core>=3.26.1; extra == "uk"
 Requires-Dist: policyengine-uk==2.88.14; extra == "uk"
 Provides-Extra: us
-Requires-Dist: policyengine_core>=3.26.0; extra == "us"
-Requires-Dist: policyengine-us==1.687.0; extra == "us"
+Requires-Dist: policyengine_core>=3.26.1; extra == "us"
+Requires-Dist: policyengine-us==1.690.7; extra == "us"
 Provides-Extra: dev
 Requires-Dist: pytest; extra == "dev"
 Requires-Dist: furo; extra == "dev"
@@ -702,9 +702,9 @@ Requires-Dist: build; extra == "dev"
 Requires-Dist: plotly>=5.0.0; extra == "dev"
 Requires-Dist: pytest-asyncio>=0.26.0; extra == "dev"
 Requires-Dist: ruff>=0.9.0; extra == "dev"
-Requires-Dist: policyengine_core>=3.26.0; extra == "dev"
+Requires-Dist: policyengine_core>=3.26.1; extra == "dev"
 Requires-Dist: policyengine-uk==2.88.14; extra == "dev"
-Requires-Dist: policyengine-us==1.687.0; extra == "dev"
+Requires-Dist: policyengine-us==1.690.7; extra == "dev"
 Requires-Dist: towncrier>=24.8.0; extra == "dev"
 Requires-Dist: mypy>=1.11.0; extra == "dev"
 Requires-Dist: pytest-cov>=5.0.0; extra == "dev"

{policyengine-4.4.2 → policyengine-4.4.3}/examples/income_distribution_us.py

@@ -10,7 +10,6 @@ Run: python examples/income_distribution_us.py
 """
 
 import time
-from pathlib import Path
 
 import plotly.graph_objects as go
 from plotly.subplots import make_subplots
@@ -19,29 +18,25 @@ from policyengine.core import Simulation
 from policyengine.outputs.aggregate import Aggregate, AggregateType
 from policyengine.tax_benefit_models.us import (
     PolicyEngineUSDataset,
+    ensure_datasets,
     us_latest,
 )
 from policyengine.utils.plotting import COLORS, format_fig
 
 
 def load_representative_data(year: int = 2024) -> PolicyEngineUSDataset:
-    """Load representative household microdata for a given year."""
-    dataset_path = Path(__file__).parent / "data" / f"enhanced_cps_2024_year_{year}.h5"
-
-    if not dataset_path.exists():
-        raise FileNotFoundError(
-            f"Dataset not found at {dataset_path}. "
-            "Run create_datasets() from policyengine.tax_benefit_models.us first."
-        )
-
-    dataset = PolicyEngineUSDataset(
-        name=f"Enhanced CPS {year}",
-        description=f"Representative household microdata for {year}",
-        filepath=str(dataset_path),
-        year=year,
+    """Load representative household microdata for a given year.
+
+    Downloads the enhanced CPS dataset from HuggingFace on first run and
+    caches it locally under ./data, matching the pattern used by the
+    other examples (e.g. us_budgetary_impact.py).
+    """
+    datasets = ensure_datasets(
+        datasets=["hf://policyengine/policyengine-us-data/enhanced_cps_2024.h5"],
+        years=[year],
+        data_folder="./data",
     )
-    dataset.load()
-    return dataset
+    return datasets[f"enhanced_cps_2024_{year}"]
 
 
 def run_simulation(dataset: PolicyEngineUSDataset) -> Simulation:

{policyengine-4.4.2 → policyengine-4.4.3}/examples/us_budgetary_impact.py

@@ -8,6 +8,14 @@ Demonstrates the canonical policyengine.py workflow:
 5. Use ChangeAggregate for targeted single-metric queries
 
 Run: python examples/us_budgetary_impact.py
+
+System requirements: this example holds two full enhanced-CPS US
+simulations in memory simultaneously to compute reform-vs-baseline
+deltas. On a 2026 dataset (~101k people / 41k households) peak memory
+has been observed at ~7.5 GiB RSS / ~80 GiB virtual, and end-to-end
+runtime is ~20 minutes on a developer laptop. Run with at least 16 GiB
+of free RAM and expect to close other memory-heavy applications. See
+https://github.com/PolicyEngine/policyengine.py/issues/328 for tracking.
 """
 
 import datetime
@@ -35,7 +43,7 @@ def main():
         data_folder="./data",
     )
     dataset = datasets[f"enhanced_cps_2024_{year}"]
-    print(f"  Loaded: {dataset}")
+    print(f"  Loaded: {dataset.name} ({len(dataset.data.person):,} people)")
 
     # ── Step 2: Define a reform ──
     # Example: double the standard deduction for single filers

{policyengine-4.4.2 → policyengine-4.4.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "policyengine"
-version = "4.4.2"
+version = "4.4.3"
 description = "A package to conduct policy analysis using PolicyEngine tax-benefit models."
 readme = "README.md"
 authors = [
@@ -41,12 +41,12 @@ graph = [
     "networkx>=3.0",
 ]
 uk = [
-    "policyengine_core>=3.26.0",
+    "policyengine_core>=3.26.1",
     "policyengine-uk==2.88.14",
 ]
 us = [
-    "policyengine_core>=3.26.0",
-    "policyengine-us==1.687.0",
+    "policyengine_core>=3.26.1",
+    "policyengine-us==1.690.7",
 ]
 dev = [
     "pytest",
@@ -59,9 +59,9 @@ dev = [
     "plotly>=5.0.0",
     "pytest-asyncio>=0.26.0",
     "ruff>=0.9.0",
-    "policyengine_core>=3.26.0",
+    "policyengine_core>=3.26.1",
     "policyengine-uk==2.88.14",
-    "policyengine-us==1.687.0",
+    "policyengine-us==1.690.7",
     "towncrier>=24.8.0",
     "mypy>=1.11.0",
     "pytest-cov>=5.0.0",

{policyengine-4.4.2 → policyengine-4.4.3}/scripts/refresh_release_bundle.py

@@ -26,8 +26,11 @@ After running:
 from __future__ import annotations
 
 import argparse
+import os
 import sys
 
+os.environ.setdefault("POLICYENGINE_SKIP_COUNTRY_IMPORTS", "1")
+
 from policyengine.provenance.bundle import (
     refresh_release_bundle,
     regenerate_trace_tro,

{policyengine-4.4.2 → policyengine-4.4.3}/src/policyengine/__init__.py

@@ -28,17 +28,20 @@ Each country module exposes ``calculate_household``, ``model``
 (the pinned ``TaxBenefitModelVersion``), and the microsim helpers.
 """
 
+import os
 from importlib.util import find_spec
 
 from policyengine import outputs as outputs
 from policyengine.core import Simulation as Simulation
 
-if find_spec("policyengine_us") is not None:
+_SKIP_COUNTRY_IMPORTS = os.environ.get("POLICYENGINE_SKIP_COUNTRY_IMPORTS") == "1"
+
+if not _SKIP_COUNTRY_IMPORTS and find_spec("policyengine_us") is not None:
     from policyengine.tax_benefit_models import us as us
 else:  # pragma: no cover
     us = None  # type: ignore[assignment]
 
-if find_spec("policyengine_uk") is not None:
+if not _SKIP_COUNTRY_IMPORTS and find_spec("policyengine_uk") is not None:
     from policyengine.tax_benefit_models import uk as uk
 else:  # pragma: no cover
     uk = None  # type: ignore[assignment]

{policyengine-4.4.2 → policyengine-4.4.3}/src/policyengine/data/release_manifests/uk.json

@@ -1,8 +1,8 @@
 {
   "schema_version": 1,
-  "bundle_id": "uk-4.4.2",
+  "bundle_id": "uk-4.4.3",
   "country_id": "uk",
-  "policyengine_version": "4.4.2",
+  "policyengine_version": "4.4.3",
   "model_package": {
     "name": "policyengine-uk",
     "version": "2.88.14",

{policyengine-4.4.2 → policyengine-4.4.3}/src/policyengine/data/release_manifests/uk.trace.tro.jsonld

@@ -18,18 +18,18 @@
         "schema:url": "https://policyengine.org"
       },
       "schema:dateCreated": "2026-05-09T07:14:50.156352Z",
-      "schema:description": "TRACE TRO for certified runtime bundle uk-4.4.2 covering the bundle manifest, the certified dataset artifact, the country model wheel, and the country data release manifest when it is available.",
+      "schema:description": "TRACE TRO for certified runtime bundle uk-4.4.3 covering the bundle manifest, the certified dataset artifact, the country model wheel, and the country data release manifest when it is available.",
       "schema:name": "policyengine uk certified bundle TRO",
       "trov:createdWith": {
         "@type": "schema:SoftwareApplication",
         "schema:name": "policyengine",
-        "schema:softwareVersion": "4.4.2"
+        "schema:softwareVersion": "4.4.3"
       },
       "trov:hasArrangement": [
         {
           "@id": "arrangement/1",
           "@type": "trov:ArtifactArrangement",
-          "rdfs:comment": "Certified arrangement for bundle uk-4.4.2.",
+          "rdfs:comment": "Certified arrangement for bundle uk-4.4.3.",
           "trov:hasArtifactLocation": [
             {
               "@id": "arrangement/1/location/bundle_manifest",
@@ -75,7 +75,7 @@
             "@type": "trov:ResearchArtifact",
             "schema:name": "policyengine.py bundle manifest for uk",
             "trov:mimeType": "application/json",
-            "trov:sha256": "68c397241d27ca5ed2b479a53e6316710eb174d02d0d4c148937dc530ad260a4"
+            "trov:sha256": "77947ecc210cad81ab4ec04aedae8a9997930d3d3cc9c7f2092c9d62aa84e105"
           },
           {
             "@id": "composition/1/artifact/data_release_manifest",
@@ -102,7 +102,7 @@
         "trov:hasFingerprint": {
           "@id": "composition/1/fingerprint",
           "@type": "trov:CompositionFingerprint",
-          "trov:sha256": "f33a4fe9983eef2ffb8a329d97eba8ae4800f311720b91683e5d7acc4679e1a8"
+          "trov:sha256": "3643c10fef547c3b0a3885ae277c699bc70b1da872f2cbe8b211b269d292e8e7"
         }
       },
       "trov:hasPerformance": {
@@ -112,8 +112,8 @@
         "pe:certifiedBy": "policyengine.py bundled manifest",
         "pe:certifiedForModelVersion": "2.88.14",
         "pe:ciGitRef": "refs/heads/main",
-        "pe:ciGitSha": "eb7334698e9116cc06a5c6232ac4b4a91e2aef9a",
-        "pe:ciRunUrl": "https://github.com/PolicyEngine/policyengine.py/actions/runs/25598363562",
+        "pe:ciGitSha": "1718b493e4749faf62f0ffdff480205abdd20011",
+        "pe:ciRunUrl": "https://github.com/PolicyEngine/policyengine.py/actions/runs/25692889721",
         "pe:compatibilityBasis": "matching_data_build_fingerprint",
         "pe:dataBuildFingerprint": "sha256:dff5c5bb976ce254fa965ecfce6a0d84859fe1629a714ae28a79b3075522a0ae",
         "pe:dataBuildId": "policyengine-uk-data-1.55.5",

policyengine-4.4.3/src/policyengine/data/release_manifests/us.json

@@ -0,0 +1,54 @@
+{
+  "schema_version": 1,
+  "bundle_id": "us-4.4.3",
+  "country_id": "us",
+  "policyengine_version": "4.4.3",
+  "model_package": {
+    "name": "policyengine-us",
+    "version": "1.690.7",
+    "sha256": "5a7a541efabac98fa069d6845902cf5924c81db67383234b55dcd2b8bfcfc3ca",
+    "wheel_url": "https://files.pythonhosted.org/packages/2a/02/52109bae5f4767237b43bd72ce0bc4edf7925650a788053b2bc168caa5ae/policyengine_us-1.690.7-py3-none-any.whl"
+  },
+  "data_package": {
+    "name": "policyengine-us-data",
+    "version": "1.110.12",
+    "repo_id": "policyengine/policyengine-us-data",
+    "release_manifest_path": "releases/1.110.12/release_manifest.json",
+    "release_manifest_revision": "3aac4505ec10d31efc1b3799a1e6458a15853ecc"
+  },
+  "certified_data_artifact": {
+    "data_package": {
+      "name": "policyengine-us-data",
+      "version": "1.110.12"
+    },
+    "build_id": "policyengine-us-data-1.110.12",
+    "dataset": "enhanced_cps_2024",
+    "uri": "hf://policyengine/policyengine-us-data/enhanced_cps_2024.h5@1.110.12",
+    "sha256": "58a6639f7511b8d804701417e2647f0c3a77f51a3d90441037eaf004b1f00761"
+  },
+  "certification": {
+    "compatibility_basis": "exact_build_model_version",
+    "data_build_id": "policyengine-us-data-1.110.12",
+    "built_with_model_version": "1.690.7",
+    "certified_for_model_version": "1.690.7",
+    "certified_by": "policyengine.py bundled manifest",
+    "data_build_fingerprint": "sha256:9961ed1c5d00943a360724da560eee425eb9f99f91896f053dca74724c46e96e"
+  },
+  "default_dataset": "enhanced_cps_2024",
+  "datasets": {
+    "enhanced_cps_2024": {
+      "path": "enhanced_cps_2024.h5"
+    }
+  },
+  "region_datasets": {
+    "national": {
+      "path_template": "enhanced_cps_2024.h5"
+    },
+    "state": {
+      "path_template": "states/{state_code}.h5"
+    },
+    "congressional_district": {
+      "path_template": "districts/{district_code}.h5"
+    }
+  }
+}

{policyengine-4.4.2 → policyengine-4.4.3}/src/policyengine/data/release_manifests/us.trace.tro.jsonld

@@ -17,19 +17,19 @@
         "schema:name": "PolicyEngine",
         "schema:url": "https://policyengine.org"
       },
-      "schema:dateCreated": "2026-05-09T09:38:11.356906Z",
-      "schema:description": "TRACE TRO for certified runtime bundle us-4.4.2 covering the bundle manifest, the certified dataset artifact, the country model wheel, and the country data release manifest when it is available.",
+      "schema:dateCreated": "2026-05-11T18:53:05.508006Z",
+      "schema:description": "TRACE TRO for certified runtime bundle us-4.4.3 covering the bundle manifest, the certified dataset artifact, the country model wheel, and the country data release manifest when it is available.",
       "schema:name": "policyengine us certified bundle TRO",
       "trov:createdWith": {
         "@type": "schema:SoftwareApplication",
         "schema:name": "policyengine",
-        "schema:softwareVersion": "4.4.2"
+        "schema:softwareVersion": "4.4.3"
       },
       "trov:hasArrangement": [
         {
           "@id": "arrangement/1",
           "@type": "trov:ArtifactArrangement",
-          "rdfs:comment": "Certified arrangement for bundle us-4.4.2.",
+          "rdfs:comment": "Certified arrangement for bundle us-4.4.3.",
           "trov:hasArtifactLocation": [
             {
               "@id": "arrangement/1/location/bundle_manifest",
@@ -45,7 +45,7 @@
               "trov:hasArtifact": {
                 "@id": "composition/1/artifact/data_release_manifest"
               },
-              "trov:hasLocation": "https://huggingface.co/policyengine/policyengine-us-data/resolve/9cb665df0a546f9c3d79b496f8eb2dd55859d38d/releases/1.78.2/release_manifest.json"
+              "trov:hasLocation": "https://huggingface.co/policyengine/policyengine-us-data/resolve/3aac4505ec10d31efc1b3799a1e6458a15853ecc/releases/1.110.12/release_manifest.json"
             },
             {
               "@id": "arrangement/1/location/dataset",
@@ -53,7 +53,7 @@
               "trov:hasArtifact": {
                 "@id": "composition/1/artifact/dataset"
               },
-              "trov:hasLocation": "https://huggingface.co/policyengine/policyengine-us-data/resolve/1.78.2/enhanced_cps_2024.h5"
+              "trov:hasLocation": "https://huggingface.co/policyengine/policyengine-us-data/resolve/1.110.12/enhanced_cps_2024.h5"
             },
             {
               "@id": "arrangement/1/location/model_wheel",
@@ -61,7 +61,7 @@
               "trov:hasArtifact": {
                 "@id": "composition/1/artifact/model_wheel"
               },
-              "trov:hasLocation": "https://files.pythonhosted.org/packages/c3/36/5633f5a3996c915494154ec3852011b1a239ea06d9f08cb6287ab709618c/policyengine_us-1.687.0-py3-none-any.whl"
+              "trov:hasLocation": "https://files.pythonhosted.org/packages/2a/02/52109bae5f4767237b43bd72ce0bc4edf7925650a788053b2bc168caa5ae/policyengine_us-1.690.7-py3-none-any.whl"
             }
           ]
         }
@@ -75,53 +75,54 @@
             "@type": "trov:ResearchArtifact",
             "schema:name": "policyengine.py bundle manifest for us",
             "trov:mimeType": "application/json",
-            "trov:sha256": "ee037378f3e655360894dc9fe66f1d0c335321ff48789e3ebb0ff38805933d72"
+            "trov:sha256": "41e196a6263b8168d403058029c52ebab795e17024ac9ebef11ff876e36959e2"
           },
           {
             "@id": "composition/1/artifact/data_release_manifest",
             "@type": "trov:ResearchArtifact",
-            "schema:name": "policyengine-us-data release manifest 1.78.2",
+            "schema:name": "policyengine-us-data release manifest 1.110.12",
             "trov:mimeType": "application/json",
-            "trov:sha256": "83aafd9fa3d33a444c0277aa4b31b2041a8785910bee30660773ab96e1b1c8b9"
+            "trov:sha256": "17cfd2fbb31064834ed82c0fd7d8ae5c272fe7f24b1e48b226a4acf97ff4c5dd"
           },
           {
             "@id": "composition/1/artifact/dataset",
             "@type": "trov:ResearchArtifact",
             "schema:name": "enhanced_cps_2024",
             "trov:mimeType": "application/x-hdf5",
-            "trov:sha256": "4e92b340c3ea3e200ed5d55edf752ee1a13baf787442956fb67d25242fed13b5"
+            "trov:sha256": "58a6639f7511b8d804701417e2647f0c3a77f51a3d90441037eaf004b1f00761"
           },
           {
             "@id": "composition/1/artifact/model_wheel",
             "@type": "trov:ResearchArtifact",
-            "schema:name": "policyengine-us==1.687.0 wheel",
+            "schema:name": "policyengine-us==1.690.7 wheel",
             "trov:mimeType": "application/zip",
-            "trov:sha256": "cac7da3aa9ba4bf57009eee75d798217bbef7e1c5ca17646d472fad715ab634f"
+            "trov:sha256": "5a7a541efabac98fa069d6845902cf5924c81db67383234b55dcd2b8bfcfc3ca"
           }
         ],
         "trov:hasFingerprint": {
           "@id": "composition/1/fingerprint",
           "@type": "trov:CompositionFingerprint",
-          "trov:sha256": "a26e89fc2cd91ee54bc7526aa4c1a32c5b8ad35caefa1e7f2572e2fb8bb303a0"
+          "trov:sha256": "b84e895b3f19ffee5ec299b94ae2155448a12af2b3dc61d00f3d17003ecdf14a"
         }
       },
       "trov:hasPerformance": {
         "@id": "trp/1",
         "@type": "trov:TransparentResearchPerformance",
-        "pe:builtWithModelVersion": "1.647.0",
+        "pe:builtWithModelVersion": "1.690.7",
         "pe:certifiedBy": "policyengine.py bundled manifest",
-        "pe:certifiedForModelVersion": "1.687.0",
+        "pe:certifiedForModelVersion": "1.690.7",
         "pe:ciGitRef": "refs/heads/main",
-        "pe:ciGitSha": "eb7334698e9116cc06a5c6232ac4b4a91e2aef9a",
-        "pe:ciRunUrl": "https://github.com/PolicyEngine/policyengine.py/actions/runs/25598363562",
-        "pe:compatibilityBasis": "matching_data_build_fingerprint",
-        "pe:dataBuildId": "policyengine-us-data-1.78.2",
+        "pe:ciGitSha": "1718b493e4749faf62f0ffdff480205abdd20011",
+        "pe:ciRunUrl": "https://github.com/PolicyEngine/policyengine.py/actions/runs/25692889721",
+        "pe:compatibilityBasis": "exact_build_model_version",
+        "pe:dataBuildFingerprint": "sha256:9961ed1c5d00943a360724da560eee425eb9f99f91896f053dca74724c46e96e",
+        "pe:dataBuildId": "policyengine-us-data-1.110.12",
         "pe:emittedIn": "github-actions",
-        "rdfs:comment": "Certification of build policyengine-us-data-1.78.2 for policyengine-us 1.687.0.",
+        "rdfs:comment": "Certification of build policyengine-us-data-1.110.12 for policyengine-us 1.690.7.",
         "trov:accessedArrangement": {
           "@id": "arrangement/1"
         },
-        "trov:startedAtTime": "2026-05-09T09:38:11.356906Z",
+        "trov:startedAtTime": "2026-05-11T18:53:05.508006Z",
         "trov:wasConductedBy": {
           "@id": "trs"
         }

{policyengine-4.4.2 → policyengine-4.4.3}/src/policyengine/provenance/bundle.py

@@ -174,6 +174,66 @@ def _hf_dataset_sha256(repo_id: str, path: str, revision: str) -> str:
     return hasher.hexdigest()
 
 
+@dataclass(frozen=True)
+class _DataReleaseManifestFetch:
+    payload: dict
+    repo_commit: Optional[str]
+
+
+def _fetch_data_release_manifest(
+    repo_id: str,
+    release_manifest_path: str,
+    revision: str,
+) -> Optional[_DataReleaseManifestFetch]:
+    """Fetch a data release manifest from HF if one is available.
+
+    Older data releases may not have a machine-readable release manifest at the
+    inferred path. In that case the bundle refresh falls back to hashing the
+    dataset artifact directly.
+
+    Data releases are stored under versioned paths, but the HF repository does
+    not necessarily create a matching git tag for each data version. Try the
+    version revision first for repositories that do publish tags, then fall
+    back to ``main`` and persist the immutable ``x-repo-commit`` header.
+    """
+    headers = {"User-Agent": "policyengine.py"}
+    token = os.environ.get("HUGGING_FACE_TOKEN") or os.environ.get("HF_TOKEN")
+    if token:
+        headers["Authorization"] = f"Bearer {token}"
+
+    revisions = [revision]
+    if revision != "main":
+        revisions.append("main")
+
+    for candidate in revisions:
+        url = (
+            f"https://huggingface.co/{repo_id}/resolve/"
+            f"{candidate}/{release_manifest_path}"
+        )
+        try:
+            with urlopen(Request(url, headers=headers)) as f:
+                payload = json.load(f)
+                repo_commit = getattr(f, "headers", {}).get("x-repo-commit")
+                return _DataReleaseManifestFetch(
+                    payload=payload,
+                    repo_commit=repo_commit,
+                )
+        except (OSError, ValueError):
+            continue
+    return None
+
+
+def _updated_release_manifest_path(
+    current_path: str,
+    old_data: str,
+    new_data: str,
+) -> str:
+    """Preserve country-specific release-manifest layout while bumping versions."""
+    if old_data in current_path:
+        return current_path.replace(old_data, new_data)
+    return current_path
+
+
 # ---------------------------------------------------------------------------
 # Refresh result
 # ---------------------------------------------------------------------------
@@ -276,19 +336,90 @@ def refresh_release_bundle(
         )
     repo_id, dataset_path, _old_revision = repo_id_match.groups()
 
+    data_package_json = manifest_json["data_package"]
+    release_manifest_json = None
+    new_release_manifest_revision = None
+    new_release_manifest_path = data_package_json.get("release_manifest_path")
+    if new_data != old_data and new_release_manifest_path is not None:
+        new_release_manifest_path = _updated_release_manifest_path(
+            current_path=new_release_manifest_path,
+            old_data=old_data,
+            new_data=new_data,
+        )
+        release_manifest_fetch = _fetch_data_release_manifest(
+            repo_id=repo_id,
+            release_manifest_path=new_release_manifest_path,
+            revision=new_data,
+        )
+        if release_manifest_fetch is None:
+            raise ValueError(
+                "Could not fetch data release manifest "
+                f"{new_release_manifest_path!r} from {repo_id}@{new_data}. "
+                "Refusing to refresh a release-manifest-backed bundle with "
+                "partial certification metadata."
+            )
+        if release_manifest_fetch.repo_commit is None:
+            raise ValueError(
+                "Could not resolve an immutable HF commit for data release "
+                f"manifest {new_release_manifest_path!r} from {repo_id}@{new_data}."
+            )
+        release_manifest_json = release_manifest_fetch.payload
+        release_manifest_data_version = release_manifest_json.get(
+            "data_package", {}
+        ).get("version")
+        if release_manifest_data_version != new_data:
+            raise ValueError(
+                "Data release manifest "
+                f"{new_release_manifest_path!r} from {repo_id} declares "
+                f"version {release_manifest_data_version!r}, expected {new_data!r}."
+            )
+        new_release_manifest_revision = release_manifest_fetch.repo_commit
+
+    certified_dataset = (
+        current.certified_data_artifact.dataset
+        if current.certified_data_artifact is not None
+        else current.default_dataset
+    )
+    data_artifact_json = {}
+    if release_manifest_json is not None:
+        data_artifact_json = release_manifest_json.get("artifacts", {}).get(
+            certified_dataset,
+            {},
+        )
+        if not data_artifact_json:
+            raise ValueError(
+                "Data release manifest "
+                f"{new_release_manifest_path!r} from {repo_id}@{new_data} "
+                f"does not include certified dataset {certified_dataset!r}."
+            )
+    dataset_repo_id = data_artifact_json.get("repo_id", repo_id)
+    dataset_path = data_artifact_json.get("path", dataset_path)
+    dataset_revision = data_artifact_json.get("revision", new_data)
+
     # Only hit HF if the data version actually changed.
     if new_data != old_data:
-        new_dataset_sha256 = _hf_dataset_sha256(repo_id, dataset_path, new_data)
+        new_dataset_sha256 = data_artifact_json.get("sha256") or _hf_dataset_sha256(
+            dataset_repo_id,
+            dataset_path,
+            dataset_revision,
+        )
     else:
         new_dataset_sha256 = old_dataset_sha256
-    new_uri = f"hf://{repo_id}/{dataset_path}@{new_data}"
+    new_uri = f"hf://{dataset_repo_id}/{dataset_path}@{dataset_revision}"
     policyengine_version = _pyproject_version(pyproject_path)
 
     # Mutate the manifest JSON in place (keep unknown fields untouched).
     manifest_json["model_package"]["version"] = new_model
     manifest_json["model_package"]["sha256"] = new_wheel_sha256
     manifest_json["model_package"]["wheel_url"] = new_wheel_url
-    manifest_json["data_package"]["version"] = new_data
+    data_package_json["version"] = new_data
+    if new_data != old_data:
+        if new_release_manifest_path is not None:
+            data_package_json["release_manifest_path"] = new_release_manifest_path
+        if new_release_manifest_revision is not None:
+            data_package_json["release_manifest_revision"] = (
+                new_release_manifest_revision
+            )
     manifest_json["certified_data_artifact"]["data_package"]["version"] = new_data
     manifest_json["certified_data_artifact"]["build_id"] = (
         f"{current.data_package.name}-{new_data}"
@@ -299,10 +430,43 @@ def refresh_release_bundle(
         f"{current.data_package.name}-{new_data}"
     )
     manifest_json["certification"]["certified_for_model_version"] = new_model
+    if release_manifest_json is not None:
+        build = release_manifest_json.get("build") or {}
+        built_with_model = build.get("built_with_model_package") or {}
+        data_build_id = (
+            build.get("build_id") or f"{current.data_package.name}-{new_data}"
+        )
+        manifest_json["certified_data_artifact"]["build_id"] = data_build_id
+        certification_json = manifest_json["certification"]
+        certification_json["data_build_id"] = data_build_id
+        certification_json["certified_for_model_version"] = new_model
+        built_with_model_version = built_with_model.get("version")
+        if built_with_model_version is not None:
+            certification_json["built_with_model_version"] = built_with_model_version
+        if built_with_model.get("git_sha") is not None:
+            certification_json["built_with_model_git_sha"] = built_with_model["git_sha"]
+        else:
+            certification_json.pop("built_with_model_git_sha", None)
+        data_build_fingerprint = built_with_model.get("data_build_fingerprint")
+        if data_build_fingerprint is not None:
+            certification_json["data_build_fingerprint"] = data_build_fingerprint
+        else:
+            certification_json.pop("data_build_fingerprint", None)
+        if built_with_model_version == new_model:
+            certification_json["compatibility_basis"] = "exact_build_model_version"
+        elif data_build_fingerprint is not None:
+            certification_json["compatibility_basis"] = (
+                "matching_data_build_fingerprint"
+            )
+        else:
+            certification_json["compatibility_basis"] = (
+                "legacy_compatible_model_package"
+            )
 
     manifest_path.write_text(
         json.dumps(manifest_json, indent=2, sort_keys=False) + "\n"
     )
+    get_release_manifest.cache_clear()
     sync_release_manifest_policyengine_version(
         policyengine_version=policyengine_version,
         manifest_dir=manifest_dir,