Enriched pipeline stage (enriched_t0p005/)
The enriched intermediate pipeline stage: adm0–adm4.geojson with PSGC psgc_code / psgc_name / psgc_status / match_confidence, output of enrich.py and input to the classifier.
The enriched tier is the intermediate enrichment stage of Pipeline B. It is the output of enrich.py and the input to classifier.py. It is kept in-repo for traceability — most users should use the curated hierarchical per-class files instead.
What it is
adm0–adm4.geojson — the full NAMRIA polygon set at each administrative level, with each polygon annotated with its psgc_code, psgc_name, psgc_status, and match_confidence.
Intermediate pipeline stage — most users want the curated hierarchical per-class files instead.
File inventory
Kept in-repo at 2023-10-24/enriched_t0p005/ for traceability:
| File | Contents |
|---|---|
adm0.geojson–adm4.geojson |
NAMRIA polygons at each admin level, each annotated with psgc_code, psgc_name, psgc_status, and match_confidence |
diff.json |
PSGC-vs-NAMRIA reconciliation diff (matches, gaps, mismatches) |
summary.md |
human-readable summary of the enrichment pass |
adm3 mixes municipalities and cities because this stage is pre-classification — features are not yet split by city class. The classifier + splitter (see Pipeline B) separate them into the curated per-class files.
Download (in-repo)
Fetch a single file via raw.githubusercontent.com:
curl -O https://raw.githubusercontent.com/bendlikeabamboo/barangay-boundaries-repository/main/2023-10-24/enriched_t0p005/adm4.geojsonHow enrichment works (name normalization + HUC mapping + fuzzy match) is documented in Pipeline B; the HUC reconciliation in HUC mapping; the match_confidence values (exact / huc_map / fuzzy) in feature properties.