Add Australian Geographic Correspondence Dataset

- Master correspondence table: 95,674 records linking POA↔LGA↔SA2↔Branch
- Complete ABS ASGS Edition 3 spatial intersections with weights
- HCF branch catchments: 9 locations with 15km service areas
- Multiple formats: Parquet (efficient) and CSV (compatible)
- Comprehensive documentation and usage examples

Dataset enables feature aggregation across Australian geographic hierarchies.
Built with real ABS data using modern spatial processing.

Files:
- master_correspondence_table.parquet (271KB) - Recommended format
- master_correspondence_table.csv (11MB) - CSV format via Git LFS
- branch_catchments.parquet (38KB) - HCF service areas
- branch_catchments.csv (38KB) - HCF service areas CSV

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <[email protected]>

.gitattributes

@@ -57,3 +57,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+*.csv filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,152 @@
+---
+license: mit
+task_categories:
+- feature-extraction
+- tabular-classification
+language:
+- en
+tags:
+- geography
+- australia
+- correspondence
+- spatial-data
+- abs
+- government
+size_categories:
+- 10K<n<100K
+---
+
+# Australian Geographic Correspondence Table (ABS ASGS Edition 3)
+
+A comprehensive correspondence table linking Australian geographic hierarchies (POA ↔ LGA ↔ SA2 ↔ Branch Catchments) built using real Australian Bureau of Statistics (ABS) data.
+
+## Dataset Description
+
+This dataset provides complete spatial correspondence relationships between different levels of the Australian Statistical Geography Standard (ASGS) Edition 3, enabling accurate feature aggregation and geographic analysis across Australia.
+
+### What's Included
+
+- **Master Correspondence Table**: 95,674 records linking POA, LGA, SA2, and HCF branch catchments
+- **Branch Catchments**: 9 unique HCF service areas with 15km radius coverage
+- **Spatial Weights**: Area-based intersection ratios for accurate data aggregation
+- **Geographic Coverage**: Complete national coverage of Australian boundaries
+
+## Files
+
+| File | Records | Size | Description |
+|------|---------|------|-------------|
+| `master_correspondence_table.parquet` | 95,674 | 271KB | Main correspondence table (recommended) |
+| `master_correspondence_table.csv` | 95,674 | 11MB | Main correspondence table (CSV format) |
+| `branch_catchments.parquet` | 9 | 38KB | HCF branch service areas |
+| `branch_catchments.csv` | 9 | 38KB | HCF branch service areas (CSV format) |
+
+## Schema
+
+### Master Correspondence Table
+
+| Column | Type | Description |
+|--------|------|-------------|
+| `poa_code` | string | 4-digit Postal Area code |
+| `poa_name` | string | Postal Area name |
+| `lga_code` | string | 5-digit Local Government Area code |
+| `lga_name` | string | Local Government Area name |
+| `sa2_code` | string | 9-digit Statistical Area Level 2 code |
+| `sa2_name` | string | Statistical Area Level 2 name |
+| `branch_id` | string | Unique branch identifier (null if >15km from any branch) |
+| `branch_name` | string | HCF branch facility name (null if >15km from any branch) |
+| `branch_type` | string | "Dental" or "Eyecare" (null if >15km from any branch) |
+| `branch_lat` | float | Branch latitude (null if >15km from any branch) |
+| `branch_long` | float | Branch longitude (null if >15km from any branch) |
+| `poa_to_lga_weight` | float | Area-weighted intersection ratio (0-1) |
+| `lga_to_sa2_weight` | float | Area-weighted intersection ratio (0-1) |
+| `catchment_overlap_pct` | float | % of catchment overlapping with area (0 if >15km) |
+| `distance_to_branch_km` | float | Distance from area centroid to nearest branch (0 if >15km) |
+| `record_type` | string | "geographic_only" or "master_correspondence" |
+
+## Geographic Coverage
+
+- **2,644 POA boundaries** (Postal Areas) - ABS 2021 data
+- **566 LGA boundaries** (Local Government Areas) - ABS 2022 data  
+- **2,473 SA2 boundaries** (Statistical Area Level 2) - ABS 2021 data
+- **9 HCF branch catchments** (15km radius service areas)
+
+## Use Cases
+
+### Feature Aggregation
+```python
+import pandas as pd
+
+# Load correspondence table
+corr = pd.read_parquet('master_correspondence_table.parquet')
+
+# Aggregate postcode-level data to LGA level
+postcode_data = pd.read_csv('your_postcode_data.csv')
+lga_aggregated = postcode_data.merge(corr, left_on='postcode', right_on='poa_code') \
+    .groupby('lga_code').agg({
+        'population': 'sum',
+        'income': lambda x: np.average(x, weights=corr.loc[x.index, 'poa_to_lga_weight'])
+    })
+```
+
+### Service Area Analysis
+```python
+# Find all areas within HCF branch catchments
+branch_areas = corr[corr['branch_id'].notna()]
+coverage_by_branch = branch_areas.groupby('branch_name').agg({
+    'sa2_code': 'count',
+    'catchment_overlap_pct': 'mean'
+})
+```
+
+## Data Quality
+
+- **Spatial Accuracy**: Built using official ABS ASGS Edition 3 boundaries
+- **Coverage**: 100% coverage of Australian POA, LGA, and SA2 areas
+- **Validation**: Comprehensive geometric and topological validation
+- **Weights**: Area-based intersection calculations for precise aggregation
+
+## Source Data
+
+- **ABS ASGS Edition 3**: Australian Statistical Geography Standard digital boundaries
+- **POA 2021**: Postal Area boundaries (GDA2020 coordinate system)
+- **LGA 2022**: Local Government Area boundaries (GDA2020 coordinate system)
+- **SA2 2021**: Statistical Area Level 2 boundaries (GDA2020 coordinate system)
+
+## Methodology
+
+1. **Data Download**: Automated download of official ABS shapefiles
+2. **Spatial Processing**: Geometric intersection calculations using GeoPandas
+3. **Weight Calculation**: Area-based ratios for accurate feature aggregation
+4. **Catchment Generation**: 15km radius buffers in Australian Albers projection
+5. **Validation**: Comprehensive quality checks and coverage analysis
+
+## License
+
+This dataset is licensed under the MIT License. The underlying ABS data is available under Creative Commons Attribution 4.0 International License.
+
+## Citation
+
+```bibtex
+@dataset{hcf_abs_correspondence_2024,
+  title={Australian Geographic Correspondence Table (ABS ASGS Edition 3)},
+  author={HCF Data Science Team},
+  year={2024},
+  publisher={Hugging Face},
+  url={https://huggingface.co/datasets/massomo/abscorrespondance}
+}
+```
+
+## Technical Details
+
+- **Built with**: Python, GeoPandas, Shapely, PyProj
+- **Coordinate Systems**: WGS84 (EPSG:4326) and Australian Albers (EPSG:3577)
+- **Processing**: Spatial intersection, area calculation, distance analysis
+- **Export**: Parquet and CSV formats with comprehensive metadata
+
+## Contact
+
+For questions about this dataset or the underlying methodology, please refer to the source repository or open an issue on the project GitHub page.
+
+---
+
+*Generated using the Master Correspondence Table Builder - a modern Python replacement for legacy SAS Enterprise Guide geographic workflows.*

branch_catchments.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4b29705cf5fe9451b5cf4a818f02b3eaa3a9fa5fcc685d0fa88fe226cd79900
+size 38512

branch_catchments.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e671cd29e6ab78e6c8ddaf7e7bc1dfdc5be8847397838f870bfd335f9d71e3b2
+size 39183

master_correspondence_table.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df823555a408de1ce0fe90af16a7e7c07d96c9abd4edf10d79b1a0b1468d4bac
+size 11365436

master_correspondence_table.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6f2eb58ceba119bc399dbb758728f59bc17d9ccde1b636a83156923aa48e0db
+size 277194