crs-l1-paris-lambert93

Rerun

# /// script
# requires-python = ">=3.12"
# dependencies = ["geopandas", "shapely", "pyproj", "pyogrio"]
# ///

import geopandas as gpd

def main():
    input_file = "paris_buildings_wgs84.geojson"
    output_file = "paris_buildings_lambert93.gpkg"
    
    # Load the data
    print(f"Reading {input_file}...")
    gdf = gpd.read_file(input_file)
    
    # Check current CRS
    print(f"Current CRS: {gdf.crs}")
    
    # Convert to Lambert-93 (EPSG:2154)
    # Lambert-93 is the standard official projection for France (RGF93 / Lambert-93)
    print("Converting to EPSG:2154 (Lambert-93)...")
    gdf = gdf.to_crs(epsg=2154)
    
    # Ensure every geometry is a plain Polygon instead of a MultiPolygon
    # The instruction says: "keep every geometry as a plain Polygon rather than turning it into a MultiPolygon"
    # This usually means if we have a MultiPolygon, we should explode it into individual Polygons.
    print("Exploding MultiPolygons into Polygons...")
    gdf = gdf.explode(index_parts=False)
    
    # The instruction says: "use `id` as the key" for the GeoPackage.
    # In GeoPackage, this usually refers to the primary key. 
    # geopandas/pyogrio will write the index or a specific column as a key if configured,
    # but typically we just ensure 'id' is a column and it's preserved.
    # Let's check if 'id' exists or if we need to create it.
    if 'id' not in gdf.columns:
        print("Column 'id' not found, creating it from index...")
        gdf['id'] = gdf.index

    # Write to GeoPackage
    print(f"Writing to {output_file}...")
    # We use layer name 'buildings' or similar, but the instruction doesn't specify layer name.
    # We will just write the file.
    gdf.to_file(output_file, driver="GPKG")
    
    print("Done.")

if __name__ == "__main__":
    main()

# /// script
# requires-python = ">=3.12"
# dependencies = ["geopandas", "shapely", "pyproj", "pyogrio"]
# ///

import geopandas as gpd

def main():
    input_file = "paris_buildings_wgs84.geojson"
    output_file = "paris_buildings_lambert93.gpkg"
    
    # Load the data
    print(f"Reading {input_file}...")
    gdf = gpd.read_file(input_file)
    
    # Check current CRS
    print(f"Current CRS: {gdf.crs}")
    
    # Convert to Lambert-93 (EPSG:2154)
    print("Converting to EPSG:2154 (Lambert-93)...")
    gdf = gdf.to_crs(epsg=2154)
    
    # Explode MultiPolygons into Polygons
    print("Exploding MultiPolygons into Polygons...")
    gdf = gdf.explode(index_parts=False)
    
    # Ensure 'id' is the first column and used as the index/key
    # Many GIS models expect a unique integer 'id' column.
    # If 'id' exists, we keep it. If not, we might need to create it.
    # The instruction says "use `id` as the key".
    
    if 'id' not in gdf.columns:
        print("Column 'id' not found, creating it...")
        gdf['id'] = range(len(gdf))
    
    # Reorder columns to put 'id' first for good measure
    cols = ['id'] + [c for c in gdf.columns if c != 'id']
    gdf = gdf[cols]

    # Write to GeoPackage
    print(f"Writing to {output_file}...")
    gdf.to_file(output_file, driver="GPKG")
    
    # Verification
    print("Verifying output...")
    check_gdf = gpd.read_file(output_file)
    print(f"Output CRS: {check_gdf.crs}")
    print(f"Geometry type: {check_gdf.geom_type.unique()}")
    print(f"Columns: {check_gdf.columns.tolist()}")
    print(f"Number of rows: {len(check_gdf)}")

    print("Done.")

if __name__ == "__main__":
    main()