geo-l1-capetown-building-centroids

Rerun

# /// script
# requires-python = ">=3.12"
# dependencies = ["geopandas", "shapely"]
# ///
import geopandas as gpd

def main():
    input_file = "capetown_buildings.shp"
    output_file = "building_centroids.geojson"

    # Read the buildings
    gdf = gpd.read_file(input_file)

    # Check for building ID column. 
    # We don't know the exact name, so let's inspect it in the script if needed, 
    # but the prompt implies there is a building identifier.
    # I will list columns first or just try to find the identifier.
    # Usually it's 'id', 'fid', 'building_id', etc.
    # The prompt says: "ensure the output carries a building_id column with the building identifier from the input"
    
    # Let's first see the columns
    print(f"Columns in input: {gdf.columns.tolist()}")

    # We need to find which column represents the building identifier. 
    # If it's not explicitly named 'building_id', we might need to rename it.
    # Common names: 'id', 'fid', 'OBJECTID', 'ID'.
    # The prompt says "ensure the output carries a building_id column".
    
    # Let's assume the identifier is in some column. 
    # I'll try to find a reasonable candidate if 'building_id' is not there.
    # However, many shapefiles have 'id' or similar.
    
    # For now, let's find the best candidate or if there's a column we should use.
    # Since I can't "guess" perfectly without looking, I'll print them.
    
    # Let's identify the ID column. 
    # I will look for columns that look like IDs.
    possible_id_cols = ['building_id', 'id', 'FID', 'fid', 'OBJECTID', 'ID']
    id_col = None
    for col in possible_id_cols:
        if col in gdf.columns:
            id_col = col
            break
    
    if id_col is None:
        # If no standard ID found, maybe the index or first column?
        # But let's assume there is one. 
        # Let's check if there's any column that's a good candidate.
        # For now, I'll just take the first column if I don't find a match? 
        # No, better to inspect.
        pass

    # Actually, let's just assume there's a column that serves as ID.
    # I'll find the index/ID column.
    # For the sake of the task, I'll try to find it.
    
    # Let's refine the script to:
    # 1. Load data.
    # 2. Identify ID column.
    # 3. Create centroids.
    # 4. Keep ONLY 'building_id' and 'geometry'.
    # 5. Save.

    # Wait, I'll just write a script that tries to find the ID column or 
    # uses the most likely one.

    # Let's run a quick inspection first.
main()

# /// script
# requires-python = ">=3.12"
# dependencies = ["geopandas", "shapely"]
# ///
import geopandas as gpd

def main():
    input_file = "capetown_buildings.shp"
    gdf = gpd.read_file(input_file)
    print(gdf.head())
    print(gdf.columns)

if __name__ == "__main__":
    main()

# /// script
# requires-python = ">=3.12"
# dependencies = ["geopandas", "shapely"]
# ///
import geopandas as gpd

def main():
    input_file = "capetown_buildings.shp"
    gdf = gpd.read_file(input_file)
    print("Columns:", gdf.columns.tolist())
    print("First few rows:\n", gdf.head())

if __name__ == "__main__":
    main()

# /// script
# requires-python = ">=3.12"
# dependencies = ["geopandas", "shapely"]
# ///
import geopandas as gpd

def main():
    input_file = "capetown_buildings.shp"
    gdf = gpd.read_file(input_file)
    print("Columns:", gdf.columns.tolist())
    print("Head:\n", gdf.head())

if __name__ == "__main__":
    main()

# /// script
# requires-python = ">=3.12"
# dependencies = ["geopandas", "shapely"]
# ///
import geopandas as gpd

def main():
    input_file = "capetown_buildings.shp"
    gdf = gpd.read_file(input_file)
    print("Columns:", gdf.columns.tolist())
    print("Head:\n", gdf.head())

if __name__ == "__main__":
    main()

# /// script
# requires-python = ">=3.12"
# dependencies = ["geopandas", "shapely"]
# ///
import geopandas as gpd

def main():
    input_file = "capetown_buildings.shp"
    output_file = "building_centroids.geojson"

    # Read the buildings
    gdf = gpd.read_file(input_file)

    # The input has 'building_i' as the ID column (truncated due to Shapefile limitations)
    # Rename it to 'building_id' as requested
    gdf = gdf.rename(columns={'building_i': 'building_id'})

    # Ensure building_id is not empty (though the prompt says "ensure... every row has a non-empty building_id")
    # We'll filter out any rows where building_id might be NaN or empty if they exist
    gdf = gdf[gdf['building_id'].notna()]
    gdf = gdf[gdf['building_id'].astype(str).str.strip() != '']

    # Create centroids
    # Use representative_point() instead of centroid to ensure the point is within the polygon
    # The prompt says "centroid layer", but for buildings, representative_point is often safer.
    # However, "centroid" is explicitly requested. I'll use centroid.
    gdf_centroids = gdf.copy()
    gdf_centroids['geometry'] = gdf['geometry'].centroid

    # Keep only 'building_id' and 'geometry'
    gdf_centroids = gdf_centroids[['building_id', 'geometry']]

    # Write to GeoJSON
    gdf_centroids.to_file(output_file, driver='GeoJSON')
    print(f"Successfully wrote {len(gdf_centroids)} centroids to {output_file}")

if __name__ == "__main__":
    main()