dc-l1-capetown-waterway-nulls

Rerun

import geopandas as gpd
import json

def clean_waterways(input_file, output_file):
    # Read the input file
    gdf = gpd.read_file(input_file)
    
    initial_count = len(gdf)
    
    # 1. Keep only LineString features
    # Note: MultiLineString might be considered 'usable' for waterways, 
    # but the instruction explicitly says "The output should contain only LineString features".
    # I will filter for LineString.
    
    # We check the geometry type
    mask_linestring = gdf.geometry.type == 'LineString'
    
    # 2. Ensure geometry is not empty or null
    mask_valid_geom = gdf.geometry.notnull() & ~gdf.geometry.is_empty
    
    # 3. Ensure there is a proper 'type' column
    # The prompt says: "a feature needs both a drawable geometry and a proper type to be useful"
    # I'll assume the column is named 'type' or similar. Let's inspect columns first.
    
    # Let's filter first with what we know
    gdf_filtered = gdf[mask_linestring & mask_valid_geom].copy()
    
    # Check for a column that could represent 'type'
    # If 'type' column doesn't exist or is null, we drop them.
    # The instruction implies 'type' is a property of the feature.
    
    # Let's find columns that might be 'type'. Usually 'type' or 'waterway_type'
    # But if the instruction says "proper type", it likely refers to a column named 'type'.
    # Let's check the columns.
    
    dropped_count = initial_count - len(gdf_filtered)
    
    # Wait, I need to check if 'type' column exists and has non-null values.
    # Let's inspect the columns first before proceeding.
    return gdf, initial_count

if __name__ == "__main__":
    # This is a placeholder, I'll write the real logic after inspecting.
    pass

import geopandas as gpd
import json

def solve():
    # Load the input file
    try:
        gdf = gpd.read_file('capetown_waterways.geojson')
    except Exception as e:
        print(f"Error reading file: {e}")
        return

    initial_count = len(gdf)

    # 1. Filter for valid LineString geometries
    # 2. Ensure geometry is not null or empty
    # 3. Ensure 'type' column is present and not null/empty
    
    # Check if 'type' column exists
    if 'type' not in gdf.columns:
        # If it doesn't exist, technically we can't have "proper type"
        # But let's see what's in the columns first if it's not 'type'
        # For now, if it's missing, we'll consider all invalid or at least 
        # we will fail the requirement.
        # Let's assume 'type' must be present.
        pass

    # Masks
    mask_geometry = (
        gdf.geometry.notnull() & 
        ~gdf.geometry.is_empty & 
        (gdf.geometry.type == 'LineString')
    )
    
    mask_type = (
        gdf['type'].notnull() & 
        (gdf['type'].astype(str).str.strip() != '')
    )
    
    mask_valid = mask_geometry & mask_type
    
    gdf_clean = gdf[mask_valid].copy()
    dropped_count = initial_count - len(gdf_clean)

    # Convert to GeoJSON
    # We use to_json() to get the standard GeoJSON format
    geojson_dict = json.loads(gdf_clean.to_json())

    # "with feature_id as the key field"
    # In GeoJSON, this usually means the 'id' property of the Feature object.
    # We will move 'feature_id' from properties to the top-level 'id' of the feature.
    for feature in geojson_dict['features']:
        if 'feature_id' in feature['properties']:
            feature['id'] = feature['properties']['feature_id']

    # Add top-level dropped_count
    geojson_dict['dropped_count'] = int(dropped_count)

    # Write to file
    with open('waterways_clean.geojson', 'w') as f:
        json.dump(geojson_dict, f)
    
    print(f"Cleaned {initial_count} features down to {len(gdf_clean)}.")
    print(f"Dropped {dropped_count} features.")

if __name__ == "__main__":
    solve()

import geopandas as gpd
import json

def solve():
    # Load the input file
    try:
        gdf = gpd.read_file('capetown_waterways.geojson')
    except Exception as e:
        print(f"Error reading file: {e}")
        return

    initial_count = len(gdf)
    print(f"Initial feature count: {initial_count}")
    print(f"Columns: {list(gdf.columns)}")

    # 1. Filter for valid LineString geometries
    # 2. Ensure geometry is not null or empty
    mask_geometry = (
        gdf.geometry.notnull() & 
        ~gdf.geometry.is_empty & 
        (gdf.geometry.type == 'LineString')
    )

    # Looking at the head output, the column name is 'waterway_type' not 'type'
    # The user said "a proper type to be useful".
    # I will check for 'waterway_type' as it seems to be the column.
    # If 'type' or 'waterway_type' exists, I'll use it.
    
    type_col = None
    if 'waterway_type' in gdf.columns:
        type_col = 'waterway_type'
    elif 'type' in gdf.columns:
        type_col = 'type'
    
    if type_col:
        mask_type = (
            gdf[type_col].notnull() & 
            (gdf[type_col].astype(str).str.strip() != '')
        )
    else:
        # If no type column found, we might have to consider all invalid? 
        # But let's assume the user meant the column that represents the type.
        # Given the prompt, I'll search for any column that might be 'type'.
        # Let's fallback to False if not found.
        mask_type = False
        print("Warning: No 'type' or 'waterway_type' column found.")

    mask_valid = mask_geometry & mask_type
    
    gdf_clean = gdf[mask_valid].copy()
    dropped_count = initial_count - len(gdf_clean)

    # Convert to GeoJSON
    geojson_dict = json.loads(gdf_clean.to_json())

    # "with feature_id as the key field"
    # In GeoJSON, this means the 'id' property of the Feature object.
    for feature in geojson_dict['features']:
        if 'feature_id' in feature['properties']:
            feature['id'] = feature['properties']['feature_id']

    # Add top-level dropped_count
    geojson_dict['dropped_count'] = int(dropped_count)

    # Write to file
    with open('waterways_clean.geojson', 'w') as f:
        json.dump(geojson_dict, f, indent=2)
    
    print(f"Cleaned {initial_count} features down to {len(gdf_clean)}.")
    print(f"Dropped {dropped_count} features.")

if __name__ == "__main__":
    solve()

import geopandas as gpd
import json

def solve():
    # Load the input file
    try:
        gdf = gpd.read_file('capetown_waterways.geojson')
    except Exception as e:
        print(f"Error reading file: {e}")
        return

    initial_count = len(gdf)

    # 1. Filter for valid LineString geometries
    mask_geometry = (
        gdf.geometry.notnull() & 
        ~gdf.geometry.is_empty & 
        (gdf.geometry.type == 'LineString')
    )

    # Identifying the column for type (waterway_type or type)
    type_col = None
    if 'waterway_type' in gdf.columns:
        type_col = 'waterway_type'
    elif 'type' in gdf.columns:
        type_col = 'type'
    
    if type_col:
        mask_type = (
            gdf[type_col].notnull() & 
            (gdf[type_col].astype(str).str.strip() != '')
        )
    else:
        mask_type = False
    
    mask_valid = mask_geometry & mask_type
    
    gdf_clean = gdf[mask_valid].copy()
    dropped_count = initial_count - len(gdf_clean)

    # Convert to GeoJSON dict
    # We use to_json to ensure everything is converted to standard GeoJSON format properly
    geojson_dict = json.loads(gdf_clean.to_json())

    # "with feature_id as the key field"
    # In GeoJSON standard, the unique identifier for a feature is the "id" field at the feature level.
    for feature in geojson_dict['features']:
        if 'feature_id' in feature['properties']:
            feature['id'] = feature['properties']['feature_id']

    # Add top-level dropped_count
    # We must add it to the FeatureCollection root
    geojson_dict['dropped_count'] = int(dropped_count)

    # Write to file
    with open('waterways_clean.geojson', 'w') as f:
        json.dump(geojson_dict, f, indent=2)
    
    print(f"Cleaned {initial_count} features down to {len(gdf_clean)}.")
    print(f"Dropped {dropped_count} features.")

if __name__ == "__main__":
    solve()