England only

pipeline/utils/england_geometry.py

@@ -0,0 +1,33 @@
+"""England boundary polygon for accurate point-in-country filtering.
+
+Uses shapely prepared geometry for fast single-point checks (osmium handlers)
+and vectorized shapely.contains for batch checks (Polars DataFrames).
+"""
+
+import json
+from pathlib import Path
+
+import numpy as np
+import shapely
+from shapely.geometry import shape
+from shapely.prepared import PreparedGeometry, prep
+
+
+def load_england_polygon(geojson_path: Path) -> PreparedGeometry:
+    """Load England boundary as a prepared shapely polygon for fast contains checks."""
+    with open(geojson_path) as f:
+        data = json.load(f)
+    geometry = shape(data["features"][0]["geometry"])
+    return prep(geometry)
+
+
+def in_england_mask(geojson_path: Path, lats: np.ndarray, lngs: np.ndarray) -> np.ndarray:
+    """Vectorized check: which (lat, lng) points are within England.
+
+    Returns a boolean numpy array.
+    """
+    with open(geojson_path) as f:
+        data = json.load(f)
+    polygon = shape(data["features"][0]["geometry"])
+    pts = shapely.points(lngs, lats)
+    return shapely.contains(polygon, pts)