gtrack 0.3.1__tar.gz → 0.3.2__tar.gz

{gtrack-0.3.1 → gtrack-0.3.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gtrack
-Version: 0.3.1
+Version: 0.3.2
 Summary: GPlates-based Tracking of Lithosphere and Kinematics
 Author: S. Ghelichkhani
 License: MIT

{gtrack-0.3.1 → gtrack-0.3.2}/gtrack/point_rotation.py

@@ -148,6 +148,55 @@ class PointCloud:
         xyz = LatLon2XYZ(latlon)
         return cls(xyz=xyz, properties=properties or {})
 
+    @classmethod
+    def from_xyz(
+        cls,
+        xyz: np.ndarray,
+        properties: Optional[Dict[str, np.ndarray]] = None,
+        normalize_to_earth: bool = False
+    ) -> "PointCloud":
+        """
+        Create PointCloud from Cartesian XYZ coordinates.
+
+        This is a convenience classmethod that provides symmetry with from_latlon.
+        It can optionally normalize points to Earth's surface.
+
+        Parameters
+        ----------
+        xyz : np.ndarray
+            Cartesian coordinates, shape (N, 3).
+        properties : dict, optional
+            Properties to attach to the points.
+        normalize_to_earth : bool, default=False
+            If True, normalize points to Earth's radius (~6.3781e6 m).
+            Useful when input points are on a unit sphere.
+
+        Returns
+        -------
+        PointCloud
+            New PointCloud with the given XYZ coordinates.
+
+        Examples
+        --------
+        >>> # Points already at Earth's radius
+        >>> xyz = np.array([[6378100, 0, 0], [0, 6378100, 0]])
+        >>> cloud = PointCloud.from_xyz(xyz)
+        >>>
+        >>> # Points on unit sphere, scale to Earth
+        >>> xyz_unit = np.array([[1, 0, 0], [0, 1, 0]])
+        >>> cloud = PointCloud.from_xyz(xyz_unit, normalize_to_earth=True)
+        """
+        from .geometry import EARTH_RADIUS
+        xyz = np.asarray(xyz)
+
+        if normalize_to_earth:
+            # Normalize to unit sphere, then scale to Earth's radius
+            norms = np.linalg.norm(xyz, axis=1, keepdims=True)
+            norms = np.maximum(norms, 1e-10)  # Avoid division by zero
+            xyz = xyz / norms * EARTH_RADIUS
+
+        return cls(xyz=xyz, properties=properties or {})
+
     def add_property(self, name: str, values: np.ndarray) -> None:
         """
         Add or update a property.

{gtrack-0.3.1 → gtrack-0.3.2}/gtrack.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gtrack
-Version: 0.3.1
+Version: 0.3.2
 Summary: GPlates-based Tracking of Lithosphere and Kinematics
 Author: S. Ghelichkhani
 License: MIT

{gtrack-0.3.1 → gtrack-0.3.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "gtrack"
-version = "0.3.1"
+version = "0.3.2"
 description = "GPlates-based Tracking of Lithosphere and Kinematics"
 readme = "README.md"
 requires-python = ">=3.9"

{gtrack-0.3.1 → gtrack-0.3.2}/tests/test_point_rotation.py

@@ -48,6 +48,67 @@ class TestPointCloud:
         distances = np.linalg.norm(cloud.xyz, axis=1)
         np.testing.assert_allclose(distances, EARTH_RADIUS, rtol=1e-10)
 
+    def test_from_xyz_basic(self):
+        """Test creating PointCloud via from_xyz classmethod."""
+        xyz = normalize_to_sphere(np.random.randn(50, 3))
+        cloud = PointCloud.from_xyz(xyz)
+
+        assert cloud.n_points == 50
+        np.testing.assert_allclose(cloud.xyz, xyz)
+        assert cloud.plate_ids is None
+        assert len(cloud.properties) == 0
+
+    def test_from_xyz_with_properties(self):
+        """Test from_xyz passes properties correctly."""
+        xyz = normalize_to_sphere(np.random.randn(30, 3))
+        props = {'depth': np.random.rand(30), 'temp': np.random.rand(30)}
+        cloud = PointCloud.from_xyz(xyz, properties=props)
+
+        assert cloud.n_points == 30
+        assert 'depth' in cloud.properties
+        assert 'temp' in cloud.properties
+
+    def test_from_xyz_normalize_to_earth(self):
+        """Test from_xyz with normalize_to_earth scales to Earth's radius."""
+        # Points on unit sphere
+        xyz_unit = np.array([[1.0, 0.0, 0.0], [0.0, 1.0, 0.0], [0.0, 0.0, 1.0]])
+        cloud = PointCloud.from_xyz(xyz_unit, normalize_to_earth=True)
+
+        distances = np.linalg.norm(cloud.xyz, axis=1)
+        np.testing.assert_allclose(distances, EARTH_RADIUS, rtol=1e-10)
+
+    def test_from_xyz_normalize_arbitrary_radius(self):
+        """Test from_xyz normalizes points at arbitrary radii to Earth's radius."""
+        # Points at radius 100
+        xyz = np.array([[100.0, 0.0, 0.0], [0.0, 0.0, 50.0]])
+        cloud = PointCloud.from_xyz(xyz, normalize_to_earth=True)
+
+        distances = np.linalg.norm(cloud.xyz, axis=1)
+        np.testing.assert_allclose(distances, EARTH_RADIUS, rtol=1e-10)
+
+        # Direction should be preserved
+        direction_original = xyz / np.linalg.norm(xyz, axis=1, keepdims=True)
+        direction_result = cloud.xyz / np.linalg.norm(cloud.xyz, axis=1, keepdims=True)
+        np.testing.assert_allclose(direction_result, direction_original, atol=1e-10)
+
+    def test_from_xyz_no_normalize(self):
+        """Test from_xyz without normalization preserves coordinates exactly."""
+        xyz = np.array([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]])
+        cloud = PointCloud.from_xyz(xyz, normalize_to_earth=False)
+
+        np.testing.assert_array_equal(cloud.xyz, xyz)
+
+    def test_from_xyz_near_zero_vector(self):
+        """Test from_xyz handles near-zero vectors without crashing."""
+        xyz = np.array([[1e-15, 1e-15, 1e-15], [1.0, 0.0, 0.0]])
+        cloud = PointCloud.from_xyz(xyz, normalize_to_earth=True)
+
+        # Should not raise; second point should be at Earth's radius
+        distances = np.linalg.norm(cloud.xyz, axis=1)
+        np.testing.assert_allclose(distances[1], EARTH_RADIUS, rtol=1e-10)
+        # First point (near-zero) should still produce a finite result
+        assert np.all(np.isfinite(cloud.xyz[0]))
+
     def test_latlon_property(self):
         """Test that latlon property correctly converts from XYZ."""
         original_latlon = np.array([