Add transform_points method to Line

skspatial/objects/line.py

@@ -1,7 +1,7 @@
 import numpy as np
 from dpcontracts import require, ensure, types
 
-from skspatial.transformation import mean_center
+from skspatial.transformation import mean_center, normalize_dimension
 from .base_line_plane import _BaseLinePlane
 from .point import Point
 from .vector import Vector
@@ -297,6 +297,7 @@ def intersect_line(self, other):
 
     @classmethod
     @types(points=np.ndarray)
+    @require("The points are all finite.", lambda args: np.isfinite(args.points).all())
     @require("There must be at least two points.", lambda args: args.points.shape[0] >= 2)
     @ensure("The output must be a line.", lambda _, result: isinstance(result, Line))
     def best_fit(cls, points):
@@ -334,3 +335,43 @@ def best_fit(cls, points):
         direction = Vector(vh[0, :])
 
         return cls(centroid, direction)
+
+    @types(points=np.ndarray)
+    @require("The points are all finite.", lambda args: np.isfinite(args.points).all())
+    @ensure("There is one coordinate for each input point.", lambda args, result: result.size == args.points.shape[0])
+    @ensure("The output is a 1D array.", lambda _, result: result.ndim == 1)
+    @ensure("The coordinates are all finite.", lambda _, result: np.isfinite(result).all())
+    def transform_points(self, points):
+        """
+        Transform points to a one-dimensional coordinate system defined by a line.
+
+        The point on the line acts as the origin of the coordinate system.
+
+        The line is analagous to an x-axis. The output coordinates represent the
+        x-values of points on this line.
+
+        Parameters
+        ----------
+        points : ndarray
+            (n, d) array of n points with dimension d.
+
+        Returns
+        -------
+        coordinates : ndarray
+            One-dimensional coordinates.
+
+        Examples
+        --------
+        >>> from skspatial.objects import Line
+
+        >>> line = Line(point=[0, 0], vector=[1, 0])
+        >>> points = np.array([[10, 2], [3, 4], [-5, 5]])
+
+        >>> line.transform_points(points)
+        array([10.,  3., -5.])
+
+        """
+        vectors_to_points = normalize_dimension(points) - self.point
+        coordinates = np.apply_along_axis(np.dot, 1, vectors_to_points, self.direction)
+
+        return coordinates

skspatial/transformation.py

@@ -3,6 +3,8 @@
 import numpy as np
 from dpcontracts import require, ensure, types
 
+from skspatial.objects import Point
+
 
 @types(points=np.ndarray)
 @require("The input must be a 2D array of points", lambda args: args.points.ndim == 2)
@@ -77,7 +79,41 @@ def mean_center(points):
 
     """
     centroid = get_centroid(points)
-
     points_centered = points - centroid
 
     return points_centered, centroid
+
+
+@ensure("The output is an ndarray.", lambda _, result: isinstance(result, np.ndarray))
+def normalize_dimension(seq_points):
+    """
+    Normalize the dimension of a set of points.
+
+    Each point is converted to a `Point` object.
+    The Point objects are then stacked into one ndarray.
+
+    Parameters
+    ----------
+    seq_points : sequence
+         Sequence of array_like objects.
+
+    Returns
+    -------
+    ndarray
+        Each row corresponds to a point in the input sequence.
+
+    Examples
+    --------
+    >>> from skspatial.transformation import normalize_dimension
+
+    >>> seq_points = ([1], [1, 2], [1, 2, 3])
+
+    >>> normalize_dimension(seq_points)
+    array([[1., 0., 0.],
+           [1., 2., 0.],
+           [1., 2., 3.]])
+
+    """
+    list_points = [Point(x) for x in seq_points]
+
+    return np.vstack(list_points)