Source code for pose_format.utils.fast_math

[docs]def distance_batch(p1s, p2s): """ Computes Euclidean distance between two sets of points in batch Parameters ---------- p1s : array-like array of shape (N, D) where N; number of points & D ; dimensionality of each point p2s : array-like array of shape (N, D) with N; number of points & D; dimensionality of each point Returns ------- array-like array of shape (N,) with euclidean distances between points in `p1s` and `p2s` Examples -------- >>> distance_batch(np.array([[0, 0], [1, 1]]), np.array([[1, 1], [2, 2]])) array([1.41421356, 1.41421356]) Note ---- Function assumes that the inputs `p1s` and `p2s` have the same shape """ squared = (p1s - p2s)**2 summed = squared.sum(axis=-1) return summed**0.5