Amritpal Singh

This snippet computes C-accelerated Dynamic Time Warping using dtaidistance on identical sequences. Setting window=1 limits temporal drift off the main diagonal, while use_pruning=False turns off C-level early-stopping heuristics—guaranteeing exact distance values instead of inf while tracking the strict 1-to-1 warping path.

Source

import numpy as np from dtaidistance import dtw # Sample 1D sequences sequence_a = np.array([0.0, 1.0, 1.2, 1.5, 2.2, 10.1, 1.1, 1.1, 1.1, 1.1, 1.1, 1.1], dtype=np.float64) sequence_b = np.array([0.0, 1.0, 1.2, 1.5, 2.2, 10.1, 1.1, 1.1, 1.1, 1.1, 1.1, 1.1], dtype=np.float64) WINDOW_SIZE = 1 # Disable C-level pruning heuristics explicitly distance = dtw.distance( sequence_a, sequence_b, window=WINDOW_SIZE, use_pruning=False # Forces full computation without inf cutoff ) path = dtw.warping_path( sequence_a, sequence_b, window=WINDOW_SIZE, use_pruning=False ) print(f"DTW Distance: {distance:.4f}") print(f"Warping Path: {path}")

Output

DTW Distance: 0.0000
Warping Path: [(0, 0), (1, 1), (2, 2), (3, 3), (4, 4), (5, 5), (6, 6), (7, 7), (8, 8), (9, 9), (10, 10), (11, 11)]