https://github.com/pollen-robotics/dtw

DTW (Dynamic Time Warping) python module
https://github.com/pollen-robotics/dtw

distance distance-measures distance-metric dtw python

Last synced: 29 days ago
JSON representation

DTW (Dynamic Time Warping) python module

• Host: GitHub
• URL: https://github.com/pollen-robotics/dtw
• Owner: pollen-robotics
• License: gpl-3.0
• Created: 2014-07-28T10:01:39.000Z (over 10 years ago)
• Default Branch: master
• Last Pushed: 2024-03-29T11:30:12.000Z (8 months ago)
• Last Synced: 2024-10-02T00:01:28.428Z (about 1 month ago)
• Topics: distance, distance-measures, distance-metric, dtw, python
• Language: Python
• Size: 191 KB
• Stars: 1,156
• Watchers: 31
• Forks: 233
• Open Issues: 16
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# [Dynamic Time Warping](https://en.wikipedia.org/wiki/Dynamic_time_warping) Python Module

[![Build Status](https://travis-ci.org/pierre-rouanet/dtw.svg?branch=master)](https://travis-ci.org/pierre-rouanet/dtw)

Dynamic time warping is used as a similarity measured between temporal sequences. This package provides two implementations:

* the basic version (see [here](https://en.wikipedia.org/wiki/Dynamic_time_warping)) for the algorithm

* an accelerated version which relies on scipy cdist (see https://github.com/pierre-rouanet/dtw/pull/8 for detail)

```python

import numpy as np

# We define two sequences x, y as numpy array

# where y is actually a sub-sequence from x

x = np.array([2, 0, 1, 1, 2, 4, 2, 1, 2, 0]).reshape(-1, 1)

y = np.array([1, 1, 2, 4, 2, 1, 2, 0]).reshape(-1, 1)

from dtw import dtw

manhattan_distance = lambda x, y: np.abs(x - y)

d, cost_matrix, acc_cost_matrix, path = dtw(x, y, dist=manhattan_distance)

print(d)

>>> 2.0 # Only the cost for the insertions is kept

# You can also visualise the accumulated cost and the shortest path

import matplotlib.pyplot as plt

plt.imshow(acc_cost_matrix.T, origin='lower', cmap='gray', interpolation='nearest')

plt.plot(path[0], path[1], 'w')

plt.show()

```

Result of the accumulated cost matrix and the shortest path (in white) found:

![Acc cost matrix and shortest path](./acc.png)

## Other examples are available as notebook

* [the code above as a notebook](./examples/simple%20example.ipynb)

* [a sound comparison based on DTW + MFCC](./examples/MFCC%20%2B%20DTW.ipynb)

* [simple speech recognition](./examples/speech-recognition.ipynb)

## Installation

```

python -m pip install dtw

```

It is tested on Python 2.7, 3.4, 3.5 and 3.6. It requires numpy and scipy.