Skip to main content

Peak detection using AMPD and ASS-AMPD algorithms

Project description

Build Status Coverage Status

AMPD algorithm in Python

Implements a function find_peaks based on the Automatic Multi-scale Peak Detection algorithm proposed by Felix Scholkmann et al. in "An Efficient Algorithm for Automatic Peak Detection in Noisy Periodic and Quasi-Periodic Signals", Algorithms 2012, 5, 588-603

Peak finding

Usage

Install from PyPI:

pip install pyampd

Or install from source:

pip install git+https://github.com/ig248/pyampd

Import function:

from pyampd.ampd import find_peaks

See notebooks/ampd.ipynb for usage examples.

Specifying maximum scale

To improve run-time on large time-series, it is possible to specify the maximum scale to consider:

peaks = find_peaks(x, scale=100)

will only consider windows up to +-100 point either side of peak candidates.

Adaptive Scale Selection

If the characteristic scale of the signal changes over time, a new algorithm called Adaptive Scale Selection can track the changes in optimal scales and detect peaks accordingly:

peaks = find_peaks_adaptive(x, window=200)

will select the optimal scale at each point using a 200-point running window.

Original implementation

find_peaks is not identical to the algorithm proposed in the original paper (especially near start and end of time series). A performance-optimized version of the original implementation is provided in find_peaks_original.

Tests

Run

pytest

Other implementations

Improvements

This Python implementation provides significant speed-ups in two areas:

  1. Efficient tracking of local minima without using random numbers
  2. Introduction of maximum window size, reducing algorithm run-time from quadratic to linear in the number of samples.
  3. Better handling of peaks near start/end of the series
  4. Addition of new Adaptive Scale Selection

ToDo

  • It may be possible to avoid repeated comparisons, and reduce worst-case runtime from O(n^2) to O(n log(n)).
  • find_peaks_adaptive could benefit from specifying both window and max_scale

References

Original paper: https://doi.org/10.1109/ICRERA.2016.7884365

Project details


Release history Release notifications

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for pyampd, version 0.0.1
Filename, size File type Python version Upload date Hashes
Filename, size pyampd-0.0.1-py2.py3-none-any.whl (5.1 kB) File type Wheel Python version py2.py3 Upload date Hashes View
Filename, size pyampd-0.0.1.tar.gz (4.6 kB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page