skip to navigation
skip to content

Not Logged In

pypeaks 0.2.7

Python module with different methods to identify peaks from data like histograms and time-series data

Identifying peaks from data is one of the most common tasks in many research and development tasks. pypeaks is a python module to detect peaks from any data like histograms and time-series.

Following are the available methods implemented in this module for peak detection: * Slope based method, where peaks are located based on how the data varies. * Intervals based method, where a set of intervals can be passed to provide apriori information that there will be at most one peak in each interval, and we just pick the maximum in each interval, filtering out irrelevant peaks at the end. * A hybrid method which combines these two methods.


$ sudo pip install --upgrade pypeaks


There is an example case included along with the code. If you don’t have this folder, please load your data instead. Or get it from

Important note

The peak finding function expects a normalized smoothed histogram. It does smoothing by default. If you want to change the smoothness, customize the corresponding argument. If the data is not normalized (so that the area under the curve comes to 1), there is a function provided to do that. If you don’t get any peaks, then you probably overlooked this!

import pickle
from pypeaks import Data, Intervals

[x, y] = pickle.load(file('examples/sample-histogram.pickle'))
data_obj = Data(x, y, smoothness=11)

#Peaks by slope method
#print data_obj.peaks

#Peaks by interval method
ji_intervals = pickle.load('examples/ji_intervals.pickle')
ji_intervals = Intervals(ji_intervals)
data_obj.get_peaks(method='interval', intervals=ji_intervals)
#print data_obj.peaks

#Read the help on Data object, and everything else is explained there.

In case you face some issue, report it on github, or write to me at gopala [dot] koduri [at] gmail [dot] com!

File Type Py Version Uploaded on Size
pypeaks-0.2.7.tar.gz (md5) Source 2014-06-06 117KB
  • Downloads (All Versions):
  • 27 downloads in the last day
  • 109 downloads in the last week
  • 600 downloads in the last month