Skip to main content

A tool to fit data to many distributions and best one(s)

Project description

https://badge.fury.io/py/fitter.svg https://pypip.in/d/fitter/badge.png https://secure.travis-ci.org/cokelaer/fitter.png https://landscape.io/github/cokelaer/fitter/master/landscape.png https://badge.waffle.io/cokelaer/fitter.png?label=ready&title=Ready

fitter package provides a simple class to figure out from whih distribution your data comes from. It uses scipy package to try 80 ditribution and allows you to plot the results to check what is the most probable distribution.

Installation

pip install fitter

Usage

Nothing complicated since there is just one class provided. First, we will need to create some data samples. Let us create a sequence of 100000 samples from a gamma distribution:

from scipy import stats
stats.gamma.rvs(2, loc=1.5, scale=2, size=100000)

Now, the question without any knowledge about the distribution of its parameter, what is a probable distribution that fit the data? scipy has 80 distribution with a method called fit that will help us here. The class Fitter will scan all the distribution, call the fit function for you, ignoring those that fail or run forever and finally give you a summary of the best distribution in the sense of sum of the square errors. The best is to give an example:

from fitter import Fitter
f = Fitter(data)
f.fit()
# make take some time since by default, all distribution are tried
f.summary()
http://pythonhosted.org/fitter/_images/index-1.png

See the online documentation for details.

Project details


Download files

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

Source Distribution

fitter-0.2.0.tar.gz (6.0 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page