Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.python.announce > #1407

ANN: Lea 1.3.1 released

From "Pierre Denis" <pie.denis@skynet.be>
Subject ANN: Lea 1.3.1 released
Date 2014-09-21 22:13 +0200
Newsgroups comp.lang.python.announce
Message-ID <mailman.14184.1411366906.18130.python-announce-list@python.org> (permalink)

Show all headers | View raw


Lea, discrete probability distributions in Python

=================================================

 

I have the pleasure to announce the release of Lea 1.3.1. 



NEW: Lea now runs on Python 3 (and still on Python 2.x) !

 

Lea is a Python package that allows you to define and play with discrete

probability distributions in an intuitive way. 

Lea can model a broad range of random discrete phenomenons. Then, it

allows calculating probabilities of events, whether atomic, aggregated or 
combined through operations. A typical example is the probabilities of the
sum of N dice having known, possibly unfair, probability distributions. 

Download (PyPi)

===============

 <http://pypi.python.org/pypi/lea> http://pypi.python.org/pypi/lea



Project page / documentation

============================

 <http://code.google.com/p/lea/> http://code.google.com/p/lea/

Features 
========
- models finite discrete probability distributions 
- standard distribution indicators (mean, standard deviation,.) 
- arithmetic and logical operators on probability distribution 
- cartesian products, conditional probabilities, joint distributions 
- generation of random samples 
- open-source project, LGPL license 

- runs on Python 2.x and 3.x
- pure Python module, lightweight - no package dependency 
- probabilities stored as rationals (no floating-point biases) 

Hoping Lea could be helpful in this uncertain universe... 

 

Pierre Denis

 

Back to comp.lang.python.announce | Previous | Next | Find similar | Unroll thread


Thread

ANN: Lea 1.3.1 released "Pierre Denis" <pie.denis@skynet.be> - 2014-09-21 22:13 +0200

csiph-web