I'm curious to know what people recommend as the best way to handle probability distributions in Sage. Using scipy requires setting up a lambda function when doing a plot, and RealDistribution only accepts one parameter (standard deviation) for the normal distribution when there should be two. Is there a better option?
scipy.stats seems quite full-featured, although I've only used rv_discrete myself. You can define a new distribution as a subclass of rv_continuous or rv_discrete, and give it methods for the density, cumulative distribution, and many other things.
You should probably also look at R.
Asked: 2012-10-28 23:15:51 +0100
Seen: 1,226 times
Last updated: Oct 30 '12
Copyright Sage, 2010. Some rights reserved under creative commons license. Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
It is possible to use the normal with a different mean, by just moving all the data - in fact, I think we showed that in the PREP workshop at some point. Obviously that's not optimal, though - maybe you should open a ticket? Just sayin' :-)