PyMC2 has some tricky tricks for reducing function evaluations if possible. A question asked and answered on Stack Overflow investigates: http://stackoverflow.com/q/27714635/1935494 and I made a IPython Notebook with more details, too: http://nbviewer.ipython.org/gist/aflaxman/c07c5261bf22f6847098

# Tag Archives: pymc

## A little PyMC2 trick

Here is a little trick for getting around a pesky initialization issue in PyMC2 models, asked and answers on Stack Overflow when thing were quiet around here: http://stackoverflow.com/a/27724637/1935494

Comments Off

Filed under software engineering

## PyMC3 with PyMC2

Did you know I have a fork of PyMC3 that you can run at the same time as PyMC2? I don’t keep it up to date, but people seem to want it every once in a while. Maybe this will help someone find it: https://github.com/aflaxman/pymc

import pymc as pm2 import pymc3 as pm3

Good for head-to-head comparisons…

Comments Off

Filed under software engineering

## Bayesian Correlation in PyMC

Here is a StackOverflow question with a nice figure:

Is there a nice, simple reference for just what exactly these graphical model figures mean? I want more of them.

Filed under statistics

## MCMC in Python: observed data for a sum of random variables in PyMC

I like answering PyMC questions on Stack Overflow, but sometimes I give an answer and end up the one with the question. Like what would you model as the sum of a Poisson and a Negative Binomial?

Comments Off

Filed under statistics

## MCMC in Python: sim and fit with same model

Here is a github issue and solution that I saw the other day. I think it’s a nice pattern.

def generate_model(values={'mu': true_param, 'm': None}): #prior mu = pymc.Uniform("mu", lower=-10, upper=10, value=values['mu'], observed=(values['mu'] is not None)) # likelihood function m = pymc.Normal("m", mu=mu, tau=tau, value=values['m'], observed=(values['m'] is not None)) return locals()

Comments Off

Filed under statistics

## MCMC in Python: Fit a non-linear function with PyMC

Here is a recent q&a on stack overflow that I did and liked.

Comments Off

Filed under statistics