[SciPy-user] Potential SoC project: black-box non-uniform random number generation

[Robert Kern]

A good number of the probability distributions in scipy.stats don't have fast
random samplers. The default implementation of the .rvs() method uses the
expensive and frequently inaccurate inversion method. However, there are
"universal" sampling algorithms that operate on very broad classes of
probability distributions.

  http://statistik.wu-wien.ac.at/projects/arvag/index.html

I think that implementing several of these algorithms and integrating them into
scipy.stats would make a tidy SoC project for a student.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco