PaCal / News: Recent posts

• Initial implementation of arithmetic of dependent random variables (Bayesian
  network style).
• Parallel computation (multicore). Set params.general.parallel=True to enable.

* beta version of arithmetic of dependent random variables (Bayesian network style)
* parts not related to dependent variables are as stable as in previous releases, so it is safe to use 1.5beta as default

* operations on two dependent random variables
* joint distribution models: normal, copulas
* joint distribution of two order statistics
* PaCal now depends on sympy

New in version 1.0:

* order statistics
* noncentral distributions: Chi^2, F, Beta and T
* optimized routines for sum, average, max, min etc. of i.i.d. random variables
* Extreme Value Distributions: Gumbel, Weibull, Frechet
* better accuracy for Beta distribution
* methods interp_error and interp_error_by_segment return estimates of
interpolation errors
* a warning is issued when dependent RVs are used in operations
(use params.general.warn_on_dependent = False to suppress)
* bug fixes