wip/py-gvar 4.6.1 NetBSD

This package provides tools for creating and manipulating Gaussian random variables from correlated multi-dimensional Gaussian distributions. The Gaussian variables are represented by objects of type :class:`gvar.GVar` which can be combined with each other and with ordinary numbers in arbitrary arithmetic expressions, including standard functions such ``sin`` and ``exp``, to create new :class:`gvar.GVar`\s whose correlations with the original variables are tracked implicitly (and handled correctly).

This package relies on ``numpy`` for efficient array arithmetic, and on Cython to compile efficient core routines and interface code