Abstraction point allowing use with numpy or Numeric
 
Chooses numpy if available because when it's installed
Numeric tends to be a bit flaky...

Modules
		numpy.add_newdocs numpy.core.defchararray numpy.ctypeslib numpy.lib.scimath numpy.fft vrml_accelerate.frustcullaccelnumpy numpy.linalg numpy.ma math numpy.random numpy.core.records vrml_accelerate.tmatrixaccelnumpy

Functions
		contiguous(a) Force to a contiguous array safeCompare(first, second) Watch out for pointless numpy truth-value checks typeCode(a) Retrieve the typecode for the given array   Depending on whether you access the classic or new API you have different access methods, so we have to use the typecode() method if .dtype.char isn't there.

Data
		__file__ = '/home/mcfletch/pylive/vrml/arrays.pyc' __name__ = 'vrml.arrays' __package__ = 'vrml' divide_safe = <ufunc 'divide'> implementation_name = 'numpy'