numba 0.6.0

[mark florisson]

On behalf of the numba team I am pleased to announce a new version of
Numba, 0.6. The release includes faster numerical codes, better type
inference support, faster autojit dispatch, python 2.6 support and
more.

Download: http://pypi.python.org/pypi/numba/0.6.0
Documentation: http://numba.pydata.org/numba-doc/0.6/
Github: https://github.com/numba/numba

Numba will be part of the next anaconda CE release 1.3.1, to be released
tomorrow.

Numba
======
Numba is an just-in-time specializing compiler for Python and NumPy
code to LLVM for annotated functions (through decorators). It's goal
is to seamlessly integrate with the Python scientific software stack
and provide optimized native code and integration with native foreign
languages.

Dependencies:
============
* llvmpy 0.10.0
* meta   (optional)
* cython
* numpy
* LLVM 3.2 (3.1 might work but is not officially supported)

Release notes:
============
* Python 2.6 support
* Programmable typing
    * Allow users to add type inference for external code
* Better NumPy type inference
    * outer, inner, dot, vdot, tensordot, nonzero, where,
      binary ufuncs + methods (reduce, accumulate, reduceat, outer)
* Type based alias analysis
    * Support for strict aliasing
* Much faster autojit dispatch when calling from Python
* Faster numerical loops through data and stride pre-loading
* Integral overflow and underflow checking for conversions from objects
* Make Meta dependency optional

Many thanks to everyone who contributed to this release!

Dan Christensen
Ilan Schnell
Jon Riehl
Lars Buitinck
Mark Florisson
Phillip Cloud
Siu Kwan Lam
Travis E. Oliphant
Timo