NumPy adds a fast and sophisticated array facility to the Python language and is the fundamental package needed for scientific computing with Python.
The NumPy package contains:
· a powerful N-dimensional array object
· sophisticated (broadcasting) functions
· basic linear algebra functions
· basic Fourier transforms
· sophisticated random number capabilities
· tools for integrating Fortran code.
· tools for integrating C/C++ code.
Users of num array can also transition their code using the numpy.numarray compatibility layer and a similar module (numpy.numarray.alter_code1).
NOTE: NumPy is licensed and distributed under the terms of an OSI-Approved Open Source License.
· Python 2.5 or later
What's New in This Release: [ read full changelog ]
· Fix `1` is printed during numpy.test()
· Backport memory leak fix
· Backport fix complex polynomial fit
· Make nansum work with booleans.
· Backport large sort fixes
· Backport object take
· Backport nditer fix op axes initialization
· npy-pkg-config ini files were missing after Bento build.
· PyArray_LexSort allocates too much temporary memory.
· Exported f2py_size symbol prevents linking multiple f2py modules.
· Add mention of PyArray_SetBaseObject stealing a reference
· Fix typo in fft docs (the indexing variable is 'm', not 'n').