Numpy

Note: The self-standing numpy module is obsolete and its use is discouraged due to potential conflicts with other python modules. If you need the numpy module, we recommend to load some version of Python - modules instead.

Description

Numpy is a Python language extension that defines the numerical array and matrix type and basic operations over them.

Availability

There are several versions of Numpy in Metacentrum.

• Numpy version 1.8.0: compiled with Intel CDK 14 (internally parallelized), which results in vastly improved computation performance
  • modules numpy-1.8.0-gcc-py2.6 and numpy-1.8.0-intel-py2.6

• Numpy version 1.7.1: compiled with Intel MKL 11.0 (internally parallelized), which results in vastly improved computation performance
  • modules numpy-1.7.1-py2.7
    • prepared to be used with Python 2.7 (module *-py2.7)
• Numpy version 1.6.2: compiled with Intel MKL 11.0 (internally parallelized), which results in vastly improved computation performance
  • modules numpy-1.6.2-py2.5 and numpy-1.6.2-py2.6
    • prepared to be used both with Python 2.5 (module *-py2.5) as well as Python 2.6 (module *-py2.6)
• Numpy version 1.6.0 beta 1: compiled with standard BLAS libraries
  • modules numpy-1.6.0b1-py2.5 and numpy-1.6.0b1-py2.6
    • prepared to be used both with Python 2.5 (module *-py2.5) as well as Python 2.6 (module *-py2.6)
• Numpy version 1.3.0
  • modules numpy-py2.4, numpy-py2.5 and numpy-py2.6

Use

Load module (module add 'module_name'), then you can import numpy libraries into Python (import numpy) or run the program f2py, by command f2py with neccessary parameters (see: man f2py).

Note: The modules version 1.6.0 do not automatically load the Python modules -- if you need them, load them manually.

Documentation

On-line documentation is available.

Supported platforms

Linux

Program administrator

meta@cesnet.cz

Homepage

URL: http://www.scipy.org/Getting_Started