Install Intel Python 3.6

Intel Distribution for Python utilizes the conda package manager, set to the intel channel.

Pros:

  • Intel Python uses highly accelerated math library versions of popular packages like Numpy, Scikit-learn, etc. reside.

Cons:

  • Intel Python is slow to update to new library versions. Anaconda updates within a couple days, but Intel might take weeks or months to update the libraries.

Note that standard Anaconda/Miniconda already gives MKL performance in some libraries like Numpy and Scipy. I still use the regular Anaconda/Miniconda (instead of Intel Python) to get updated libraries with most of them having MKL.

Install

  1. Download Stand-Alone Intel Python 3.6
  2. Extract and Install as User with
    ./install_GUI.sh
    

    which installs to ~/intel/intelpython3/bin

  3. Add to ~/.bashrc
    export PATH="$HOME/intel/intelpython3/bin:$PATH"
    
  4. Open a new Terminal and enjoy accelerated Intel Python 3.6

Spyder IDE

As usual with conda

conda install spyder

Fix broken history up arrow

If the up arrow for Python command history is broken, instead giving ^[[A, this can be fixed by

  1. install rlwrap
    apt install rlwrap
    
  2. add to ~/.bash_aliases
    alias python="rlwrap python"
    

That won’t bother other Pythons like system Python /usr/bin/python.

Tags:

Categories:

Written by Michael Hirsch, Ph.D. //

Comments