How to Install the Python* Version of Intel® Data Analytics Acceleration Library (Intel® DAAL)

Introduction

Intel® Data Analytics Acceleration Library (Intel® DAAL) is a software solution that offers building blocks covering all the stages of data analytics, from preprocessing to decision making. The beta version of Intel DAAL 2017 provides support for the Python* language.

Installation

This article gives you three options to install the Python* version of Intel DAAL, or pyDAAL.

• Anaconda* distribution on Linux*

  The easiest way to install pyDAAL on Linux* is to get it through the Anaconda* distribution. For details, review this step-by-step instruction.

• Intel® Distribution for Python* on Windows*, Linux*, and OS X*

  pyDAAL is also shipped with the Intel® Distribution for Python*. For more information, refer to the product page.

• Building pyDAAL from the source code

  If the options above do not work for you, it is possible to build pyDAAL yourself. The sources are provided in the beta package of Intel DAAL 2017.

  1. Set the DAAL environment variables by running the <DAAL install dir>/daal/bin/daalvars.sh script as appropriate for your target architecture:
     • IA-32 architecture: daalvars.sh ia32
     • Intel® 64 architecture: daalvars.sh intel64
  2. Go to the directory with the Python* sources of Intel DAAL and run the installation script:

     cd <DAAL install dir>/pydaal_sources
<python home>/python setup.py install

     The setup.py script builds and installs the pyDAAL package to use Intel DAAL in Python* programs. For details, refer to the Intel DAAL getting started guide available in <DAAL install dir>/documentation/en/ps2017.

Conclusion

The three options above are the most common ways of installing the Python* version of Intel DAAL on your system. Let us know which one you like the most. If you have another preferred method not covered in this article, we would also love to hear about it.