Installation

System requirements

auto-sklearn has the following system requirements:

  • Linux operating system (for example Ubuntu),
  • Python (>=3.4).
  • C++ compiler (with C++11 supports) and SWIG (version 3.0 or later)

For an explanation of missing Microsoft Windows and MAC OSX support please check the Section Windows/OSX compability.

Installing auto-sklearn

Please install all dependencies manually with:

curl https://raw.githubusercontent.com/automl/auto-sklearn/master/requirements.txt | xargs -n 1 -L 1 pip install

Then install auto-sklearn:

pip install auto-sklearn

We recommend installing auto-sklearn into a virtual environment or an Anaconda environment.

Ubuntu installation

To provide a C++11 building environment and the lateste SWIG version on Ubuntu, run:

sudo apt-get install build-essential swig

Anaconda installation

Anaconda does not ship auto-sklearn, and there are no conda packages for auto-sklearn. Thus, it is easiest to install auto-sklearn as detailed in the Section Installing auto-sklearn.

A common installation problem under recent Linux distribution is the incompability of the compiler version used to compile the Python binary shipped by AnaConda and the compiler installed by the distribution. This can be solved by installing the gcc compiler shipped with AnaConda (as well as swig):

conda install gcc swig

Windows/OSX compability

Windows

auto-sklearn relies heavily on the Python module resource. resource is part of Python’s Unix Specific Services and not available on a Windows machine. Therefore, it is not possible to run auto-sklearn on a Windows machine.

Possible solutions (not tested):

  • Windows 10 bash shell
  • virtual machine
  • docker image

Mac OSX

Auto-sklearn is known to work on OSX systems. Nevertheless, there are two issues holding us back from actively supporting OSX:

  • The resource module cannot enforce a memory limit on a Python process (see SMAC3/issues/115).
  • OSX machines on travis-ci take more than 30 minutes to spawn. This makes it impossible for us to run unit tests for auto-sklearn and its dependencies SMAC3 and ConfigSpace.

Possible solutions (not tested):

  • virtual machine
  • docker image