Preparing a virtualenv¶
We recommend using a virtualenv, because (A) you can control the version of each python package, (B) installing and removing packages becomes easy as (C) you don’t need sudo rights and (D) it becomes harder to mess up up the python installation on your system.
pip install virtualenv virtualenv virtualSurrogates source virtualSurrogates/bin/activate
matplotlib, as this doesn’t work through setup.py.
easy_install -U distribute pip install numpy==1.8.1 pip install scipy==0.14.0 pip install matplotlib pip install scikit-learn==0.15.1
This may take some time. Afterwards you can verify having those libs installed with:
pip freeze argparse==1.2.1 mock==1.0.1 nose==1.3.4 numpy==1.8.1 pyparsing==2.0.3 python-dateutil==2.3 pytz==2014.10 scipy==0.14.0 six==1.8.0 wsgiref==0.1.2 scikit-learn==0.15.1 matplotlib==1.4.2
Install the Surrogate Benchmark Library¶
- Clone the repository:
git clone https://github.com/KEggensperger/SurrogateBenchmarks.git cd SurrogateBenchmarks
python setup.py install
This will install tools, scripts and some requirements (
python-daemon). This might take a while. When your environment is ready it could/should look like this:
pip freeze Surrogates==Nan argparse==1.2.1 decorator==3.4.0 lockfile==0.10.2 mock==1.0.1 networkx==1.9.1 numpy==1.8.1 pyparsing==2.0.3 python-daemon==1.6.1 python-dateutil==2.3 pytz==2014.10 scikit-learn==0.15.1 scipy==0.14.0 six==1.8.0 wsgiref==0.1.2
If the installation was successful you can run some tests. NOTE: Some tests will fail, if you are using different versions of numpy, scipy, and/or scikit-learn. This is not problematic as some of the tests only assert that you retrieve exactly the same results as me and as the numeric results only slightly differ.
python setup.py test
NOTE: If you cannot install the library, because you cannot upgrade scikit-learn, numpy, scipy, etc. Make sure some version of these modules is installed and uncomment the respective lines in install_requires.