Description
Please install the following packages (either with pip or conda):
- scikit-learn 0.19.0 (along with numpy and scipy as dependencies)
- pandas
- matplotlib
- jupyter
Check the scikit-learn version with:
python -c "import sklearn; print(sklearn.__version__)"
If you have an old version you can upgrade to scikit-learn 0.19.0 with conda update scikit-learn if you use conda or pip install -U scikit-learn otherwise.
We recommend using Python 3.6 but the tutorial material should also work with older Python versions (including Python 2.7, but please please consider using Python 3.6 instead).
In addition you can also install scikit-optimize (if you use a conda environment, you need to first conda install pip and then pip install scikit-optimize) in that environment. scikit-optimize has a dependency on scikit-garden which does not yet provide any binary package and can fail to install automatically from source on systems that lack a properly configured C/C++ compiler. If you have trouble installing scikit-garden and scikit-optimize, don't worry you will be able to follow those sections on the video projector while still being able to run the exercises of the other sections that only require scikit-learn.
The notebooks will be published on https://github.com/ogrisel/euroscipy_2017_sklearn (please update on Sunday evening or at the beginning of the tutorial assuming wifi is good enough).