ML-Ensemble is available through PyPI:

pip install -U mlens

Bleeding edge

For the bleeding edge version install via Github. You can optionally checkout pending pull requests before running the install command.

git clone; cd mlens;
python install


To install mlens the following dependencies are required:

Package Version Module
scipy >= 0.17 All
numpy >= 1.11 All

Additionally, to use the visualization module, the following libraries are necessary:

Package Version
matplotlib >= 1.5
seaborn >= 0.7

If you want to run examples, you also need:

Package Version
Scikit-learn >= 0.17

Test build

To test the installation, run:

cd mlens;

Note that this requires the Nose unit testing suite: if not found, the test script will automatically try to install it using pip install nose-exclude. The expected output should look like:

>>> python
Setting up tests... Ready.
Checking build... Build ok.

If the build fails, a log file will be created named check_build_log.txt that contains the traceback for the failed test for debugging.