Moss is a library of functions, classes, and scripts to that may be useful for analyzing scientific data. Because this package is developed for neuroimaging and cognitive science, there is probably some bias towards applications that are useful in that domain. However, the functions are intended to be written in as general and lightweight a fashion as possible.

• Python 2.7

• numpy

• scipy

• matplotlib

• pandas

• statsmodels

• scikit-learn

• six

To install the released version, just do

pip install -U moss

However, I update the code pretty frequently, so you may want to clone the github repository and install with

python setup.py install

You can exercise the full test suite by running nosetests in the source directory.

Note that some of the statistical tests depend on randomly generated data and fail from time to time because of this.

Moss is released under a BSD (3-clause) license.

http://github.com/mwaskom/moss

Please report any bugs you encounter on the Github issue tracker.