v1.4.0

• pip installation

v1.3.5

• Fixed recommendation example
• Fixed SNMF
• Minor bugs

v1.3.2

Merge pull request #36 from haesemeyer/haesemeyer-patch-1

Update snmf.py to fix issue #35

v1.3.1

• Nimfa is now released under the 3-clause BSD license

v1.3

New algorithms for separable nonnegative matrix factorization.

1.2.2

v1.2.2

1.2.2

1.2.1

Minor improvements to the changed API. Algorithms used for matrix factorization are now accessible as Python objects rather than string parameters.

Enhanced API.

The API has been changed. Algorithms used for matrix factorization are now accessible as Python objects rather than string parameters. This makes Nimfa more intuitive and easier to use, especially in environments such as IPython.

Python 3 compatibility.

v1.1

Re-added normalization to Medulloblastoma (needed in examples)