LCM: Linear time Closed itemset Miner

TODO

• Refactoring lcm and publish on github. Now, lcm install method is too old.
• Use Cython to call lcm. Maybe Cython is most useful to call a module implemented by c language.
• version control and publish to PyPI

Install

To install LCM ver. 5.3., run install/install_lcm.sh. If you need, set INSTALL_PATH in shell script. ref. http://research.nii.ac.jp/~uno/codes-j.htm

Install as module

pip install git+ssh://git@github.com/YuseiYokoyama/python-lcm.git
#pip install git+ssh://git@github.com/YuseiYokoyama/python-lcm.git@master
#pip install git+ssh://git@github.com/YuseiYokoyama/python-lcm.git@v0.1.0

Uninstall

pip uninstall lcm

Related

fp grouth is also frequent itemset mining algorithm.

• provided a python implemented module in mlxtend. https://github.com/rasbt/mlxtend/blob/2362d9f53cb715a88cda63a8a7e69793afd64cbd/mlxtend/frequent_patterns/fpgrowth.py
• also output not closed pattern

So, LCM is faster.