The python implementation of Apriori Algorithm.
Frequently used for association rule mining. e.g. People like coke, also like
lime
$ sudo python setup.py install
>>> from yapa.apriori import NaiveApriori
>>> # Instantiate a apriori with universal set, and a bunch of parameters
>>> apriori = NaiveApriori(universal_set = set(range(1,10)),
support_criterion=0.2,
confident_criterion=0.7,
maximum_cardinality=4)
>>> # Prepare the sample sets, AKA, training sets.
>>> data_sets = (set(1,2,5), set(2,3,4), set(1,2,3), set(2,10))
>>> # Generate the rules
>>> apriori.generate_rules(data_sets)
>>> # Predict a associated element based on an input
>>> for result, confident in apriori.predict([0,1]):
print set([0,1]),"->", result, confident
set([1,2])->set([4]), 0.75
Please reference the docstrings
$ nosetests test
or
$ python -m unittest test
See LICENSE