This package is to select the order of Makrov models for both smart home dataset and synthetic data.

The orders of Markov models ranges from 0 to 6th.

The measures to select orders are AIC, BIC, EDC, and HQIC.

The mc_class.py has the class MarkovModel with below properties.

m.alph is a list of unique activities. print("m.alph", m.alph)

m.tran is a dictionary that shows the transtion matrix (with count not probability)

e.g.: {(('adf', 'dafae', 'dafae'), 'dafae'): 1.0} print("m.tran", m.tran, len(m.tran))

m.kgrams is a dictionary that shows the count of each k successive activities

e.g.: {('abc', 'abc', 'abc'): 1} print("m.kgrams", m.kgrams)

m.freq shows the count of each k successive activities:

e.g.: ('abc', 'abc', 'abc'): 1 call from m.kgrams print("m.freq", m.freq(('dafae', 'adf', 'abc')))

m.freq2(('dafae', 'adf', 'abc'), 'abc') shows the count of the transition to 'abc'

call from m.tran print("m.freq2", m.freq2(('dafae', 'adf', 'abc'), 'abc'))