How to write and read HMM in GHMM
This article shows how to write and read Hidden Markov Model in GHMM.
It is very easy.
Setup
from ghmm import * A = [[0.6, 0.4, 0], [0, 0.6, 0.4], [0.0, 0.0, 1.0]] B = [[0.25, 0.25, 0.25, 0.25], [0.25, 0.25, 0.25, 0.25], [0.25, 0.25, 0.25, 0.25]] pi = [1.0, 0, 0] sigma = IntegerRange(0, 4) hmm = HMMFromMatrices(sigma, DiscreteDistribution(sigma), A, B, pi) hmm.normalize() train_seq = EmissionSequence(sigma, [0, 0, 0, 1, 3, 2, 2, 3]) hmm.baumWelch(train_seq)
Write HMM to a file
filename = 'trained_hmm'
hmm.write(filename)
Read HMM from a file
from ghmm import * filename = 'trained_hmm' hmm = HMMOpenXML.openNewXML(filename, None) #can be use hmm normally test_seq = EmissionSequence(sigma, [0, 0, 0, 1, 3, 2, 2]) result = hmm.viterbi(test_seq)