A side by side comparison of hmmlearn and pomegranate HMMs on a chaotic dataset

hmm_hmmlearn.py

#run in python 2.7
from hmmlearn.hmm import GaussianHMM
data=np.fromfile("np_input_file.txt").reshape([200001,6])

#MODEL PARAMS
K=3 
iter_num=3000
convergence_tolerance=0.001
stat_vec_tol=0.999
 
#FIT MODEL
model= GaussianHMM(n_components=K, covariance_type="full", n_iter=iter_num,tol=convergence_tolerance).fit(data)
hidden_states=model.predict(data) #encodes a time series of hidden states
means=model.means_

#PLOT MODEL
plt.plot(data.T[0],data.T[3],c='k',lw=0.005) #the data
for i in range(0,3): #the model means
    plt.scatter(model.states[i].distribution.mu[0],model.states[i].distribution.mu[3],c='r',label="HMM clusters upon convergence")

hmm_pomegranate.py

#run in python 3.6
from pomegranate import *
from sklearn.cluster import KMeans
import matplotlib.pyplot as plt

#IMPORT DATA AND RUN KMEANS
data=np.fromfile("np_input_file.txt").reshape([1,200001,6])
cov = numpy.eye(6)
kmodel=KMeans(3).fit(data[0,:,:])
mu=kmodel.cluster_centers_

#BUILD MODEL
s1=State(MultivariateGaussianDistribution(mu[0],cov))
s2=State(MultivariateGaussianDistribution(mu[1],cov))
s3=State(MultivariateGaussianDistribution(mu[2],cov))
model=HiddenMarkovModel()
model.add_states(s1, s2, s3)
model.add_transition(model.start, s1, 0.33)
model.add_transition(model.start, s2, 0.33)
model.add_transition(model.start, s3, 0.34)

model.add_transition(s1, s1, 0.33)
model.add_transition(s1, s2, 0.33)
model.add_transition(s1, s3, 0.34)

model.add_transition(s2, s1, 0.33)
model.add_transition(s2, s2, 0.33)
model.add_transition(s2, s3, 0.34)

model.add_transition(s3, s1, 0.33)
model.add_transition(s3, s2, 0.33)
model.add_transition(s3, s3, 0.34)

model.bake()

data=data.reshape([200001,6])
#FIT MODEL
model.fit(data, verbose=True)

##PLOT MODEL
plt.plot(data.T[0], data.T[3],c='k',lw=0.005) #the data
for i in range(0,3): #the model means
    plt.scatter(model.states[i].distribution.mu[0],model.states[i].distribution.mu[3],c='r',label="HMM clusters upon convergence")
    plt.scatter(kmodel.cluster_centers_[i][0],kmodel.cluster_centers_[i][3],c='g',zorder=30,label="k_means initial clusters")