def test_fit(self, params='ste', n_iter=15, verbose=False, **kwargs):
np.random.seed(0)
h = hmm.MultinomialHMM(self.n_components,
startprob=self.startprob,
transmat=self.transmat)
h.emissionprob = self.emissionprob
# Create training data by sampling from the HMM.
train_obs = [h.rvs(n=10) for x in xrange(10)]
# Mess up the parameters and see if we can re-learn them.
h.startprob = hmm.normalize(self.prng.rand(self.n_components))
h.transmat = hmm.normalize(self.prng.rand(self.n_components,
self.n_components),
axis=1)
h.emissionprob = hmm.normalize(self.prng.rand(self.n_components,
self.n_symbols),
axis=1)
trainll = train_hmm_and_keep_track_of_log_likelihood(h,
train_obs,
n_iter=n_iter,
params=params,
**kwargs)[1:]
if not np.all(np.diff(trainll) > 0) and verbose:
print
print 'Test train: (%s)\n %s\n %s' % (params, trainll,
np.diff(trainll))
self.assertTrue(np.all(np.diff(trainll) > 0))
def __call__(self):
hmmdata = dict()
for (name, tracks, probs, finfo) in \
self.dtable.iterby(self.ecopts.sortby, True):
if tracks is probs is finfo is None:
hmmdata[name] = None
continue
est = self._get_estimator(probs, tracks)
est.constrain(self.hmmc(est))
# ugly sklearn
hmm_ = hmm.MultinomialHMM(n_components=est.nstates)
hmm_._set_startprob(est.startprob)
hmm_._set_transmat(est.trans)
hmm_._set_emissionprob(est.emis)
# trackwise error correction
tracks2 = []
for track in self.classdef.label2index(tracks):
tracks2.append(hmm_.predict(track))
tracks2 = self.classdef.index2labels(np.array(tracks2, dtype=int))
bucket = HmmBucket(tracks,
tracks2,
est.startprob,
est.emis,
est.trans,
self.dtable.groups(self.ecopts.sortby, name),
tracks.shape[0],
self.ecopts.timelapse, finfo)
hmmdata[name] = bucket
return hmmdata
def test_attributes(self):
h = hmm.MultinomialHMM(self.n_components)
self.assertEquals(h.n_components, self.n_components)
h.startprob = self.startprob
assert_array_almost_equal(h.startprob, self.startprob)
self.assertRaises(ValueError, h._set_startprob, 2 * self.startprob)
self.assertRaises(ValueError, h._set_startprob, [])
self.assertRaises(ValueError, h._set_startprob,
np.zeros((self.n_components - 2, self.n_symbols)))
h._set_transmat(self.transmat)
assert_array_almost_equal(h._get_transmat(), self.transmat)
self.assertRaises(ValueError, h._set_transmat, 2 * self.transmat)
self.assertRaises(ValueError, h._set_transmat, [])
self.assertRaises(ValueError, h._set_transmat,
np.zeros((self.n_components - 2, self.n_components)))
h._set_emissionprob(self.emissionprob)
assert_array_almost_equal(h._get_emissionprob(), self.emissionprob)
self.assertRaises(ValueError, h._set_emissionprob, [])
self.assertRaises(ValueError, h._set_emissionprob,
np.zeros((self.n_components - 2, self.n_symbols)))
self.assertEquals(h.n_symbols, self.n_symbols)
def test_attributes(self):
h = hmm.MultinomialHMM(self.n_components)
self.assertEquals(h.n_components, self.n_components)
h.startprob_ = self.startprob
assert_array_almost_equal(h.startprob_, self.startprob)
self.assertRaises(ValueError, h.__setattr__, 'startprob_',
2 * self.startprob)
self.assertRaises(ValueError, h.__setattr__, 'startprob_', [])
self.assertRaises(ValueError, h.__setattr__, 'startprob_',
np.zeros((self.n_components - 2, self.n_symbols)))
h.transmat_ = self.transmat
assert_array_almost_equal(h.transmat_, self.transmat)
self.assertRaises(ValueError, h.__setattr__, 'transmat_',
2 * self.transmat)
self.assertRaises(ValueError, h.__setattr__, 'transmat_', [])
self.assertRaises(ValueError, h.__setattr__, 'transmat_',
np.zeros((self.n_components - 2, self.n_components)))
h.emissionprob_ = self.emissionprob
assert_array_almost_equal(h.emissionprob_, self.emissionprob)
self.assertRaises(ValueError, h.__setattr__, 'emissionprob_', [])
self.assertRaises(ValueError, h.__setattr__, 'emissionprob_',
np.zeros((self.n_components - 2, self.n_symbols)))
self.assertEquals(h.n_symbols, self.n_symbols)
def train(trees):
corpus = [list(t.tokens) for t in trees]
words = list(sorted(set([t.word for c in corpus for t in c])))
tags = list(sorted(set([t.tag for c in corpus for t in c])))
word_index = _index_positions(words)
tag_index = _index_positions(tags)
transition = np.zeros((len(tags), len(tags)))
start = np.zeros((len(tags),))
emission = np.zeros((len(tags), len(words)+1))
for s in corpus:
for i,t in enumerate(s):
tag_ix = tag_index[t.tag]
word_ix = word_index[t.word]
emission[tag_ix, word_ix] += 1
if i == 0:
start[tag_ix] += 1
else:
transition[tag_index[s[i-1].tag], tag_ix] += 1
start /= start.sum()
normalize_rows = lambda x: (x.T / x.sum(axis=1)).T
transition = normalize_rows(transition+1)
emission = normalize_rows(emission+1)
model = hmm.MultinomialHMM(n_components=len(tags),
transmat=transition,
startprob=start)
model.emissionprob_ = emission
return HMMTagger(words, tags, model)
def test_fit_with_init(self,
params='ste',
n_iter=5,
verbose=False,
**kwargs):
h = self.h
learner = hmm.MultinomialHMM(self.n_components)
# Create training data by sampling from the HMM.
train_obs = [h.sample(n=10)[0] for x in xrange(10)]
# use init_function to initialize paramerters
learner._init(train_obs, params)
trainll = train_hmm_and_keep_track_of_log_likelihood(learner,
train_obs,
n_iter=n_iter,
params=params,
**kwargs)[1:]
# Check that the loglik is always increasing during training
if not np.all(np.diff(trainll) > 0) and verbose:
print
print 'Test train: (%s)\n %s\n %s' % (params, trainll,
np.diff(trainll))
self.assertTrue(np.all(np.diff(trainll) > -1.e-3))
def test_decode_map_algorithm(self):
observations = [0, 1, 2]
h = hmm.MultinomialHMM(self.n_components, startprob=self.startprob,
transmat=self.transmat, algorithm="map",)
h.emissionprob_ = self.emissionprob
logprob, state_sequence = h.decode(observations)
assert_array_equal(state_sequence, [1, 0, 0])
def test_rvs(self, n=1000):
h = hmm.MultinomialHMM(self.n_components,
startprob=self.startprob,
transmat=self.transmat)
h.emissionprob = self.emissionprob
samples = h.rvs(n)
self.assertEquals(len(samples), n)
self.assertEquals(len(np.unique(samples)), self.n_symbols)
def learn(obs, n_states=10):
i2t, t2i = token_features(obs)
tokens_i = list(map(t2i, obs))
model = hmm.MultinomialHMM(n_states, n_iter=20)
model.fit(tokens_i)
return model, i2t, t2i
def setUp(self):
self.prng = np.random.RandomState(9)
self.n_components = 2 # ('Rainy', 'Sunny')
self.n_symbols = 3 # ('walk', 'shop', 'clean')
self.emissionprob = [[0.1, 0.4, 0.5], [0.6, 0.3, 0.1]]
self.startprob = [0.6, 0.4]
self.transmat = [[0.7, 0.3], [0.4, 0.6]]
self.h = hmm.MultinomialHMM(self.n_components,
startprob=self.startprob,
transmat=self.transmat)
self.h.emissionprob_ = self.emissionprob
def test_eval(self):
h = hmm.MultinomialHMM(self.n_components,
startprob=self.startprob,
transmat=self.transmat)
h.emissionprob = self.emissionprob
idx = np.repeat(range(self.n_components), 10)
nobs = len(idx)
obs = [int(x) for x in np.floor(self.prng.rand(nobs) * self.n_symbols)]
ll, posteriors = h.eval(obs)
self.assertEqual(posteriors.shape, (nobs, self.n_components))
assert_array_almost_equal(posteriors.sum(axis=1), np.ones(nobs))
def f(marks,prb):
if len(marks)!=1:
print "Error"
else:
#states = ["A+","A","A-", "B+","B","B-","C+","C","C-", "F"]
states = ["A","B","C", "F"]
n_states = len(states)
observations = ["Sub1:A+","Sub1:A","Sub1:A-", "Sub1:B+", "Sub1:B", "Sub1:B-", "Sub1:C+", "Sub1:C", "Sub1:C_","Sub1:F"]
n_observations = len(observations)
#Starting proberbilities
start_probability = np.array([0.25, 0.25,0.25, 0.25])
#Transision proberbilities
transition_probability = np.array([[0.4, 0.3,0.2, 0.1],[0.3, 0.4,0.2, 0.1],[0.1, 0.3,0.4, 0.2],[0.1, 0.2,0.3, 0.4]])
# Emission proberbilities are read from file
emission_probability = np.array(prb)
model = hmm.MultinomialHMM(n_components=n_states)
model._set_startprob(start_probability)
model._set_transmat(transition_probability)
model._set_emissionprob(emission_probability)
bob_says = marks
#bob_says=[0]
logprob, pred = model.decode(bob_says, algorithm="viterbi")
# Display the result
print "Previous Exam Result:", ", ".join(map(lambda x: observations[x], bob_says))
print "Predicted Software Engineering Result:", ", ".join(map(lambda x: states[x], pred))
'''
If C and F values are not get as a predicted then student is displayed as approved
'''
if (pred[0]!=3)&(pred[0]!=2):
con="Student Approved"
else:con="Not Sufficient Requirement"
# Create GUI
window = Tk()
window.configure(background='#8A002E')
window.title('Margov Prediction')
window.geometry('480x200') # Size 200, 200
frame = Frame(window,background='#8A002E')
frame.pack()
def back():
window.withdraw()
Label(frame, background='#8A002E',font=("Helvetica", 14),foreground="white",text="Student's predicted success:").grid(row=0, column=0)
Label(frame, background='#8A002E',font=("Helvetica", 14),foreground="white",text=map(lambda x: states[x], pred)).grid(row=0, column=1)
Label(frame,background='#8A002E', text=con,fg="red",font=("Helvetica", 20)).grid(row=1, column=0)
Button(frame, text='Next', command=back).grid(row=2, column=2, sticky=W, pady=4)
return pred[0]
def test_wikipedia_viterbi_example(self):
# From http://en.wikipedia.org/wiki/Viterbi_algorithm:
# "This reveals that the observations ['walk', 'shop', 'clean']
# were most likely generated by states ['Sunny', 'Rainy',
# 'Rainy'], with probability 0.01344."
observations = [0, 1, 2]
h = hmm.MultinomialHMM(self.n_components,
startprob=self.startprob,
transmat=self.transmat)
h.emissionprob = self.emissionprob
logprob, state_sequence = h.decode(observations)
self.assertAlmostEqual(np.exp(logprob), 0.01344)
assert_array_equal(state_sequence, [1, 0, 0])
def absolute_train(self, data):
#save the data so that we may use it some other time (e.g. visualization)
self.data = data
#gather the codebook from VQ.py module
self.codebook = VQModel.get_codebook(data)
#quantize the data according to the codebook
quantized_value = quantize(self.codebook, data)
#Now train the HMM
#WARNING: The number of components have to be researched.
self.hmm = hmm.MultinomialHMM(n_components=5)
self.hmm.fit([quantized_value])
return
def f(marks):
if len(marks) != 4:
print "Error"
else:
states = ["pass", "fail"]
n_states = len(states)
observations = [
"OOP CA Passed", "OOP CA Failed", "OOP Exam Passed",
"OOP Exam Failed", "OOSD CA Passed", "OOSD CA Failed",
"OOSD Exam Passed", "OOSD Exam Failed"
]
n_observations = len(observations)
start_probability = np.array([0.5, 0.5])
transition_probability = np.array([[0.7, 0.3], [0.3, 0.7]])
emission_probability = np.array(
[[0.0, 0.0, 0.1, 0.1, 0.1, 0.1, 0.4, 0.2],
[0.0, 0.05, 0.1, 0.2, 0.1, 0.1, 0.05, 0.4]])
model = hmm.MultinomialHMM(n_components=n_states)
model._set_startprob(start_probability)
model._set_transmat(transition_probability)
model._set_emissionprob(emission_probability)
bob_says = marks
logprob, pred = model.decode(bob_says, algorithm="viterbi")
prob = 0
if pred[0] == 0:
prob = prob + 1
if pred[1] == 0:
prob = prob + 1
if pred[2] == 0:
prob = prob + 1
if pred[3] == 0:
prob = prob + 1
print "Persentage of Passing : ", float(
(float(prob) / 4) * float(100)), "%"
print "Previous Exam Result:", ", ".join(
map(lambda x: observations[x], bob_says))
print "Predicted Software Engineering Result:", ", ".join(
map(lambda x: states[x], pred))
return float((float(prob) / 4) * float(100))
import numpy as np
from sklearn import hmm
prng = np.random.RandomState(9)
##############################################################
n_components = 3 # ('Urna A', 'Urna B', 'Urna C')
n_symbols = 3 # ('Bola Red', 'Bola Blue', 'Bola Green')
emissionprob = [[0.34, 0.33, 0.33], [0.4, 0.55, 0.05], [0.05, 0.55, 0.4]]
startprob = [0.34, 0.33, 0.33]
transmat = [[0.34, 0.33, 0.33], [0.9, 0.05, 0.05], [0.9, 0.05, 0.05]]
h = hmm.MultinomialHMM(n_components, startprob=startprob, transmat=transmat)
h.emissionprob = emissionprob
observations = [1, 1, 2, 2, 1, 0, 1, 2, 2, 0]
h.fit(observations, n_iter=2)
print h.startprob
def f(marks, prb):
if len(marks) != 3:
print "Error"
else:
states = ["A", "B", "C", "F"]
n_states = len(states)
observations = [
"sub1-A", "sub1-B", "sub1-C", "sub1-D", "sub2-A", "sub2-B",
"sub2-C", "sub2-D", "sub3-A", "sub3-B", "sub3-C", "sub3-D"
]
#observations = ["OOP-A", "OOP-B", "OOP-C","OOP-D"]
n_observations = len(observations)
start_probability = np.array([0.25, 0.25, 0.25, 0.25])
transition_probability = np.array([[0.4, 0.3, 0.2, 0.1],
[0.3, 0.4, 0.2, 0.1],
[0.1, 0.3, 0.4, 0.2],
[0.1, 0.2, 0.3, 0.4]])
#emission_probability = np.array([[0.2, 0.1,0.1, 0.1,0.2, 0.1,0.1, 0.1],[0.2, 0.1,0.1, 0.1,0.2, 0.1,0.1, 0.1],[0.2, 0.1,0.1, 0.1,0.2, 0.1,0.1, 0.1],[0.2, 0.1,0.1, 0.1,0.2, 0.1,0.1, 0.1]])
#emission_probability = np.array([[0.4, 0.2,0.2, 0.2],[0.4, 0.2,0.2, 0.2],[0.4, 0.2,0.2, 0.2],[0.4, 0.2,0.2, 0.2]])
emission_probability = np.array(prb)
model = hmm.MultinomialHMM(n_components=n_states)
model._set_startprob(start_probability)
model._set_transmat(transition_probability)
model._set_emissionprob(emission_probability)
bob_says = marks
logprob, pred = model.decode(bob_says, algorithm="viterbi")
print "Previous Exam Result:", ", ".join(
map(lambda x: observations[x], bob_says))
print "Predicted Software Engineering Result:", ", ".join(
map(lambda x: states[x], pred))
if (pred[0] != 3) & (pred[1] != 3) & (pred[0] != 2) & (
pred[1] != 2) & (pred[2] != 3) & (pred[2] != 2):
con = "Student Approved"
else:
con = "Not Sufficient Requirenment"
window = Tk()
window.configure(background='#8A002E')
window.title('Margov Prediction')
window.geometry('380x180') # Size 200, 200
frame = Frame(window, background='#8A002E')
frame.pack()
Label(frame,
background='#8A002E',
font=("Helvetica", 14),
foreground="white",
text="Student's predicted success:").grid(row=0, column=0)
Label(frame,
background='#8A002E',
font=("Helvetica", 14),
foreground="white",
text=", ".join(map(lambda x: states[x], pred))).grid(row=0, column=1)
Label(frame,
background='#8A002E',
text=con,
fg="red",
font=("Helvetica", 24)).grid(row=1, column=0)
Button(frame, text='Quit', command=window.quit).grid(row=2,
column=2,
sticky=W,
pady=4)
return 0
n_states = len(states)
observations = list_day
n_observations = len(observations)
start_probability = np.array(start_prob_member) # has nothing to do with users
#print "start=", start_probability
transition_probability = np.array(trans_probality) # has nothing to do with users
#print "transition", transition_probability
emission_probability = np.array(emission_probab) # here there are users involved
#print "emmission", emission_probability
model = hmm.MultinomialHMM(n_components=n_states)
model._set_startprob(start_probability)
model._set_transmat(transition_probability)
model._set_emissionprob(emission_probability) # here there are users involved
# predict a sequence of hidden states based on visible states
LIST_OF_DICT_FOR_PANDAS_DATAFRAME = []
for y in range(0,MAX_USER_ID):
user_id = y+1
#print "\n\nUser " , user_id ;
for x in range (1,7):
DICT_FOR_PANDAS_DATAFRAME = {}
# -*- coding: utf-8 -*- """ Created on Fri Apr 15 15:59:17 2016 @author: gong @description:这是HMM模型 """ #from sklearn.hmm import GaussianHMM from sklearn import hmm X = [0, 1, 2, 1, 1, 2, 0, 3, 9] model = hmm.MultinomialHMM(n_components=4) model.n_symbols = 10 model.fit([X], n_iter=100) print model.predict(X) #print model.eval(X)
def f(marks, prb):
if len(marks) != 2:
print "Error"
else:
print "hidden"
states = ["A", "B", "C", "F"]
n_states = len(states)
observations = [
"Sub1:A+", "Sub1:A", "Sub1:A-", "Sub1:B+", "Sub1:B", "Sub1:B-",
"Sub1:C+", "Sub1:C", "Sub1:C_", "Sub1:F", "Sub2:A+", "Sub2:A",
"Sub2:A-", "Sub2:B+", "Sub2:B", "Sub2:B-", "Sub2:C+", "Sub2:C",
"Sub2:C_", "Sub2:F"
]
n_observations = len(observations)
#Starting proberbilities
start_probability = np.array([0.25, 0.25, 0.25, 0.25])
#Transision proberbilities
transition_probability = np.array([[0.4, 0.3, 0.2, 0.1],
[0.3, 0.4, 0.2, 0.1],
[0.1, 0.3, 0.4, 0.2],
[0.1, 0.2, 0.3, 0.4]])
# Emission proberbilities are read from file
emission_probability = np.array(prb)
model = hmm.MultinomialHMM(n_components=n_states)
model._set_startprob(start_probability)
model._set_transmat(transition_probability)
model._set_emissionprob(emission_probability)
bob_says = marks
logprob, pred = model.decode(bob_says, algorithm="viterbi")
'''
Create the mean value for general prediction
'''
val1 = 0
val2 = 0
totval = 0
if (pred[0] == 0):
val1 = 85
if (pred[0] == 1):
val1 = 65
if (pred[0] == 2):
val1 = 47.5
if (pred[0] == 3):
val1 = 20
if (pred[1] == 0):
val2 = 85
if (pred[1] == 1):
val2 = 65
if (pred[1] == 2):
val2 = 47.5
if (pred[1] == 3):
val2 = 20
totval = (val1 + val2) / 2
if (totval > 70):
gr = "A"
elif (totval > 55):
gr = "B"
elif (totval > 40):
gr = "C"
else:
gr = "F"
#If C and F values are not get as a predicted then student is displayed as approved
print "Previous Exam Result:", ", ".join(
map(lambda x: observations[x], bob_says))
print "Predicted Software Engineering Result:", ", ".join(
map(lambda x: states[x], pred))
if (pred[0] != 3) & (pred[1] != 3) & (pred[0] != 2) & (pred[1] != 2):
con = "Student Approved"
else:
con = "Not Sufficient Requirement"
#GUI Preparation
window = Tk()
window.configure(background='#8A002E')
window.title('Margov Prediction')
window.geometry('480x200') # Size 200, 200
frame = Frame(window, background='#8A002E')
frame.pack()
#Call Back Function to the main window
def back():
window.withdraw()
Label(frame,
background='#8A002E',
font=("Helvetica", 12),
foreground="white",
text="Student's predicted success:").grid(row=0, column=0)
Label(frame,
background='#8A002E',
font=("Helvetica", 12),
foreground="white",
text=", ".join(map(lambda x: states[x], pred))).grid(row=0, column=1)
Label(frame,
background='#8A002E',
text="Overall Grade is :",
fg="white",
font=("Helvetica", 16)).grid(row=1, column=0)
Label(frame,
background='#8A002E',
text=gr,
fg="white",
font=("Helvetica", 16)).grid(row=1, column=1)
Label(frame,
background='#8A002E',
text=con,
fg="red",
font=("Helvetica", 20)).grid(row=2, column=0)
Button(frame, text='Next', command=back).grid(row=3,
column=2,
sticky=W,
pady=4)
#Return Grade
return gr
def f(marks,prb):
if len(marks)!=2:
print "Error"
else:
print "hidden"
states = ["A", "B","C", "F"]
n_states = len(states)
observations = ["Sub1:A+","Sub1:A","Sub1:A-", "Sub1:B+", "Sub1:B", "Sub1:B-", "Sub1:C+", "Sub1:C", "Sub1:C_","Sub1:F","Sub2:A+","Sub2:A","Sub2:A-", "Sub2:B+", "Sub2:B", "Sub2:B-", "Sub2:C+", "Sub2:C", "Sub2:C_","Sub2:F"]
#observations = ["OOP-A", "OOP-B", "OOP-C","OOP-D"]
n_observations = len(observations)
start_probability = np.array([0.25, 0.25,0.25, 0.25])
transition_probability = np.array([[0.4, 0.3,0.2, 0.1],[0.3, 0.4,0.2, 0.1],[0.1, 0.3,0.4, 0.2],[0.1, 0.2,0.3, 0.4]])
#emission_probability = np.array([[0.2, 0.1,0.1, 0.1,0.2, 0.1,0.1, 0.1],[0.2, 0.1,0.1, 0.1,0.2, 0.1,0.1, 0.1],[0.2, 0.1,0.1, 0.1,0.2, 0.1,0.1, 0.1],[0.2, 0.1,0.1, 0.1,0.2, 0.1,0.1, 0.1]])
#emission_probability = np.array([[0.4, 0.2,0.2, 0.2],[0.4, 0.2,0.2, 0.2],[0.4, 0.2,0.2, 0.2],[0.4, 0.2,0.2, 0.2]])
emission_probability = np.array(prb)
model = hmm.MultinomialHMM(n_components=n_states)
model._set_startprob(start_probability)
model._set_transmat(transition_probability)
model._set_emissionprob(emission_probability)
bob_says = marks
logprob, pred = model.decode(bob_says, algorithm="viterbi")
val1=0
val2=0
totval=0
if(pred[0]==0):
val1=85
if(pred[0]==1):
val1=65
if(pred[0]==2):
val1=47.5
if(pred[0]==3):
val1=20
if(pred[1]==0):
val2=85
if(pred[1]==1):
val2=65
if(pred[1]==2):
val2=47.5
if(pred[1]==3):
val2=20
totval=(val1+val2)/2
if (totval>70):
gr=0
elif (totval>55):
gr=1
elif (totval>40):
gr=2
else:
gr=3
print "Previous Exam Result:", ", ".join(map(lambda x: observations[x], bob_says))
print "Predicted Software Engineering Result:", ", ".join(map(lambda x: states[x], pred))
if (pred[0]!=3)&(pred[1]!=3)&(pred[0]!=2)&(pred[1]!=2):
con="Student Approved"
else:con="Not Sufficient Requirenment"
window = Tk()
window.configure(background='#8A002E')
window.title('Margov Prediction')
window.geometry('450x200') # Size 200, 200
frame = Frame(window,background='#8A002E')
frame.pack()
Label(frame, background='#8A002E',font=("Helvetica", 14),foreground="white",text="Student's predicted success:").grid(row=0, column=0)
Label(frame, background='#8A002E',font=("Helvetica", 14),foreground="white",text=", ".join(map(lambda x: states[x], pred))).grid(row=0, column=1)
Label(frame,background='#8A002E', text=con,fg="red",font=("Helvetica", 24)).grid(row=1, column=0)
Button(frame, text='Quit', command=window.quit).grid(row=2, column=2, sticky=W, pady=4)
return gr
def hmm_model_fit(obs):
model = hmm.MultinomialHMM(n_components=2)
model.fit([obs])
return model
def test_set_emissionprob(self):
h = hmm.MultinomialHMM(self.n_components)
emissionprob = np.array([[0.8, 0.2, 0.0], [0.7, 0.2, 1.0]])
h.emissionprob = emissionprob
assert np.allclose(emissionprob, h.emissionprob)
