def _generate_words_letters(number_of_words=100,
order=1,
max_transitions=30,
words_cache=None):
if words_cache is None:
words_cache = WordsCache()
entries = entries_from_cmudict()
existing_words = [word for word, phones in entries]
entries = filter_entries(entries, 'Austen')
sequences = [[letter for letter in word] for word, phones in entries]
markov_tree = generate_markov_tree(sequences, order=order)
number_generated = 0
while number_generated < number_of_words:
cached_words = words_cache.get_words('letters')
new_sequence = generate_new_sequence(markov_tree, max_transitions)
new_word = ''.join(new_sequence)
# Reject words already in the corpus
if new_word in existing_words or new_word in cached_words:
continue
else:
words_cache.add_word('letters', new_word)
print "New word: %s" % new_word
number_generated += 1
return words_cache
def _generate_words_letters(number_of_words=100,
order=1,
max_transitions=30,
words_cache=None):
if words_cache is None:
words_cache = WordsCache()
entries = entries_from_cmudict()
existing_words = [word for word, phones in entries]
entries = filter_entries(entries, 'Austen')
sequences = [[letter for letter in word] for word, phones in entries]
markov_tree = generate_markov_tree(sequences, order=order)
number_generated = 0
while number_generated < number_of_words:
cached_words = words_cache.get_words('letters')
new_sequence = generate_new_sequence(markov_tree, max_transitions)
new_word = ''.join(new_sequence)
# Reject words already in the corpus
if new_word in existing_words or new_word in cached_words:
continue
else:
words_cache.add_word('letters', new_word)
print "New word: %s" % new_word
number_generated += 1
return words_cache
def test_first_order(self):
markov_tree = {
'START': {
'B': 1
},
'B': {
'UH': 1
},
'UH': {
'STOP': 1
}
}
sequence = generate_new_sequence(markov_tree, 10)
self.assertEqual(sequence, ['B', 'UH'])
def test_cutoff(self):
markov_tree = {
'START': {
'B': 1
},
'B': {
'UH': 1
},
'UH': {
'L': 1
},
'L': {
'STOP': 1
}
}
sequence = generate_new_sequence(markov_tree, 1)
self.assertEqual(sequence, ['B'])
def test_cutoff(self):
markov_tree = {
'START': {
'B': 1
},
'B': {
'UH': 1
},
'UH': {
'L': 1
},
'L': {
'STOP': 1
}
}
sequence = generate_new_sequence(markov_tree, 1)
self.assertEqual(sequence, ['B'])
def test_second_order(self):
markov_tree = {
'START': {
'START': {
'B': 1
},
'B': {
'UH': 1
}
},
'B': {
'UH': {
'STOP': 1
}
},
'UH': {
'STOP': {
'STOP': 1
}
}
}
sequence = generate_new_sequence(markov_tree, 10)
self.assertEqual(sequence, ['B', 'UH'])