本文整理匯總了Python中nltk.metrics.BigramAssocMeasures.pmi方法的典型用法代碼示例。如果您正苦於以下問題：Python BigramAssocMeasures.pmi方法的具體用法？Python BigramAssocMeasures.pmi怎麽用？Python BigramAssocMeasures.pmi使用的例子？那麽, 這裏精選的方法代碼示例或許可以為您提供幫助。您也可以進一步了解該方法所在類nltk.metrics.BigramAssocMeasures的用法示例。
在下文中一共展示了BigramAssocMeasures.pmi方法的3個代碼示例，這些例子默認根據受歡迎程度排序。您可以為喜歡或者感覺有用的代碼點讚，您的評價將有助於係統推薦出更棒的Python代碼示例。
示例1: test_bigram2
# 需要導入模塊: from nltk.metrics import BigramAssocMeasures [as 別名]
# 或者: from nltk.metrics.BigramAssocMeasures import pmi [as 別名]
def test_bigram2(self):
sent = 'this this is is a a test test'.split()
b = BigramCollocationFinder.from_words(sent)
#python 2.6 does not have assertItemsEqual or assertListEqual
self.assertEqual(
sorted(b.ngram_fd.items()),
sorted([(('a', 'a'), 1), (('a', 'test'), 1), (('is', 'a'), 1), (('is', 'is'), 1), (('test', 'test'), 1), (('this', 'is'), 1), (('this', 'this'), 1)])
)
self.assertEqual(
sorted(b.word_fd.items()),
sorted([('a', 2), ('is', 2), ('test', 2), ('this', 2)])
)
self.assertTrue(len(sent) == sum(b.word_fd.values()) == sum(b.ngram_fd.values()) + 1)
self.assertTrue(close_enough(
sorted(b.score_ngrams(BigramAssocMeasures.pmi)),
sorted([(('a', 'a'), 1.0), (('a', 'test'), 1.0), (('is', 'a'), 1.0), (('is', 'is'), 1.0), (('test', 'test'), 1.0), (('this', 'is'), 1.0), (('this', 'this'), 1.0)])
)) 示例2: test_bigram3
# 需要導入模塊: from nltk.metrics import BigramAssocMeasures [as 別名]
# 或者: from nltk.metrics.BigramAssocMeasures import pmi [as 別名]
def test_bigram3(self):
sent = 'this this is is a a test test'.split()
b = BigramCollocationFinder.from_words(sent, window_size=3)
self.assertEqual(
sorted(b.ngram_fd.items()),
sorted([(('a', 'test'), 3), (('is', 'a'), 3), (('this', 'is'), 3), (('a', 'a'), 1), (('is', 'is'), 1), (('test', 'test'), 1), (('this', 'this'), 1)])
)
self.assertEqual(
sorted(b.word_fd.items()),
sorted([('a', 2), ('is', 2), ('test', 2), ('this', 2)])
)
self.assertTrue(len(sent) == sum(b.word_fd.values()) == (sum(b.ngram_fd.values()) + 2 + 1) / 2.0)
self.assertTrue(close_enough(
sorted(b.score_ngrams(BigramAssocMeasures.pmi)),
sorted([(('a', 'test'), 1.584962500721156), (('is', 'a'), 1.584962500721156), (('this', 'is'), 1.584962500721156), (('a', 'a'), 0.0), (('is', 'is'), 0.0), (('test', 'test'), 0.0), (('this', 'this'), 0.0)])
)) 示例3: test_bigram5
# 需要導入模塊: from nltk.metrics import BigramAssocMeasures [as 別名]
# 或者: from nltk.metrics.BigramAssocMeasures import pmi [as 別名]
def test_bigram5(self):
sent = 'this this is is a a test test'.split()
b = BigramCollocationFinder.from_words(sent, window_size=5)
self.assertEqual(
sorted(b.ngram_fd.items()),
sorted([(('a', 'test'), 4), (('is', 'a'), 4), (('this', 'is'), 4), (('is', 'test'), 3), (('this', 'a'), 3), (('a', 'a'), 1), (('is', 'is'), 1), (('test', 'test'), 1), (('this', 'this'), 1)])
)
self.assertEqual(
sorted(b.word_fd.items()),
sorted([('a', 2), ('is', 2), ('test', 2), ('this', 2)])
)
self.assertTrue(len(sent) == sum(b.word_fd.values()) == (sum(b.ngram_fd.values()) + 4 + 3 + 2 + 1) / 4.0)
self.assertTrue(close_enough(
sorted(b.score_ngrams(BigramAssocMeasures.pmi)),
sorted([(('a', 'test'), 1.0), (('is', 'a'), 1.0), (('this', 'is'), 1.0), (('is', 'test'), 0.5849625007211562), (('this', 'a'), 0.5849625007211562), (('a', 'a'), -1.0), (('is', 'is'), -1.0), (('test', 'test'), -1.0), (('this', 'this'), -1.0)])
))