def setUp(self):
self.phraseVector = {
'project': 1,
'cluster': 1,
'highdimensional': 1,
'streams': 1
}
self.phraseTextAndDimensionMap = TwoWayMap()
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'project', 0)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 1)
self.phraseTextToPhraseObjectMap = {
'project':
Phrase('project', test_time, score=8),
'cluster':
Phrase('cluster', test_time, score=8),
'abcd':
Phrase(
'abcd',
test_time -
3 * stream_settings['max_phrase_inactivity_time_in_seconds'],
score=8)
}
self.vector = Vector({0: 1, 1: 1, 2: 1, 3: 1})
self.initial_max_dimensions = stream_settings['dimensions']
stream_settings['dimensions'] = 2
def setUp(self):
self.dimension, self.signatureLength = 50, 23
self.phraseTextAndDimensionMap = TwoWayMap()
for i in range(self.dimension):
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, i, i)
self.unitRandomVectors = [
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1)
for i in range(self.signatureLength)
]
self.doc1 = Document(
1,
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1))
self.doc2 = Document(
2,
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1))
self.doc1.setSignatureUsingVectors(self.unitRandomVectors,
self.phraseTextAndDimensionMap)
self.doc2.setSignatureUsingVectors(self.unitRandomVectors,
self.phraseTextAndDimensionMap)
self.pm = SignaturePermutationWithTrie(
signatureLength=self.signatureLength)
self.pm.addDocument(self.doc1)
self.pm.addDocument(self.doc2)
def test_setSignatureUsingVectors(self):
phraseTextAndDimensionMap = TwoWayMap()
phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'a', 1)
phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'b', 2)
documentWithDimensionsInVector = Document(1, {'a':1, 'b':4})
documentWithDimensionsNotInVector = Document(1, {'a':1, 'c':4})
vectors = [ Vector({1: 3/5., 2: -4/5.}), Vector({1:-5/13., 2: 12/13.})]
documentWithDimensionsInVector.setSignatureUsingVectors(vectors, phraseTextAndDimensionMap)
documentWithDimensionsNotInVector.setSignatureUsingVectors(vectors, phraseTextAndDimensionMap)
self.assertEqual(Signature('01'), documentWithDimensionsInVector.signature)
self.assertEqual(Signature('10'), documentWithDimensionsNotInVector.signature)
def setUp(self):
self.phraseVector = {'project':1, 'cluster':1, 'highdimensional':1, 'streams':1}
self.phraseTextAndDimensionMap = TwoWayMap()
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'project', 0)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 1)
self.finalPhraseToIdMap = {'project': 0, 'cluster': 1, 'streams': 2, 'highdimensional': 3}
settings['dimensions'] = 2
class UtilityMethodsTests(unittest.TestCase):
def setUp(self):
self.phraseVector = {'project':1, 'cluster':1, 'highdimensional':1, 'streams':1}
self.phraseTextAndDimensionMap = TwoWayMap()
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'project', 0)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 1)
self.finalPhraseToIdMap = {'project': 0, 'cluster': 1, 'streams': 2, 'highdimensional': 3}
settings['dimensions'] = 2
def test_updatePhraseTextAndDimensionsMap_PhraseMapHasLesserDimensions(self):
settings['dimensions'] = 4
UtilityMethods.updatePhraseTextAndDimensionsMap(self.phraseVector, self.phraseTextAndDimensionMap, **settings)
self.assertEqual(self.finalPhraseToIdMap, self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
def test_updatePhraseTextAndDimensionsMap_PhraseMapHasMaximumDimensions(self):
UtilityMethods.updatePhraseTextAndDimensionsMap(self.phraseVector, self.phraseTextAndDimensionMap, **settings)
for k in ['streams', 'highdimensional']: del self.finalPhraseToIdMap[k]
self.assertEqual(self.finalPhraseToIdMap, self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
def __init__(self, **settings):
self.settings = settings
self.nearestNeighborThreshold = settings['nearest_neighbor_threshold']
self.unitVector = RandomGaussianUnitVector(
dimensions=settings['dimensions'], mu=0, sigma=1)
self.vectorPermutations = VectorPermutation.getPermutations(
settings['signature_length'], settings['dimensions'],
self.unitVector)
# self.signaturePermutations = [SignaturePermutationWithTrie(settings['signature_length']) for i in range(settings['number_of_permutations'])]
signatureType = settings.get('signature_type', 'signature_type_trie')
if signatureType == 'signature_type_trie':
self.signaturePermutations = [
SignaturePermutationWithTrie(settings['signature_length'])
for i in range(settings['number_of_permutations'])
]
else:
self.signaturePermutations = [
SignaturePermutationWithSortedList(
settings['signature_length'])
for i in range(settings['number_of_permutations'])
]
self.phraseTextAndDimensionMap = TwoWayMap()
self.documentIdToDocumentMap = {}
class SignaturePermutationTests(unittest.TestCase):
def setUp(self):
self.dimension, self.signatureLength = 50, 23
self.phraseTextAndDimensionMap = TwoWayMap()
for i in range(self.dimension): self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, i,i)
self.unitRandomVectors = [VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1) for i in range(self.signatureLength)]
self.doc1=Document(1, VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1))
self.doc2=Document(2, VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1))
self.doc1.setSignatureUsingVectors(self.unitRandomVectors, self.phraseTextAndDimensionMap); self.doc2.setSignatureUsingVectors(self.unitRandomVectors, self.phraseTextAndDimensionMap)
self.pm = SignaturePermutationWithTrie(signatureLength=self.signatureLength)
self.pm.addDocument(self.doc1)
self.pm.addDocument(self.doc2)
def test_addDocument_newKey(self):
doc1=Document(1, VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1))
doc1.setSignatureUsingVectors(self.unitRandomVectors, self.phraseTextAndDimensionMap)
pm = SignaturePermutationWithTrie(signatureLength=self.signatureLength)
pm.addDocument(doc1)
self.assertEqual(pm.signatureTrie[doc1.signature.permutate(pm).to01()], set([1]))
def test_addDocument_existingKey(self):
newDocModifiedWithExistingSignature = Document(3, VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1))
newDocModifiedWithExistingSignature.signature = Signature(self.doc1.signature.to01())
self.pm.addDocument(newDocModifiedWithExistingSignature)
self.assertEqual(self.pm.signatureTrie[self.doc1.signature.permutate(self.pm).to01()], set([1, 3]))
def test_getNearestDocument_usingAKeyAlreadyInTrie(self): self.assertEqual(self.pm.getNearestDocuments(self.doc1), set([1]))
def test_getNearestDocument_usingANearbyKeyInTrie(self):
digitReplacement = {'0': '1', '1': '0'}
newDocWithANearbySignature = Document(3, VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1))
exactSignature = self.doc1.signature.to01()
newDocWithANearbySignature.signature = Signature(exactSignature[:-1]+digitReplacement[exactSignature[-1]])
self.assertNotEquals(self.doc1.signature.to01(), newDocWithANearbySignature.signature.to01())
self.assertEqual(self.pm.getNearestDocuments(newDocWithANearbySignature), set([1])) # This assertion can sometimes fail because of randomization. Run the tests again. It's OK!
def test_getNearestDocument_emptyTrie(self):
permutationWithEmptyTrie = SignaturePermutationWithTrie(signatureLength=self.signatureLength)
self.assertEqual(permutationWithEmptyTrie.getNearestDocuments(self.doc1), set())
def test_removeDocument_documents(self):
newDocModifiedWithExistingSignature = Document(3, VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1))
newDocModifiedWithExistingSignature.signature = Signature(self.doc1.signature.to01())
self.pm.addDocument(newDocModifiedWithExistingSignature)
self.assertEqual(self.pm.signatureTrie[self.doc1.signature.permutate(self.pm).to01()], set([1, 3]))
self.pm.removeDocument(newDocModifiedWithExistingSignature)
self.assertEqual(self.pm.signatureTrie[self.doc1.signature.permutate(self.pm).to01()], set([1]))
self.pm.removeDocument(self.doc1)
self.assertEqual(None, self.pm.signatureTrie.get(self.doc1.signature.permutate(self.pm).to01()))
def test_resetSignatureTrie(self):
self.assertTrue(len(self.pm.signatureTrie)>0)
self.pm.resetSignatureDataStructure()
self.assertTrue(len(self.pm.signatureTrie)==0)
def setUp(self):
self.phraseVector = {
'project': 1,
'cluster': 1,
'highdimensional': 1,
'streams': 1
}
self.phraseTextAndDimensionMap = TwoWayMap()
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'project', 0)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 1)
self.finalPhraseToIdMap = {
'project': 0,
'cluster': 1,
'streams': 2,
'highdimensional': 3
}
settings['dimensions'] = 2
def setUp(self):
self.phraseVector = {'project':1, 'cluster':1, 'highdimensional':1, 'streams':1}
self.phraseTextAndDimensionMap = TwoWayMap()
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'project', 0)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 1)
self.phraseTextToPhraseObjectMap = {'project': Phrase('project', test_time, score=8), 'cluster': Phrase('cluster', test_time, score=8), 'abcd': Phrase('abcd', test_time-3*stream_settings['max_phrase_inactivity_time_in_seconds'], score=8)}
self.vector = Vector({0:1, 1:1, 2:1, 3:1})
self.initial_max_dimensions = stream_settings['dimensions']
stream_settings['dimensions'] = 2
def setUp(self):
self.dimension, self.signatureLength = 50, 23
self.phraseTextAndDimensionMap = TwoWayMap()
for i in range(self.dimension): self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, i,i)
self.unitRandomVectors = [VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1) for i in range(self.signatureLength)]
self.doc1=Document(1, VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1))
self.doc2=Document(2, VectorGenerator.getRandomGaussianUnitVector(dimension=self.dimension, mu=0, sigma=1))
self.doc1.setSignatureUsingVectors(self.unitRandomVectors, self.phraseTextAndDimensionMap); self.doc2.setSignatureUsingVectors(self.unitRandomVectors, self.phraseTextAndDimensionMap)
self.pm = SignaturePermutationWithTrie(signatureLength=self.signatureLength)
self.pm.addDocument(self.doc1)
self.pm.addDocument(self.doc2)
class UtilityMethodsTests(unittest.TestCase):
def setUp(self):
self.phraseVector = {
'project': 1,
'cluster': 1,
'highdimensional': 1,
'streams': 1
}
self.phraseTextAndDimensionMap = TwoWayMap()
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'project', 0)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 1)
self.finalPhraseToIdMap = {
'project': 0,
'cluster': 1,
'streams': 2,
'highdimensional': 3
}
settings['dimensions'] = 2
def test_updatePhraseTextAndDimensionsMap_PhraseMapHasLesserDimensions(
self):
settings['dimensions'] = 4
UtilityMethods.updatePhraseTextAndDimensionsMap(
self.phraseVector, self.phraseTextAndDimensionMap, **settings)
self.assertEqual(
self.finalPhraseToIdMap,
self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
def test_updatePhraseTextAndDimensionsMap_PhraseMapHasMaximumDimensions(
self):
UtilityMethods.updatePhraseTextAndDimensionsMap(
self.phraseVector, self.phraseTextAndDimensionMap, **settings)
for k in ['streams', 'highdimensional']:
del self.finalPhraseToIdMap[k]
self.assertEqual(
self.finalPhraseToIdMap,
self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
def test_setSignatureUsingVectors(self):
phraseTextAndDimensionMap = TwoWayMap()
phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'a', 1)
phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'b', 2)
documentWithDimensionsInVector = Document(1, {'a': 1, 'b': 4})
documentWithDimensionsNotInVector = Document(1, {'a': 1, 'c': 4})
vectors = [
Vector({
1: 3 / 5.,
2: -4 / 5.
}),
Vector({
1: -5 / 13.,
2: 12 / 13.
})
]
documentWithDimensionsInVector.setSignatureUsingVectors(
vectors, phraseTextAndDimensionMap)
documentWithDimensionsNotInVector.setSignatureUsingVectors(
vectors, phraseTextAndDimensionMap)
self.assertEqual(Signature('01'),
documentWithDimensionsInVector.signature)
self.assertEqual(Signature('10'),
documentWithDimensionsNotInVector.signature)
def test_setSignatureUsingVectorPermutations(self):
dimensions, signatureLength = 53, 13
phraseTextAndDimensionMap = TwoWayMap()
for i in range(dimensions):
phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, i, i)
phraseTextAndDimensionMapWithMissingDimensions = TwoWayMap()
for i in range(dimensions - 50):
phraseTextAndDimensionMapWithMissingDimensions.set(
TwoWayMap.MAP_FORWARD, i, i)
unitVector = RandomGaussianUnitVector(dimensions=dimensions,
mu=0,
sigma=1)
vectorPermutations = VectorPermutation.getPermutations(
signatureLength, dimensions, unitVector)
permutatedUnitVectors = [
unitVector.getPermutedVector(r) for r in vectorPermutations
]
documentVector = VectorGenerator.getRandomGaussianUnitVector(
dimension=dimensions, mu=0, sigma=1)
documentWithSignatureByVectors = Document(1, documentVector)
documentWithSignatureByVectorPermutations = Document(2, documentVector)
documentWithSignatureByVectors.setSignatureUsingVectors(
permutatedUnitVectors, phraseTextAndDimensionMap)
documentWithSignatureByVectorPermutations.setSignatureUsingVectorPermutations(
unitVector, vectorPermutations, phraseTextAndDimensionMap)
self.assertEqual(documentWithSignatureByVectors.signature,
documentWithSignatureByVectorPermutations.signature)
documentWithSignatureByVectors.setSignatureUsingVectors(
permutatedUnitVectors,
phraseTextAndDimensionMapWithMissingDimensions)
documentWithSignatureByVectorPermutations.setSignatureUsingVectorPermutations(
unitVector, vectorPermutations,
phraseTextAndDimensionMapWithMissingDimensions)
self.assertEqual(documentWithSignatureByVectors.signature,
documentWithSignatureByVectorPermutations.signature)
def __init__(self, **clustering_settings):
self.thresholdForDocumentToBeInACluster = clustering_settings[
'threshold_for_document_to_be_in_cluster']
self.unitVector = RandomGaussianUnitVector(
dimensions=clustering_settings['dimensions'], mu=0, sigma=1)
self.vectorPermutations = VectorPermutation.getPermutations(
clustering_settings['signature_length'],
clustering_settings['dimensions'], self.unitVector)
signatureType = clustering_settings.get('signature_type',
'signature_type_trie')
if signatureType == 'signature_type_trie':
self.signaturePermutations = [
SignaturePermutationWithTrie(
clustering_settings['signature_length'])
for i in range(clustering_settings['number_of_permutations'])
]
else:
self.signaturePermutations = [
SignaturePermutationWithSortedList(
clustering_settings['signature_length'])
for i in range(clustering_settings['number_of_permutations'])
]
self.phraseTextAndDimensionMap, self.clusters = TwoWayMap(), {}
self.clustering_settings = clustering_settings
def test_setSignatureUsingVectorPermutations(self):
dimensions, signatureLength = 53, 13
phraseTextAndDimensionMap = TwoWayMap()
for i in range(dimensions): phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, i,i)
phraseTextAndDimensionMapWithMissingDimensions = TwoWayMap()
for i in range(dimensions-50): phraseTextAndDimensionMapWithMissingDimensions.set(TwoWayMap.MAP_FORWARD, i,i)
unitVector = RandomGaussianUnitVector(dimensions=dimensions, mu=0, sigma=1)
vectorPermutations = VectorPermutation.getPermutations(signatureLength, dimensions, unitVector)
permutatedUnitVectors = [unitVector.getPermutedVector(r) for r in vectorPermutations]
documentVector = VectorGenerator.getRandomGaussianUnitVector(dimension=dimensions, mu=0, sigma=1)
documentWithSignatureByVectors=Document(1, documentVector)
documentWithSignatureByVectorPermutations=Document(2, documentVector)
documentWithSignatureByVectors.setSignatureUsingVectors(permutatedUnitVectors, phraseTextAndDimensionMap)
documentWithSignatureByVectorPermutations.setSignatureUsingVectorPermutations(unitVector, vectorPermutations, phraseTextAndDimensionMap)
self.assertEqual(documentWithSignatureByVectors.signature, documentWithSignatureByVectorPermutations.signature)
documentWithSignatureByVectors.setSignatureUsingVectors(permutatedUnitVectors, phraseTextAndDimensionMapWithMissingDimensions)
documentWithSignatureByVectorPermutations.setSignatureUsingVectorPermutations(unitVector, vectorPermutations, phraseTextAndDimensionMapWithMissingDimensions)
self.assertEqual(documentWithSignatureByVectors.signature, documentWithSignatureByVectorPermutations.signature)
class UtilityMethodsTests(unittest.TestCase):
def setUp(self):
self.phraseVector = {'project':1, 'cluster':1, 'highdimensional':1, 'streams':1}
self.phraseTextAndDimensionMap = TwoWayMap()
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'project', 0)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 1)
self.phraseTextToPhraseObjectMap = {'project': Phrase('project', test_time, score=8), 'cluster': Phrase('cluster', test_time, score=8), 'abcd': Phrase('abcd', test_time-3*stream_settings['max_phrase_inactivity_time_in_seconds'], score=8)}
self.vector = Vector({0:1, 1:1, 2:1, 3:1})
self.initial_max_dimensions = stream_settings['dimensions']
stream_settings['dimensions'] = 2
def tearDown(self): stream_settings['dimensions'] = self.initial_max_dimensions
def test_updatedPhraseObject_PhraseObjectScoresAreUpdatedCorrectly(self):
UtilityMethods.updatePhraseTextToPhraseObject(self.phraseVector, test_time+timedelta(seconds=60), self.phraseTextToPhraseObjectMap, **stream_settings)
self.assertEqual(5, len(self.phraseTextToPhraseObjectMap))
self.assertEqual(5, self.phraseTextToPhraseObjectMap['project'].score)
self.assertEqual(1, self.phraseTextToPhraseObjectMap['streams'].score)
def test_updatedPhraseObject_phrase_does_not_exist_in_phraseToIdMap_but_exists_in_phraseTextToPhraseObjectMap_with_dimensions_full(self):
stream_settings['dimensions'] = 1
self.phraseTextAndDimensionMap.remove(TwoWayMap.MAP_FORWARD, 'cluster')
UtilityMethods.updatePhraseTextToPhraseObject(self.phraseVector, test_time+timedelta(seconds=60), self.phraseTextToPhraseObjectMap, **stream_settings)
self.assertEqual({'project':0}, self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
self.assertEqual(5, len(self.phraseTextToPhraseObjectMap))
self.assertEqual(5, self.phraseTextToPhraseObjectMap['project'].score)
self.assertEqual(5, self.phraseTextToPhraseObjectMap['cluster'].score)
self.assertEqual(1, self.phraseTextToPhraseObjectMap['streams'].score)
def test_createOrAddNewPhraseObject(self):
UtilityMethods.createOrAddNewPhraseObject('new_phrase', self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
UtilityMethods.createOrAddNewPhraseObject('project', self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
self.assertEqual(4, len(self.phraseTextToPhraseObjectMap))
self.assertEqual(1, self.phraseTextToPhraseObjectMap['new_phrase'].score)
self.assertEqual(9, self.phraseTextToPhraseObjectMap['project'].score)
def test_updateDimensions_when_phraseTextToIdMap_is_filled_to_max_dimensions(self):
for phrase, score in zip(['added'], range(10,11)): self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase, test_time, score=score)
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap, self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
self.assertEqual({'project':0, 'added': 1}, self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
def test_updateDimensions_when_phraseTextToIdMap_is_filled_to_max_dimensions_and_entire_map_is_changed(self):
for phrase, score in zip(['added', 'are'], range(10,12)): self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase, test_time, score=score)
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap, self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
self.assertEqual({'added':1, 'are': 0}, self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
def test_updateDimensions_when_phraseTextToIdMap_has_lesser_than_max_dimensions(self):
stream_settings['dimensions'] = 4
for phrase, score in zip(['new', 'phrases', 'are', 'added'], range(7,11)): self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase, test_time, score=score)
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap, self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
self.assertEqual(set({'project':0, 'phrases': 1, 'are':2, 'added':3}), set(self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD)))
self.assertEqual(4, len(self.phraseTextAndDimensionMap))
def test_updateDimensions_when_phrases_with_lower_id_are_removed_from_phraseTextToIdMap(self):
stream_settings['dimensions'] = 3
for phrase, score in zip(['new', 'phrases', 'are'], range(100,103)): self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase, test_time, score=score)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 2)
self.phraseTextToPhraseObjectMap['cluster'].score=100
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap, self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
self.assertEqual(range(3), sorted(self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD).values()))
def test_updateDimensions_remove_old_phrases(self):
originalTime=self.phraseTextToPhraseObjectMap['abcd'].latestOccuranceTime
self.phraseTextToPhraseObjectMap['abcd'].latestOccuranceTime=test_time
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap, self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
self.assertTrue('abcd' in self.phraseTextToPhraseObjectMap)
self.phraseTextToPhraseObjectMap['abcd'].latestOccuranceTime=originalTime
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap, self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
self.assertTrue('abcd' not in self.phraseTextToPhraseObjectMap)
def test_updateDimensions_when_dimensions_have_to_be_removed(self):
stream_settings['dimensions'] = 4
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'abcdx', 2)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'abcdxy', 3)
for phrase, score in zip(['new_text'], range(7,8)): self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase, test_time, score=score)
self.phraseTextToPhraseObjectMap['cluster'].latestOccuranceTime=test_time-3*stream_settings['max_phrase_inactivity_time_in_seconds']
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap, self.phraseTextToPhraseObjectMap, test_time, **stream_settings)
self.assertEqual(set({'project':0, 'new_text': 1}), set(self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD)))
def test_checkCriticalErrorsInPhraseTextToIdMap_larger_than_expected_dimensions(self):
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'sdfsd', 3)
print 'Ignore this message: ',
self.assertRaises(SystemExit, UtilityMethods.checkCriticalErrorsInPhraseTextToIdMap, self.phraseTextAndDimensionMap, **stream_settings)
def test_pruningConditionDeterministic(self):
phrase1 = Phrase('dsf', test_time-3*stream_settings['max_phrase_inactivity_time_in_seconds'], 1)
phrase2 = Phrase('dsf', test_time, 1)
self.assertTrue(UtilityMethods.pruningConditionDeterministic(phrase1, test_time, **stream_settings))
self.assertFalse(UtilityMethods.pruningConditionDeterministic(phrase2, test_time, **stream_settings))
def test_pruningConditionRandom(self):
phrase1 = Phrase('dsf', test_time-3*stream_settings['max_phrase_inactivity_time_in_seconds'], 1)
phrase2 = Phrase('dsf', test_time, 1)
self.assertTrue(UtilityMethods.pruningConditionRandom(phrase1, test_time, **stream_settings))
self.assertFalse(UtilityMethods.pruningConditionRandom(phrase2, test_time, **stream_settings))
def test_pruneUnnecessaryPhrases(self):
phraseTextToPhraseObjectMap = {'dsf': Phrase('dsf', test_time-3*stream_settings['max_phrase_inactivity_time_in_seconds'], 1), 'abc': Phrase('abc', test_time, 1)}
UtilityMethods.pruneUnnecessaryPhrases(phraseTextToPhraseObjectMap, test_time, UtilityMethods.pruningConditionRandom, **stream_settings)
self.assertTrue('dsf' not in phraseTextToPhraseObjectMap)
self.assertTrue('abc' in phraseTextToPhraseObjectMap)
class UtilityMethodsTests(unittest.TestCase):
def setUp(self):
self.phraseVector = {
'project': 1,
'cluster': 1,
'highdimensional': 1,
'streams': 1
}
self.phraseTextAndDimensionMap = TwoWayMap()
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'project', 0)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 1)
self.phraseTextToPhraseObjectMap = {
'project':
Phrase('project', test_time, score=8),
'cluster':
Phrase('cluster', test_time, score=8),
'abcd':
Phrase(
'abcd',
test_time -
3 * stream_settings['max_phrase_inactivity_time_in_seconds'],
score=8)
}
self.vector = Vector({0: 1, 1: 1, 2: 1, 3: 1})
self.initial_max_dimensions = stream_settings['dimensions']
stream_settings['dimensions'] = 2
def tearDown(self):
stream_settings['dimensions'] = self.initial_max_dimensions
def test_updatedPhraseObject_PhraseObjectScoresAreUpdatedCorrectly(self):
UtilityMethods.updatePhraseTextToPhraseObject(
self.phraseVector, test_time + timedelta(seconds=60),
self.phraseTextToPhraseObjectMap, **stream_settings)
self.assertEqual(5, len(self.phraseTextToPhraseObjectMap))
self.assertEqual(5, self.phraseTextToPhraseObjectMap['project'].score)
self.assertEqual(1, self.phraseTextToPhraseObjectMap['streams'].score)
def test_updatedPhraseObject_phrase_does_not_exist_in_phraseToIdMap_but_exists_in_phraseTextToPhraseObjectMap_with_dimensions_full(
self):
stream_settings['dimensions'] = 1
self.phraseTextAndDimensionMap.remove(TwoWayMap.MAP_FORWARD, 'cluster')
UtilityMethods.updatePhraseTextToPhraseObject(
self.phraseVector, test_time + timedelta(seconds=60),
self.phraseTextToPhraseObjectMap, **stream_settings)
self.assertEqual({'project': 0},
self.phraseTextAndDimensionMap.getMap(
TwoWayMap.MAP_FORWARD))
self.assertEqual(5, len(self.phraseTextToPhraseObjectMap))
self.assertEqual(5, self.phraseTextToPhraseObjectMap['project'].score)
self.assertEqual(5, self.phraseTextToPhraseObjectMap['cluster'].score)
self.assertEqual(1, self.phraseTextToPhraseObjectMap['streams'].score)
def test_createOrAddNewPhraseObject(self):
UtilityMethods.createOrAddNewPhraseObject(
'new_phrase', self.phraseTextToPhraseObjectMap, test_time,
**stream_settings)
UtilityMethods.createOrAddNewPhraseObject(
'project', self.phraseTextToPhraseObjectMap, test_time,
**stream_settings)
self.assertEqual(4, len(self.phraseTextToPhraseObjectMap))
self.assertEqual(1,
self.phraseTextToPhraseObjectMap['new_phrase'].score)
self.assertEqual(9, self.phraseTextToPhraseObjectMap['project'].score)
def test_updateDimensions_when_phraseTextToIdMap_is_filled_to_max_dimensions(
self):
for phrase, score in zip(['added'], range(10, 11)):
self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase,
test_time,
score=score)
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap,
self.phraseTextToPhraseObjectMap,
test_time, **stream_settings)
self.assertEqual({
'project': 0,
'added': 1
}, self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
def test_updateDimensions_when_phraseTextToIdMap_is_filled_to_max_dimensions_and_entire_map_is_changed(
self):
for phrase, score in zip(['added', 'are'], range(10, 12)):
self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase,
test_time,
score=score)
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap,
self.phraseTextToPhraseObjectMap,
test_time, **stream_settings)
self.assertEqual({
'added': 1,
'are': 0
}, self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD))
def test_updateDimensions_when_phraseTextToIdMap_has_lesser_than_max_dimensions(
self):
stream_settings['dimensions'] = 4
for phrase, score in zip(['new', 'phrases', 'are', 'added'],
range(7, 11)):
self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase,
test_time,
score=score)
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap,
self.phraseTextToPhraseObjectMap,
test_time, **stream_settings)
self.assertEqual(
set({
'project': 0,
'phrases': 1,
'are': 2,
'added': 3
}),
set(self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD)))
self.assertEqual(4, len(self.phraseTextAndDimensionMap))
def test_updateDimensions_when_phrases_with_lower_id_are_removed_from_phraseTextToIdMap(
self):
stream_settings['dimensions'] = 3
for phrase, score in zip(['new', 'phrases', 'are'], range(100, 103)):
self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase,
test_time,
score=score)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'cluster', 2)
self.phraseTextToPhraseObjectMap['cluster'].score = 100
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap,
self.phraseTextToPhraseObjectMap,
test_time, **stream_settings)
self.assertEqual(
range(3),
sorted(
self.phraseTextAndDimensionMap.getMap(
TwoWayMap.MAP_FORWARD).values()))
def test_updateDimensions_remove_old_phrases(self):
originalTime = self.phraseTextToPhraseObjectMap[
'abcd'].latestOccuranceTime
self.phraseTextToPhraseObjectMap[
'abcd'].latestOccuranceTime = test_time
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap,
self.phraseTextToPhraseObjectMap,
test_time, **stream_settings)
self.assertTrue('abcd' in self.phraseTextToPhraseObjectMap)
self.phraseTextToPhraseObjectMap[
'abcd'].latestOccuranceTime = originalTime
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap,
self.phraseTextToPhraseObjectMap,
test_time, **stream_settings)
self.assertTrue('abcd' not in self.phraseTextToPhraseObjectMap)
def test_updateDimensions_when_dimensions_have_to_be_removed(self):
stream_settings['dimensions'] = 4
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'abcdx', 2)
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'abcdxy', 3)
for phrase, score in zip(['new_text'], range(7, 8)):
self.phraseTextToPhraseObjectMap[phrase] = Phrase(phrase,
test_time,
score=score)
self.phraseTextToPhraseObjectMap[
'cluster'].latestOccuranceTime = test_time - 3 * stream_settings[
'max_phrase_inactivity_time_in_seconds']
UtilityMethods.updateDimensions(self.phraseTextAndDimensionMap,
self.phraseTextToPhraseObjectMap,
test_time, **stream_settings)
self.assertEqual(
set({
'project': 0,
'new_text': 1
}),
set(self.phraseTextAndDimensionMap.getMap(TwoWayMap.MAP_FORWARD)))
def test_checkCriticalErrorsInPhraseTextToIdMap_larger_than_expected_dimensions(
self):
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, 'sdfsd', 3)
print 'Ignore this message: ',
self.assertRaises(
SystemExit, UtilityMethods.checkCriticalErrorsInPhraseTextToIdMap,
self.phraseTextAndDimensionMap, **stream_settings)
def test_pruningConditionDeterministic(self):
phrase1 = Phrase(
'dsf', test_time -
3 * stream_settings['max_phrase_inactivity_time_in_seconds'], 1)
phrase2 = Phrase('dsf', test_time, 1)
self.assertTrue(
UtilityMethods.pruningConditionDeterministic(
phrase1, test_time, **stream_settings))
self.assertFalse(
UtilityMethods.pruningConditionDeterministic(
phrase2, test_time, **stream_settings))
def test_pruningConditionRandom(self):
phrase1 = Phrase(
'dsf', test_time -
3 * stream_settings['max_phrase_inactivity_time_in_seconds'], 1)
phrase2 = Phrase('dsf', test_time, 1)
self.assertTrue(
UtilityMethods.pruningConditionRandom(phrase1, test_time,
**stream_settings))
self.assertFalse(
UtilityMethods.pruningConditionRandom(phrase2, test_time,
**stream_settings))
def test_pruneUnnecessaryPhrases(self):
phraseTextToPhraseObjectMap = {
'dsf':
Phrase(
'dsf', test_time -
3 * stream_settings['max_phrase_inactivity_time_in_seconds'],
1),
'abc':
Phrase('abc', test_time, 1)
}
UtilityMethods.pruneUnnecessaryPhrases(
phraseTextToPhraseObjectMap, test_time,
UtilityMethods.pruningConditionRandom, **stream_settings)
self.assertTrue('dsf' not in phraseTextToPhraseObjectMap)
self.assertTrue('abc' in phraseTextToPhraseObjectMap)
class SignaturePermutationTests(unittest.TestCase):
def setUp(self):
self.dimension, self.signatureLength = 50, 23
self.phraseTextAndDimensionMap = TwoWayMap()
for i in range(self.dimension):
self.phraseTextAndDimensionMap.set(TwoWayMap.MAP_FORWARD, i, i)
self.unitRandomVectors = [
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1)
for i in range(self.signatureLength)
]
self.doc1 = Document(
1,
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1))
self.doc2 = Document(
2,
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1))
self.doc1.setSignatureUsingVectors(self.unitRandomVectors,
self.phraseTextAndDimensionMap)
self.doc2.setSignatureUsingVectors(self.unitRandomVectors,
self.phraseTextAndDimensionMap)
self.pm = SignaturePermutationWithTrie(
signatureLength=self.signatureLength)
self.pm.addDocument(self.doc1)
self.pm.addDocument(self.doc2)
def test_addDocument_newKey(self):
doc1 = Document(
1,
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1))
doc1.setSignatureUsingVectors(self.unitRandomVectors,
self.phraseTextAndDimensionMap)
pm = SignaturePermutationWithTrie(signatureLength=self.signatureLength)
pm.addDocument(doc1)
self.assertEqual(pm.signatureTrie[doc1.signature.permutate(pm).to01()],
set([1]))
def test_addDocument_existingKey(self):
newDocModifiedWithExistingSignature = Document(
3,
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1))
newDocModifiedWithExistingSignature.signature = Signature(
self.doc1.signature.to01())
self.pm.addDocument(newDocModifiedWithExistingSignature)
self.assertEqual(
self.pm.signatureTrie[self.doc1.signature.permutate(
self.pm).to01()], set([1, 3]))
def test_getNearestDocument_usingAKeyAlreadyInTrie(self):
self.assertEqual(self.pm.getNearestDocuments(self.doc1), set([1]))
def test_getNearestDocument_usingANearbyKeyInTrie(self):
digitReplacement = {'0': '1', '1': '0'}
newDocWithANearbySignature = Document(
3,
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1))
exactSignature = self.doc1.signature.to01()
newDocWithANearbySignature.signature = Signature(
exactSignature[:-1] + digitReplacement[exactSignature[-1]])
self.assertNotEquals(self.doc1.signature.to01(),
newDocWithANearbySignature.signature.to01())
self.assertEqual(
self.pm.getNearestDocuments(newDocWithANearbySignature), set([1])
) # This assertion can sometimes fail because of randomization. Run the tests again. It's OK!
def test_getNearestDocument_emptyTrie(self):
permutationWithEmptyTrie = SignaturePermutationWithTrie(
signatureLength=self.signatureLength)
self.assertEqual(
permutationWithEmptyTrie.getNearestDocuments(self.doc1), set())
def test_removeDocument_documents(self):
newDocModifiedWithExistingSignature = Document(
3,
VectorGenerator.getRandomGaussianUnitVector(
dimension=self.dimension, mu=0, sigma=1))
newDocModifiedWithExistingSignature.signature = Signature(
self.doc1.signature.to01())
self.pm.addDocument(newDocModifiedWithExistingSignature)
self.assertEqual(
self.pm.signatureTrie[self.doc1.signature.permutate(
self.pm).to01()], set([1, 3]))
self.pm.removeDocument(newDocModifiedWithExistingSignature)
self.assertEqual(
self.pm.signatureTrie[self.doc1.signature.permutate(
self.pm).to01()], set([1]))
self.pm.removeDocument(self.doc1)
self.assertEqual(
None,
self.pm.signatureTrie.get(
self.doc1.signature.permutate(self.pm).to01()))
def test_resetSignatureTrie(self):
self.assertTrue(len(self.pm.signatureTrie) > 0)
self.pm.resetSignatureDataStructure()
self.assertTrue(len(self.pm.signatureTrie) == 0)
