def test_target_set(self):
'''
Tests Target set function
'''
# Test normal cases
target_example_int = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
predicted_sentiment = 0
target_example_int['predicted'] = 0
self.assertEqual(predicted_sentiment, target_example_int['predicted'],
msg='Predicted sentiment value should be 0 and '\
'not {}'.format(target_example_int['predicted']))
target_example_string = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 'pos')
predicted_sentiment = 'neg'
target_example_string['predicted'] = 'neg'
self.assertEqual(predicted_sentiment, target_example_string['predicted'],
msg='Predicted sentiment value should be `neg` and '\
'not {}'.format(target_example_string['predicted']))
# Testing the errors
with self.assertRaises(KeyError, msg='Should not allow you to set keys '\
'other than `predicted`'):
target_example_int['sentiment'] = 0
def test_targetcoll_get(self):
'''
Test the __getitem__ function of TargetCollection
'''
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 1)
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', 1)
target_col = TargetCollection([target_example_0, target_example_1,
target_example_2])
# Test normal case
self.assertEqual(target_example_2, target_col['2'], msg='Cannot access '\
'data using keys. key used {} collection {}'\
.format('2', target_col))
self.assertEqual(target_example_1, target_col.get('3'), msg='Cannot '\
'access data using the get method.')
self.assertEqual(None, target_col.get('5'), msg='Default value for '\
'get not working should be None not {}'\
.format(target_col.get('5')))
# Test that it raises a KeyError when key does not exist
with self.assertRaises(KeyError, msg='Should produce a key error when '\
'the data does not exist'):
target_col['5']
def test_targetcoll_add(self):
'''
Test the add function of TargetCollection
'''
target_col = TargetCollection()
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 1)
# Ensure the normal case works
target_col.add(target_example_0)
self.assertEqual(target_col['1'], target_example_0, msg='Test that {}' \
' has been added to {}'\
.format(target_example_0, target_col))
with self.assertRaises(TypeError, msg='Should not be able to add a dict'):
target_col.add({'target_id' : '2'})
with self.assertRaises(ValueError, msg='Should not be able to add a '\
'Target that has no `id`'):
del target_example_1['target_id']
if 'target_id' in target_example_1:
raise KeyError('{} should not contain `id` key'\
.format(target_example_1))
target_col.add(target_example_1)
def test_targetcoll_constructor(self):
'''
Tests TargetCollection constructor
'''
target_example = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
# Test that it will not accept anything but Target instances
with self.assertRaises(TypeError, msg='The constructor should only '\
'accept an interator as an argument'):
TargetCollection(1)
with self.assertRaises(TypeError, msg='The constructor should only '\
'accept an interator of Target instances'):
TargetCollection([1, 2, 3, 4])
# Should accept the following without any problems
TargetCollection([])
TargetCollection([target_example])
TargetCollection()
# Testing the case where the list of Targets contains duplicate keys
another_example = Target([(3, 4)], '2', 'Keys',
'text with Keys', -1)
dup_example = Target([(3, 10)], '1', 'Pixel',
'text with Pixels', 0)
with self.assertRaises(KeyError, msg='Should raise an error as two of '\
'the target instances have the same key'):
TargetCollection([target_example, another_example, dup_example])
def test_target_eq(self):
'''
Test the Target __eq__ method
'''
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 1)
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', 1)
target_example_dup0 = Target([(1, 2)], '1', 'S8',
'text with Samsung S8', 1)
# Targets with the same ID should be True even if they have different
# data
self.assertEqual(target_example_0, target_example_dup0,
msg='Should be equal as they have the same ID `1`')
# Targets with the same minimum keys should be True
target_example_dup1 = copy.deepcopy(target_example_1)
del target_example_dup1['target_id']
self.assertEqual(target_example_1, target_example_dup1, msg='Should be'\
' equal as they have the same minimum keys')
# Normal case
target_example_dup2 = copy.deepcopy(target_example_2)
self.assertEqual(target_example_2, target_example_dup2, msg='Copies of'\
' the same Target instance should be equal')
# Test that it won't accept dicts with the same minimum keys
dict_example_1 = {'target' : 'Iphone', 'sentiment' : 1,
'spans' : [(1, 5)], 'text' : 'text with Iphone'}
test_equality = dict_example_1 == target_example_1
self.assertEqual(False, test_equality, msg='Should not accept dicts '\
'even with the same minimum_keys')
def extract_targets(current_target, end_span, start_span, targets,
target_spans, target_index, tweet_text, sentiment_data,
tweet_id, target_sentiments):
if current_target != []:
target_word = ' '.join(current_target)
end_span = start_span + len(target_word)
targets.append(target_word)
target_spans.append((start_span, end_span))
start_span, end_span = None, None
current_target = []
target_index += 1
tweet_text = ' '.join(tweet_text)
for index, target in enumerate(targets):
target_id = '{}#{}'.format(tweet_id, index)
target_sentiment = target_sentiments[index]
target_span = target_spans[index]
if tweet_text[target_span[0] : target_span[1]] != target:
raise Exception('The target span {} does not match the '\
'target word {} in {}'\
.format(target_span, target, tweet_text))
target_data = {'spans' : [target_span], 'target_id' : target_id,
'target' : target, 'text' : tweet_text,
'sentiment' : target_sentiment,
'sentence_id' : tweet_id}
target_data = Target(**target_data)
sentiment_data.add(target_data)
return sentiment_data
def test_targetcoll_stored_sent(self):
'''
Test the stored_sentiments function of TargetCollection
'''
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 1)
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', -1)
target_col = TargetCollection([target_example_0, target_example_1,
target_example_2])
valid_sentiments = set([1, -1])
test_sentiments = target_col.stored_sentiments()
self.assertEqual(valid_sentiments, test_sentiments, msg='The unique '\
'sentiments in the TargetCollection should be {} and '\
'not {}'.format(valid_sentiments, test_sentiments))
def dong(file_path):
'''
Given file path to the
`Li Dong <https://github.com/bluemonk482/tdparse/tree/master/data/lidong>`_
sentiment data it will parse the data and return it as a list of dictionaries.
:param file_path: File Path to the annotated data
:type file_path: String
:returns: A TargetCollection containing Target instances.
:rtype: TargetCollection
'''
file_path = os.path.abspath(file_path)
if not os.path.isfile(file_path):
raise FileNotFoundError('This file does not exist {}'.format(file_path))
file_name, _ = os.path.splitext(os.path.basename(file_path))
sentiment_range = [-1, 0, 1]
sentiment_data = TargetCollection()
with open(file_path, 'r') as dong_file:
sent_dict = {}
for index, line in enumerate(dong_file):
divisible = index + 1
line = line.strip()
if divisible % 3 == 1:
sent_dict['text'] = line
elif divisible % 3 == 2:
sent_dict['target'] = line
elif divisible % 3 == 0:
sentiment = int(line)
if sentiment not in sentiment_range:
raise ValueError('The sentiment has to be one of the '\
'following values {} not {}'\
.format(sentiment_range, sentiment))
sent_dict['sentiment'] = int(line)
text = sent_dict['text'].lower()
target = sent_dict['target'].lower()
offsets = [match.span() for match in re.finditer(target, text)]
if len(target.split()) > 1:
joined_target = ''.join(target.split())
offsets.extend([match.span()
for match in re.finditer(joined_target, text)])
sent_dict['spans'] = offsets
sent_id = file_name + str(len(sentiment_data))
# Sentence ID is the same as the target as there is only one
# target per sentence
sent_dict['sentence_id'] = sent_id
sent_dict['target_id'] = sent_id
sent_target = Target(**sent_dict)
sentiment_data.add(sent_target)
sent_dict = {}
else:
raise Exception('Problem')
return sentiment_data
def test_targetcoll_add_preds(self):
'''
Tests the add_pred_sentiment function of TargetCollection
'''
target_example_int_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
target_example_int_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 1)
target_example_int_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', -1)
target_col_int = TargetCollection([target_example_int_0,
target_example_int_1,
target_example_int_2])
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 'pos')
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 'pos')
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', 'neg')
target_col_str = TargetCollection([target_example_0, target_example_1,
target_example_2])
pred_sents = [2, 5, 4]
target_col_int.add_pred_sentiment(pred_sents)
self.assertEqual(5, target_col_int['3']['predicted'], msg='Predicted '\
'sentiment not set correctly for `id` 3 should be 5 '\
'and not {}'.format(target_col_int))
with self.assertRaises(KeyError, msg='The original Target instances '\
'predicted sentiment should not be set as the '\
'TargetCollection should have copied them.'):
target_example_int_1['predicted']
pred_sents = ['neg', 'neu', 'pos']
target_col_str.add_pred_sentiment(pred_sents)
self.assertEqual('neu', target_col_str['3']['predicted'], msg='Predicted '\
'sentiment not set correctly for `id` 3 should be `neu` '\
'and not {}'.format(target_col_str))
pred_sents = [1, 0, -1]
target_col_str.add_pred_sentiment(pred_sents, mapper={1 : 'pos', 0 : 'neu',
-1 : 'neg'})
self.assertEqual('neu', target_col_str['3']['predicted'], msg='Predicted '\
'sentiment not set correctly for `id` 3 should be `neu` '\
'and not {} using the mapper'.format(target_col_str))
with self.assertRaises(KeyError, msg='The original Target instances '\
'predicted sentiment should not be set as the '\
'TargetCollection should have copied them.'):
target_example_1['predicted']
#with self.assertRaises(TypeError, msg='Should only accept list type '\
# 'not tuples'):
# target_col_int.add_pred_sentiment((2, 5, 4))
with self.assertRaises(ValueError, msg='Should accept lists that are '\
' the same size as the TargetCollection'):
target_col_int.add_pred_sentiment([1, 2, 3, 4])
def test_targetcoll_data(self):
'''
Test the data function of TargetCollection
'''
target_col = TargetCollection()
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 1)
target_col.add(target_example_0)
target_col.add(target_example_1)
all_data = target_col.data()
self.assertEqual(target_example_0, all_data[0], msg='First data '\
'returned should be the first inserted {} and not '\
'{}'.format(target_example_0, all_data[0]))
self.assertEqual(target_example_1, all_data[1], msg='Second data '\
'returned should be the second inserted {} and not '\
'{}'.format(target_example_1, all_data[1]))
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', 1)
del target_col['1']
target_col.add(target_example_2)
all_data = target_col.data()
self.assertEqual(target_example_1, all_data[0], msg='First data '\
'returned should be the second inserted {} and not '\
'{} as the first has been removed'\
.format(target_example_1, all_data[0]))
self.assertEqual(target_example_2, all_data[1], msg='Second data '\
'returned should be the third inserted {} and not '\
'{} as the first has been removed'\
.format(target_example_2, all_data[1]))
self.assertEqual(2, len(all_data), msg='The length of the data returned'\
'shoudl be 2 and not {}'.format(len(all_data)))
def generate_stats(data_path: Path) -> Dict[str, Union[int, float]]:
target_data = []
with data_path.open('r') as data_lines:
for line in data_lines:
line = json.loads(line)
line['spans'] = [tuple(span) for span in line['spans']]
target_data.append(Target(**line))
target_data = TargetCollection(target_data)
target_stats = defaultdict(lambda: 0)
data_size = len(target_data)
target_stats['size'] = data_size
for i in range(1, 3):
target_stats[f'Distinct sentiment {i}'] = len(
target_data.subset_by_sentiment(i))
for data in target_data.data_dict():
target_stats[data['sentiment']] += 1
for key, value in target_stats.items():
if key == 'size':
continue
target_stats[key] = value / data_size
return target_stats
def test_targetcoll_set(self):
'''
Test the __setitem__ function of TargetCollection
'''
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 1)
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', 1)
target_col = TargetCollection([target_example_0, target_example_1,
target_example_2])
target_example_3 = Target([(2, 4)], '5', 'new', 'new text', 0)
target_example_4 = Target([(1, 3)], '6', 'another', 'another text', 1)
target_example_5 = Target([(1, 3)], '7', 'another', 'another text', 1)
target_diff_1 = Target([(4, 5)], '3', 'test', 'test text', 0)
# Normal case adding a new value
target_col['5'] = target_example_3
self.assertEqual(target_col['5'], target_example_3, msg='Cannot add '\
'new value. store {} value added {}'\
.format(target_col, target_example_3))
# If key already exists it cannot be added
with self.assertRaises(KeyError, msg='Should not be able to add value '\
'{} as its key {} already exists {}'\
.format(target_diff_1, '3', target_col)):
target_col['3'] = target_diff_1
with self.assertRaises(KeyError, msg='Value with a different `id` to '\
'the key should fail. Key {} Value {}'\
.format('7', target_example_4)):
target_col['7'] = target_example_4
# Should accept Target instance with no `id`
del target_example_5['target_id']
if 'target_id' in target_example_5:
raise KeyError('{} should not contain `id` key'\
.format(target_example_5))
target_col['8'] = target_example_5
def test_target_normlisation(self):
'''
Tests target_normalisation
'''
with self.assertRaises(TypeError, msg='target_dict parameter has to '\
'be of type dict only'):
target_normalisation(['anything'])
test_values = [{'target_id':str(0),
'sentiment':-1,
'text':'This is a fake news articledd that is to represent a '\
'Tweet!!!! and it was an awful News Articless I think.',
'target':'news article',
'spans':[(15, 27), (83, 95)]},
{'target_id':str(1),
'sentiment':1,
'text':'I had a great ssDay however I did not get much '\
'work done in the days',
'target':'day',
'spans':[(16, 19), (64, 67)]},
{'target_id':str(2),
'sentiment':1,
'text':'I had a great ssDay however I did not get much '\
'work done in the days',
'target':'day',
'spans':[(16, 19)]},
{'target_id':str(3),
'sentiment':1,
'text':'Day however I did not get much done',
'target':'day',
'spans':[(0, 3)]},
{'target_id':str(4),
'sentiment':1,
'text':'however I did not get much done in the day',
'target':'day',
'spans':[(39, 42)]},
{'spans': [(47, 80)],
'target_id': '2',
'target': 'Core Processing Unit temperatures',
'text': 'Temperatures were ok but I was not tracking'\
' in Core Processing Unit temperatures.',
'sentiment': 0},
{'spans': [(1, 14), (15, 29)],
'target_id': '8',
'target': 'britney spears',
'text': "#britneyspears Britney Spears 's new single "\
"-3' debuts at #1: video: congratulations are "\
"in order ..",
'sentiment': 0}]
valid_results = [('This is a fake news_article dd that is to represent '\
'a Tweet!!!! and it was an awful news_article ss I '\
'think.', 'news_article'),
('I had a great ss day however I did not get much work'\
' done in the day s', 'day'),
('I had a great ss $day$ however I did not get much '\
'work done in the days', '$day$'),
('day however I did not get much done', 'day'),
('however I did not get much done in the day',
'day'),
('Temperatures were ok but I was not tracking in '\
'Core_ProcessingUnittemperatures .',
'Core_ProcessingUnittemperatures'),
("# britney_spears britney_spears 's new single "
"-3' debuts at #1: video: congratulations are in order ..",
'britney_spears')]
test_values = [Target(**test_value) for test_value in test_values]
for index, test_value in enumerate(test_values):
test_result = target_normalisation(test_value)
valid_result = valid_results[index]
self.assertEqual(valid_result, test_result, msg='Results is '\
'{} and should be {}. Test value {}'\
.format(test_result, valid_result, test_value))
similarity_field = np.array(target_data[similarity_field_name])
if is_language_model and threshold:
similarity_field = target_data['original_perplexity'] - similarity_field
above_original_perplexity_index = np.argmax((similarity_field <= 0) + 0)
similarity_field = similarity_field[:above_original_perplexity_index]
if len(similarity_field) < k_similar:
temp_k = len(similarity_field)
elif (not is_language_model) and threshold:
above_threshold_index = np.argmin((similarity_field >= threshold) + 0)
similarity_field = similarity_field[:above_threshold_index]
if len(similarity_field) < k_similar:
temp_k = len(similarity_field)
# For each of the filtered alternative targets it creates a json
# like object that will be used to store it in a collection to then
# save to a json file
alternative_targets = target_data['alternative_targets'][:temp_k]
for index, alternative_target in enumerate(alternative_targets):
new_target_id = f'{target_id}_{index}'
new_target_data = exchange_targets(target_data, alternative_target,
new_target_id)
# sanitizing the target dataset.
new_target_data.pop('alternative_targets')
new_target_data.pop(similarity_field_name)
if is_language_model:
new_target_data.pop('original_perplexity')
new_target_dataset.append(Target(**new_target_data))
print(f'Size of the expanded dataset {len(new_target_dataset)}')
new_target_dataset = TargetCollection(new_target_dataset)
new_target_dataset.to_json_file(str(args.dataset_save_fp),cache=False)
def _semeval_extract_data(sentences, file_name, conflict=False,
sentence_ids_skip=None):
'''
:param sentences: A `sentences` named element
:param file_name: Name of the file being parsed
:param conflict: Determine if to keep the target data that has a conflict \
sentiment label.
:param sentence_ids_skip: IDs of sentences that should be skipped
:type sentences: xml.etree.ElementTree.Element
:type file_name: String
:type conflict: bool. Defailt False
:type sentence_ids_skip: list. Default None
:returns: A TargetCollection containing Target instances.
:rtype: TargetCollection
'''
# Converts the sentiment tags from Strings to ints
sentiment_mapper = {'conflict' : -2, 'negative' : -1,
'neutral' : 0, 'positive' : 1}
def extract_aspect_terms(aspect_terms, sentence_id):
'''
:param aspect_terms: An aspectTerms element within the xml tree
:param sentence_id: Id of the sentence that the aspects came from.
:type aspect_terms: xml.etree.ElementTree.Element
:type sentence_id: String
:returns: A list of dictioanries containg id, span, sentiment and \
target
:rtype: list
'''
aspect_terms_data = []
for index, aspect_term in enumerate(aspect_terms):
aspect_term = aspect_term.attrib
aspect_term_data = {}
sentiment = sentiment_mapper[aspect_term['polarity']]
if sentiment == -2 and not conflict:
continue
aspect_id = '{}{}'.format(sentence_id, index)
aspect_term_data['target_id'] = aspect_id
if 'term' in aspect_term:
aspect_term_data['target'] = aspect_term['term']
elif 'target' in aspect_term:
aspect_term_data['target'] = aspect_term['target']
else:
raise KeyError('There is no `target` attribute in the opinions '\
'element {}'.format(aspect_term))
aspect_term_data['sentiment'] = sentiment
aspect_term_data['spans'] = [(int(aspect_term['from']),
int(aspect_term['to']))]
aspect_term_data['sentence_id'] = sentence_id
# If the target is NULL then there is no target
if aspect_term_data['target'] == 'NULL':
continue
aspect_terms_data.append(aspect_term_data)
return aspect_terms_data
def add_text(aspect_data, text):
'''
:param aspect_data: A list of dicts containing `span`, `target` and \
`sentiment` keys.
:param text: The text of the sentence that is associated to all of the \
aspects in the aspect_data list
:type aspect_data: list
:type text: String
:returns: The list of dicts in the aspect_data parameter but with a \
`text` key with the value that the text parameter contains
:rtype: list
'''
for data in aspect_data:
data['text'] = text
return aspect_data
all_aspect_term_data = TargetCollection()
for sentence in sentences:
aspect_term_data = None
text_index = None
sentence_id = file_name + sentence.attrib['id']
# Allow the parser to skip certain sentences
if sentence_ids_skip is not None:
if sentence.attrib['id'] in sentence_ids_skip:
continue
for index, data in enumerate(sentence):
if data.tag == 'sentence':
raise Exception(sentence.attrib['id'])
if data.tag == 'text':
text_index = index
elif data.tag == 'aspectTerms' or data.tag == 'Opinions':
aspect_term_data = extract_aspect_terms(data, sentence_id)
if aspect_term_data is None:
continue
if text_index is None:
raise ValueError('A semeval sentence should always have text '\
'semeval file {} sentence id {}'\
.format(file_name, sentence.attrib['id']))
sentence_text = sentence[text_index].text
aspect_term_data = add_text(aspect_term_data, sentence_text)
for aspect in aspect_term_data:
sent_target = Target(**aspect)
all_aspect_term_data.add(sent_target)
return all_aspect_term_data
def test_target_constructor(self):
'''
Test that target constructor
'''
# Testing the spans types
with self.assertRaises(TypeError, msg='Spans should be of type list'):
Target('span', '1', 'Iphone', 'text with Iphone', 'Pos')
with self.assertRaises(TypeError, msg='Spans should be list of tuples'):
Target([1, 2], '1', 'Iphone', 'text with Iphone', 'Pos')
with self.assertRaises(ValueError, msg='Spans should contain tuples of '\
'length 2'):
Target([(1, 2, 3), (3, 4, 5)], '1', 'Iphone', 'text with Iphone', 'Pos')
with self.assertRaises(ValueError, msg='Spans should contain tuples of '\
'length 2'):
Target([(1, 2), (3, 4, 5)], '1', 'Iphone', 'text with Iphone', 'Pos')
with self.assertRaises(TypeError, msg='Spans should contain tuples of '\
'length 2 and are Ints'):
Target([('1', '2')], '1', 'Iphone', 'text with Iphone', 'Pos')
with self.assertRaises(TypeError, msg='Spans should contain tuples of '\
'length 2 and are Ints'):
Target([(1, '2')], '1', 'Iphone', 'text with Iphone', 'Pos')
with self.assertRaises(TypeError, msg='Spans should contain tuples of '\
'length 2 and are Ints'):
Target([('1', 2)], '1', 'Iphone', 'text with Iphone', 'Pos')
with self.assertRaises(ValueError, msg='Spans should contain tuples of '\
'Ints where the first Int < second Int'):
Target([(7, 5)], '1', 'Iphone', 'text with Iphone', 'Pos')
with self.assertRaises(ValueError, msg='Spans should contain tuples of '\
'Ints where the first Int < second Int'):
Target([(3, 5), (6, 6)], '1', 'Iphone', 'text with Iphone', 'Pos')
# Testing that the spans work in a normal case
Target([(3, 5), (6, 8)], '1', 'Iphone', 'text with Iphone', 'Pos')
# Testing the target ID type
with self.assertRaises(TypeError, msg='Target ID should be a String'):
Target([(3, 5), (6, 8)], 1, 'Iphone', 'text with Iphone', 'Pos')
# Testing target type
with self.assertRaises(TypeError, msg='Target should be a String'):
Target([(3, 5), (6, 8)], '1', ('Iphone',), 'text with Iphone', 'Pos')
# Testing text type
with self.assertRaises(TypeError, msg='Text should be a String'):
Target([(3, 5), (6, 8)], '1', 'Iphone', ('text with Iphone',), 'Pos')
# Testing sentiment type
with self.assertRaises(TypeError, msg='Sentiment should be a String or '\
'Int'):
Target([(3, 5), (6, 8)], '1', 'Iphone', 'text with Iphone', ('Pos',))
# Testing the sentence_id type
with self.assertRaises(TypeError, msg='sentence_id should be a String'):
Target([(3, 5), (6, 8)], '1', 'Iphone', 'text with Iphone', 'pos',
sentence_id=1)
# Testing the sentiment type works as an Integer (Normal case)
span = [(3, 5)]
target = 'Iphone'
sentiment = 1
text = 'text with Iphone'
target_id = '210#1'
predicted = 0
sentence_id = '210'
target_example = Target(span, target_id, target, text, sentiment,
predicted, sentence_id)
# Testing that the dictionary mapping is correct
self.assertEqual(target_id, target_example['target_id'],
msg='The target ID should {} and not {}'\
.format(target_id, target_example['target_id']))
self.assertEqual(text, target_example['text'],
msg='The text should be {} and not {}'\
.format(text, target_example['text']))
self.assertEqual(sentiment, target_example['sentiment'],
msg='The sentiment should be {} and not {}'\
.format(sentiment, target_example['sentiment']))
self.assertEqual(target, target_example['target'],
msg='The target should be {} and not {}'\
.format(target, target_example['target']))
self.assertEqual(span, target_example['spans'],
msg='The spans should be {} and not {}'\
.format(span, target_example['spans']))
self.assertEqual(predicted, target_example['predicted'],
msg='The predicted sentiment should be {} and not {}'\
.format(predicted, target_example['predicted']))
self.assertEqual(sentence_id, target_example['sentence_id'],
msg='The sentence_id should be {} and not {}'\
.format(sentence_id, target_example['sentence_id']))
def hu_liu(file_path):
'''
Parser for the datasets from the following two papers (DOES NOT WORK):
1. `A Holistic Lexicon-Based Approach to Opinion Mining \
<https://www.cs.uic.edu/~liub/FBS/opinion-mining-final-WSDM.pdf>`_
2. `Mining and Summarizing Customer Reviews \
<https://www.cs.uic.edu/~liub/publications/kdd04-revSummary.pdf>`_
Currently this does not work. This is due to the dataset not containing
enough data to determine where the targets are in the text.
:param file_path: The path to a file containing annotations in the format \
of hu and liu sentiment datasets.
:type file_path: String
:returns: A TargetCollection containing Target instances.
:rtype: TargetCollection
'''
file_path = os.path.abspath(file_path)
file_name = os.path.basename(file_path)
sentiment_data = TargetCollection()
with open(file_path, 'r', encoding='cp1252') as annotations:
for sentence_index, annotation in enumerate(annotations):
# If it does not contain ## then not a sentence
if '##' not in annotation:
continue
targets_text = annotation.split('##')
if len(targets_text) > 2 or len(targets_text) < 1:
raise ValueError('The annotation {} when split on `##` should '\
'contain at least the sentence text and at'\
' most the text and the targets and not {}'\
.format(annotation, targets_text))
# If it just contains the sentence text then go to next
elif len(targets_text) == 1:
continue
elif targets_text[0].strip() == '':
continue
targets, text = targets_text
targets = targets.strip()
text = text.strip()
sentence_id = file_name + '#{}'.format(sentence_index)
targets = targets.split(',')
for target_index, target in enumerate(targets):
target = target.strip()
sentiment_match = re.search(r'\[[+-]\d\]$', target)
is_implicit = re.search(r'\[[up]\]', target)
if is_implicit:
print('Target {} is implicit {}'.format(target, text))
continue
if not sentiment_match:
raise ValueError('Target {} does not have a corresponding'\
' sentiment value. annotation {}'\
.format(target, annotation))
target_text = target[:sentiment_match.start()].strip()
sentiment_text = sentiment_match.group().strip().strip('[]')
sentiment_value = int(sentiment_text)
target_matches = list(re.finditer(target_text, text))
if len(target_matches) != 1:
print('The Target {} can only occur once in the '\
'text {}'.format(target_text, text))
continue
raise ValueError('The Target {} can only occur once in the '\
'text {}'.format(target_text, text))
target_span = target_matches[0].span()
target_id = sentence_id + '#{}'.format(target_index)
data_dict = {}
data_dict['spans'] = [target_span]
data_dict['target'] = target_text
data_dict['sentiment'] = sentiment_value
data_dict['text'] = text
data_dict['sentence_id'] = sentence_id
data_dict['target_id'] = target_id
sentiment_data.add(Target(**data_dict))
return sentiment_data
def parse_tweet(tweet_data, anno_data, tweet_id):
def get_offsets(entity, tweet_text, target):
offset_shifts = [0, -1, 1]
from_offset = entity['offset']
for offset_shift in offset_shifts:
from_offset_shift = from_offset + offset_shift
to_offset = from_offset_shift + len(target)
offsets = [(from_offset_shift, to_offset)]
offset_text = tweet_text[from_offset_shift : to_offset].lower()
if offset_text == target.lower():
return offsets
raise ValueError('Offset {} does not match target text {}. Full '\
'text {}\nid {}'\
.format(from_offset, target, tweet_text, tweet_id))
def fuzzy_target_match(tweet_text, target):
low_target = target.lower()
target_searches = [low_target, r'[^\w]' + low_target,
r'[^\w]' + low_target + r'[^\w]',
low_target + r'[^\w]',
low_target.replace(' ', ''),
low_target.replace(" '", '')]
for target_search in target_searches:
target_matches = list(re.finditer(target_search,
tweet_text.lower()))
if len(target_matches) == 1:
return target_matches
if tweet_id in set(['81211671026352128', '78689580104290305',
'81209490499960832']):
return None
if tweet_id == '75270720671973376' and target == 'kippers':
return None
if tweet_id == '65855178264686592' and target == 'tax':
return None
print(tweet_data)
print(anno_data)
raise ValueError('Cannot find the exact additional '\
'entity {} within the tweet {}'\
.format(target, tweet_text))
target_instances = []
tweet_id = str(tweet_id)
tweet_text = tweet_data['content']
target_ids = []
# Parse all of the entities that have been detected automatically
for entity in tweet_data['entities']:
data_dict = {}
target = entity['entity']
target_ids.append(entity['id'])
entity_id = str(entity['id'])
data_dict['spans'] = get_offsets(entity, tweet_text, target)
data_dict['target'] = entity['entity']
data_dict['target_id'] = folder_name + tweet_id + '#' + entity_id
data_dict['sentence_id'] = folder_name + tweet_id
data_dict['sentiment'] = anno_data['items'][entity_id]
if data_dict['sentiment'] == 'doesnotapply' and not include_dnr:
continue
# Convert from Strings to Integer
data_dict['sentiment'] = sentiment_mapper[data_dict['sentiment']]
data_dict['text'] = tweet_text
target_instances.append(Target(**data_dict))
# Parse all of the entities that have been selected by the user
if include_additional:
additional_data = anno_data['additional_items']
if isinstance(additional_data, dict):
for target, sentiment in additional_data.items():
target_matches = fuzzy_target_match(tweet_text, target)
if target_matches is None:
continue
target_id = max(target_ids) + 1
target_ids.append(target_id)
data_dict['spans'] = [target_matches[0].span()]
data_dict['target'] = target
data_dict['sentiment'] = sentiment
data_dict['text'] = tweet_text
data_dict['sentence_id'] = tweet_id
data_dict['target_id'] = tweet_id + '#' + str(target_id)
target_instances.append(Target(**data_dict))
return target_instances
def test_dependency_context(self):
'''
Tests dependency_context
'''
# Test the multiple span cases
test_values = [{'target_id':str(0),
'sentiment':-1,
'text':'This is a fake news articledd that is to represent a '\
'Tweet!!!! and it was an awful News Articless I think.',
'target':'news article',
'spans':[(15, 27), (83, 95)]},
{'target_id':str(1),
'sentiment':1,
'text':'I had a great Day however I did not get much '\
'work done in the days',
'target':'day',
'spans':[(14, 17), (62, 65)]},
{'spans': [(1, 14), (15, 29)],
'target_id': '8',
'target': 'britney spears',
'text': "#britneyspears Britney Spears 's new single "\
"-3' debuts at #1: video: congratulations are "\
"in order ..",
'sentiment': 0},
{'spans': [(5, 14)],
'target_id': '9',
'target': 'wait time',
'text': "less wait time for me!",
'sentiment': 0},
{'spans': [(81, 91)],
'target_id': '9',
'target': '@RichardWS',
'text': "Lower taxes rendered null & void with the "\
"planned increase in VAT by the tories. "\
"@RichardWS #battlefornumber10 @RichardWS",
'sentiment': 0},
{'spans': [(46, 57)],
'target_id': '9',
'target': '@Shaun_XL5:',
'text': "< My main #GE2015 policy worry is now defence "\
"@Shaun_XL5: I think we need Nato target of 2% "\
"GDP with hardcore auditing @paullewismoney",
'sentiment': 0},
{'spans': [(102, 119)],
'target_id': '9',
'target': 'first-time buyers',
'text': "Despite spin, no new government money for "\
"#housing: transferring resources from affordable"\
" renting to first-time buyers. #GE2015 #GE15",
'sentiment': 0},
{'spans': [(130, 140)],
'target_id': '9',
'target': '@UNICEF_uk!',
'text': "Violence against children around the world is "\
"like a hidden epidemic. Michael Sheen talking "\
"well abt SDGs on #marrshow Nice work @UNICEF_uk!",
'sentiment': 0},
{'spans': [(21, 32)],
'target_id': '9',
'target': 'Tony #Blair',
'text': "Corrupt war criminal Tony #Blair's rats are "\
"running amok inside Labour today. If I was "\
"thinking of voting Labour today. I'd stop. #GE15",
'sentiment': 0},
{'spans': [(39, 49)],
'target_id': '9',
'target': '@UKLabour;',
'text': "Day 1 of #GE2015 & I'm already sick of "\
"@UKLabour; their lies & their forgetfulness. "\
"Whose policies created the need for austerity?"\
" Yes. Lab!",
'sentiment': 0},
{'spans': [(33, 46)],
'target_id': '9',
'target': '@GrantShapps’',
'text': "Lest Cameron forgets: Police say @GrantShapps’ "\
"firm sales “may constitute offence of fraud” "\
"#bbcdp #pmqs : http://t.co/Gu9Ke6sRtX",
'sentiment': 0}]
valid_results = [[{'text' : 'This is a fake news article',
'span' : (15, 27)},
{'text' : 'dd that is to represent a Tweet and it was '\
'an awful news article',
'span' : (52, 64)}],
[{'text' : 'I had a great day however I did not get '\
'much work done in the day',
'span' : (14, 17)},
{'text' : 'I had a great day however I did not get '\
'much work done in the day',
'span' : (62, 65)}],
[{'text' : "britney spears britney spears",
'span' : (0, 14)},
{'text' : "britney spears britney spears",
'span' : (15, 29)}],
[{'text' : 'less wait time for me',
'span' : (5, 14)}],
[{'text' : '@RichardWS',
'span' : (0, 10)}],
[{'text' : "My main #GE2015 policy worry is now defence"\
" @Shaun_XL5:", 'span' : (44, 55)}],
[{'text' : "transferring resources from affordable "\
"renting to first-time buyers",
'span' : (50, 67)}],
[{'text' : "@UNICEF_uk!", 'span' : (0, 11)}],
[{'text' : "criminal Tony #Blair",
'span' : (9, 20)}],
[{'text' : "Day 1 of #GE2015 & I'm already sick of "\
"@UKLabour; their lies & their forgetfulness",
'span' : (39, 49)}],
[{'text' : "Lest Cameron forgets Police say "\
"@GrantShapps’ firm sales",
'span' : (32, 45)}]]
test_values = [Target(**test_value) for test_value in test_values]
test_results = dependency_context(test_values, tweebo)
for index, valid_result in enumerate(valid_results):
test_result = test_results[index]
for dict_index, valid_dict in enumerate(valid_result):
test_dict = test_result[dict_index]
self.assertEqual(valid_dict['text'], test_dict['text'],
msg='texts are different correct `{}` test `{}`'\
.format(valid_dict['text'], test_dict['text']))
self.assertEqual(valid_dict['span'], test_dict['span'],
msg='spans are different correct `{}` test `{}`'\
' text `{}`'.format(valid_dict['span'],
test_dict['span'],
test_dict['text']))
# Test the lower casing case of the text and if the target is upper case
test_values = [{'target_id':str(0),
'sentiment':-1,
'text':'This is a fake news articledd that is to represent a '\
'Tweet!!!! and it was an awful News Articless I think.',
'target':'news article',
'spans':[(15, 27), (83, 95)]},
{'target_id':str(0),
'sentiment':-1,
'text':'This is a fake news articledd that is to represent a '\
'Tweet!!!! and it was an awful News Articless I think.',
'target':'News Article',
'spans':[(15, 27), (83, 95)]},
{'target_id':str(1),
'sentiment':1,
'text':'I had a great Day however I did not get much '\
'work done in the days',
'target':'day',
'spans':[(14, 17), (62, 65)]}]
valid_results = [[{'text' : 'this is a fake news article',
'span' : (15, 27)},
{'text' : 'dd that is to represent a tweet and it was '\
'an awful news article',
'span' : (52, 64)}],
[{'text' : 'this is a fake news article',
'span' : (15, 27)},
{'text' : 'dd that is to represent a tweet and it was '\
'an awful news article',
'span' : (52, 64)}],
[{'text' : 'i had a great day however i did not get '\
'much work done in the day',
'span' : (14, 17)},
{'text' : 'i had a great day however i did not get '\
'much work done in the day',
'span' : (62, 65)}]]
test_values = [Target(**test_value) for test_value in test_values]
test_results = dependency_context(test_values, tweebo, lower=True)
for index, valid_result in enumerate(valid_results):
test_result = test_results[index]
for dict_index, valid_dict in enumerate(valid_result):
test_dict = test_result[dict_index]
self.assertEqual(valid_dict['text'], test_dict['text'],
msg='texts are different correct `{}` test `{}`'\
.format(valid_dict['text'], test_dict['text']))
self.assertEqual(valid_dict['span'], test_dict['span'],
msg='spans are different correct `{}` test `{}`'\
.format(valid_dict['span'], test_dict['span']))
# Test the case where the target is mentioned twice but only 1 is relevant
# to one of the mentions
test_values = [{'target_id':str(1),
'sentiment':1,
'text':'I had a great Day however I did not get much '\
'work done in the day',
'target':'day',
'spans':[(14, 17)]},
{'target_id':str(3),
'sentiment':1,
'text':'I had a great Day however I did not get much '\
'work done in the Day',
'target':'day',
'spans':[(14, 17)]},
{'spans': [(47, 80)],
'target_id': '2',
'target': 'Core Processing Unit temperatures',
'text': 'Temperatures were ok but I was not tracking'\
' in Core Processing Unit temperatures.',
'sentiment': 0},
{'text' : 'I also recommend the rice dishes or the '\
'different varieties of congee (rice porridge).',
'spans' : [(63, 85)],
'target_id' : '4',
'target' : 'congee (rice porridge)',
'sentiment' : 1}]
valid_results = [[{'text' : 'I had a great day however I did not get '\
'much work done in the day',
'span' : (14, 17)}],
[{'text' : 'I had a great day however I did not get '\
'much work done in the Day',
'span' : (14, 17)}],
[{'text' : 'Temperatures were ok but I was not tracking'\
' in Core Processing Unit temperatures',
'span' : (47, 80)}],
[{'text' : 'I also recommend the rice dishes or the '\
'different varieties of congee (rice '\
'porridge)',
'span' : (63, 85)}]]
test_values = [Target(**test_value) for test_value in test_values]
test_results = dependency_context(test_values, tweebo)
for index, valid_result in enumerate(valid_results):
test_result = test_results[index]
for dict_index, valid_dict in enumerate(valid_result):
test_dict = test_result[dict_index]
self.assertEqual(valid_dict['text'], test_dict['text'],
msg='texts are different correct `{}` test `{}`'\
.format(valid_dict['text'], test_dict['text']))
self.assertEqual(valid_dict['span'], test_dict['span'],
msg='spans are different correct `{}` test `{}`'\
.format(valid_dict['span'], test_dict['span']))
def test_targetcoll_sent_data(self):
'''
Test the sentiment_data function of TargetCollection
'''
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 1)
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 1)
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', -1)
target_col_int = TargetCollection([target_example_0, target_example_1,
target_example_2])
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 'pos')
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 'pos')
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', 'neg')
target_col_str = TargetCollection([target_example_0, target_example_1,
target_example_2])
# Testing the basic example
test_sentiments = target_col_int.sentiment_data()
valid_sentiments = [1, 1, -1]
self.assertEqual(valid_sentiments, test_sentiments, msg='The Integer '\
'sentiments returned should be {} and not {}'\
.format(valid_sentiments, test_sentiments))
test_sentiments = target_col_str.sentiment_data()
valid_sentiments = ['pos', 'pos', 'neg']
self.assertEqual(valid_sentiments, test_sentiments, msg='The String '\
'sentiments returned should be {} and not {}'\
.format(valid_sentiments, test_sentiments))
# Testing the mapping function
str_mapper = {'pos' : 1, 'neg' : -1}
test_sentiments = target_col_str.sentiment_data(mapper=str_mapper)
valid_sentiments = [1, 1, -1]
self.assertEqual(valid_sentiments, test_sentiments, msg='The String '\
'sentiments should be mapped to Integers. Valid {} '\
'not {}'.format(valid_sentiments, test_sentiments))
int_mapper = {1 : 'pos', -1 : 'neg'}
test_sentiments = target_col_int.sentiment_data(mapper=int_mapper)
valid_sentiments = ['pos', 'pos', 'neg']
self.assertEqual(valid_sentiments, test_sentiments, msg='The Integer '\
'sentiments should be mapped to String. Valid {} '\
'not {}'.format(valid_sentiments, test_sentiments))
with self.assertRaises(TypeError, msg='Should only accept dict mapper'):
target_col_int.sentiment_data(mapper=[(1, 'pos'), (-1, 'neg')])
with self.assertRaises(ValueError, msg='Mapper should refuse dicts that'\
' may have valid mappings but not all the mappings'):
target_col_int.sentiment_data(mapper={1 : 'pos'})
with self.assertRaises(ValueError, msg='Mapper should refuse dicts that'\
' contain the correct number of mappings but not '\
'the correct mappings'):
target_col_int.sentiment_data({0 : 'pos', -1 : 'neg'})
with self.assertRaises(ValueError, msg='Mapper should refuse dicts that '\
'have all the correct mappings but contain some '\
'in-correct mappings'):
target_col_int.sentiment_data(mapper={1 : 'pos', -1 : 'neg',
0 : 'neu'})
# Testing the sentiment_field parameter
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 'pos', 'neg')
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 'pos', 'neu')
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', 'neg', 'pos')
target_col = TargetCollection([target_example_0, target_example_1,
target_example_2])
test_sentiments = target_col.sentiment_data(sentiment_field='predicted')
valid_sentiments = ['neg', 'neu', 'pos']
self.assertEqual(valid_sentiments, test_sentiments, msg='The predicted '\
'sentiments returned should be {} and not {}'\
.format(valid_sentiments, test_sentiments))
def test_dependency_relation_context(self):
'''
Tests dependency_relation_context
'''
# Test the normalise case
test_values = [{'target_id':str(0),
'sentiment':-1,
'text':'This is a fake news articledd that is to represent a '\
'Tweet!!!! and it was an awful News Articless I think.',
'target':'news article',
'spans':[(15, 27), (83, 95)]},
{'target_id':str(1),
'sentiment':1,
'text':'I had a great Day however I did not get much '\
'work done in the days',
'target':'day',
'spans':[(14, 17), (62, 65)]},
{'target_id':str(2),
'sentiment':1,
'text':'this is an alternative sentence to see how it '\
'performs',
'target':'sentence',
'spans':[(23, 31)]},
{'target_id':str(3),
'sentiment':1,
'text':'Teen pop star heartthrobe is all the rage on '\
'social media',
'target':'teen',
'spans':[(0, 4)]}]
valid_results = [['a fake', 'an awful'], ['a great', 'the'],
['an alternative to see performs it how'], ['']]
test_values = [Target(**test_value) for test_value in test_values]
test_results = dependency_relation_context(test_values,
tweebo,
n_relations=(1, -1))
for index, valid_result in enumerate(valid_results):
test_result = test_results[index]
self.assertEqual(valid_result, test_result, msg='Incorrect context'\
' correct {} test {}'.format(valid_result, test_result))
# Testing when we only want the first dependency relation
valid_results = [['a fake', 'an awful'], ['a great', 'the'],
['an alternative to'], ['']]
test_results = dependency_relation_context(test_values, tweebo)
for index, valid_result in enumerate(valid_results):
test_result = test_results[index]
self.assertEqual(valid_result, test_result, msg='Incorrect context'\
' correct {} test {}'.format(valid_result, test_result))
# Testing to ensure it will lower case the words before processing
valid_results = [['a fake', 'an awful'], ['a great', 'the'],
['an alternative to see performs it how'], ['']]
test_results = dependency_relation_context(test_values, tweebo, True,
(1, -1))
for index, valid_result in enumerate(valid_results):
test_result = test_results[index]
self.assertEqual(valid_result, test_result, msg='Incorrect context'\
' correct {} test {}'.format(valid_result, test_result))
# Testing for when a sentence mentions the target more than once but we
# are only interested in the first mention
test_values = [{'target_id':str(1),
'sentiment':1,
'text':'I had a great Day however I did not get much '\
'work done in the day',
'target':'day',
'spans':[(14, 17)]}]
valid_results = [['a great']]
test_values = [Target(**test_value) for test_value in test_values]
test_results = dependency_relation_context(test_values,
tweebo,
n_relations=(1, -1))
for index, valid_result in enumerate(valid_results):
test_result = test_results[index]
self.assertEqual(valid_result, test_result, msg='Incorrect context'\
' for more than one mention correct {} test {}'\
.format(valid_result, test_result))
def test_target_coll_subset_by_sent(self):
'''
Test the subset_by_sentiment function of TargetCollection
'''
target_example_0 = Target([(3, 5), (6, 8)], '1', 'Iphone',
'text with Iphone', 'pos', sentence_id='4')
target_example_1 = Target([(1, 5)], '3', 'Iphone',
'text with Iphone', 'neg', sentence_id='4')
target_example_2 = Target([(1, 2)], '2', 'Iphone',
'text with Iphone', 'neg', sentence_id='5')
target_example_3 = Target([(1, 2)], '4', 'Iphone',
'text with Iphone', 'neg', sentence_id='5')
target_example_4 = Target([(1, 2)], '5', 'Iphone',
'text with Iphone', 'pos', sentence_id='6')
target_example_5 = Target([(1, 2)], '6', 'Iphone',
'text with Iphone', 'neu', sentence_id='6')
target_example_6 = Target([(1, 2)], '7', 'Iphone',
'text with Iphone', 'neg', sentence_id='6')
target_example_7 = Target([(1, 2)], '8', 'Iphone',
'text with Iphone', 'neg', sentence_id='7')
target_example_8 = Target([(1, 2)], '9', 'Iphone',
'text with Iphone', 'neg', sentence_id='8')
target_example_9 = Target([(1, 2)], '10', 'Iphone',
'text with Iphone', 'neg', sentence_id='8')
target_example_10 = Target([(1, 2)], '11', 'Iphone',
'text with Iphone', 'pos', sentence_id='8')
all_targets = [target_example_0, target_example_1, target_example_2,
target_example_3, target_example_4, target_example_5,
target_example_6, target_example_7, target_example_8,
target_example_9, target_example_10]
target_col = TargetCollection(all_targets)
# Test for 2 unique sentiments per sentence
test_col = target_col.subset_by_sentiment(2)
valid_col = TargetCollection([target_example_0, target_example_1,
target_example_8, target_example_9,
target_example_10])
self.assertEqual(valid_col, test_col, msg='Should only return these {}'\
' but has returned this {}'.format(valid_col, test_col))
# Test for 1 unique sentiments per sentence
test_col = target_col.subset_by_sentiment(1)
valid_col = TargetCollection([target_example_7, target_example_2,
target_example_3])
self.assertEqual(valid_col, test_col, msg='Should only return these {}'\
' but has returned this {}'.format(valid_col, test_col))
# Test for 3 unique sentiments per sentence
test_col = target_col.subset_by_sentiment(3)
valid_col = TargetCollection([target_example_4, target_example_5,
target_example_6])
self.assertEqual(valid_col, test_col, msg='Should only return these {}'\
' but has returned this {}'.format(valid_col, test_col))
