def create_tri_model(span_eng_dict):
trigram_span_dict = collections.defaultdict(lambda: 0)
trigram_eng_dict = collections.defaultdict(lambda: 0)
trigram_span_eng_dict = collections.defaultdict(lambda: tuple)
text = codecs.open('SpanishText.txt', encoding='utf-8')
for sentence in text.readlines():
line = [re.sub('[.?!",]', '', word) for word in sentence.split()]
for word1, word2, word3 in trigrams(line):
trigram_span_dict[(word1.lower(), word2.lower(),
word3.lower())] += 1
for word1, word2, word3 in trigrams(line):
#print(word1, span_eng_dict[word1])
trigram_span_eng_dict[(
word1.lower(), word2.lower(),
word3.lower())] = (span_eng_dict[word1.lower()],
span_eng_dict[word2.lower()],
span_eng_dict[word3.lower()])
eng_text = open('DMT_output.txt')
for sentence in eng_text.readlines():
line = [re.sub('[.?!",]', '', word) for word in sentence.split()]
for word1, word2, word3 in trigrams(line):
trigram_eng_dict[(word1.lower(), word2.lower(),
word3.lower())] += 1
text = ''
for k, v in trigram_span_eng_dict.items():
try:
if (trigram_span_dict.get(k) == trigram_eng_dict.get(v)) and (
trigram_span_dict.get(k) >= 1):
#print(k, v)
text += k[0] + k[1]
except:
pass
return
def score_by_topic(pkg, scores):
'''Examines the pkg and adds scores according to topics in it.'''
themes = Themes.instance()
for level in range(3):
pkg_text = package_text(pkg, level)
words, words_without_stopwords = normalize_text(pkg_text)
for num_words in (1, 2, 3):
if num_words == 1:
ngrams = words_without_stopwords
topic_ngrams = themes.topic_words
topic_ngrams_set = themes.topic_words_set
elif num_words == 2:
ngrams = bigrams(words)
topic_ngrams = themes.topic_bigrams
topic_ngrams_set = themes.topic_bigrams_set
elif num_words == 3:
ngrams = trigrams(words)
topic_ngrams = themes.topic_trigrams
topic_ngrams_set = themes.topic_trigrams_set
matching_ngrams = set(ngrams) & topic_ngrams_set
if matching_ngrams:
for ngram in matching_ngrams:
occurrences = ngrams.count(ngram)
score = (3-level) * occurrences * num_words
theme = topic_ngrams[ngram]
ngram_printable = ' '.join(ngram) if isinstance(ngram, tuple) else ngram
reason = '"%s" matched %s' % (ngram_printable, LEVELS[level])
if occurrences > 1:
reason += ' (%s times)' % occurrences
scores[theme].append((score, reason))
log.debug(' %s %s %s', theme, score, reason)
def act(self):
"""
Add words in the last observation to the dictionary.
This checks any fields in the message present in the --dict-textfields
argument (e.g. "text,labels").
"""
for textfield in self.textfields:
source = self.observation.get(textfield)
if source is None:
continue
# fields may be singleton strings or lists of strings.
# wrap the singleton strings in a list to iterate over them
if type(source) is str:
source = [source]
for text in source:
if text:
tokens = self.tokenize(text)
self.add_to_dict(tokens)
unigram_ = nltk.ngrams(tokens, 1)
bigrams_ = bigrams(tokens)
trigrams_ = trigrams(tokens)
self.unigram_freq.update(unigram_)
self.bigram_freq.update(bigrams_)
self.trigram_freq.update(trigrams_)
return {'id': 'Dictionary'}
def trigram_format( test_corpus ):
"""
>>> trigram_format(["the dog runs STOP", "the cat walks STOP", "the dog runs STOP"])
[[('the', 'dog', 'runs'), ('dog', 'runs', 'STOP')], [('the', 'cat', 'walks'), ('cat', 'walks', 'STOP')], [('the', 'dog', 'runs'), ('dog', 'runs', 'STOP')]]
"""
wl = [ [word for word in sentence.split()] for sentence in test_corpus]
return [ util.trigrams( l ) for l in wl ]
def generate_unibitrigrams(key_score_file):
with open(key_score_file, 'rb') as infile:
infile.readline()
key_list = list()
for line in infile:
row = list(line.split(','))
key_list.append(row[0])
uni_bi_trigrams = []
for phrase in key_list:
words = []
unigrams_ls = []
bigrams_ls = []
trigrams_ls = []
for word in nltk.word_tokenize(phrase):
word = re.sub('[!"#$%&\'\(\)*+,-./:;<=>?@[\]\^_`{|}~]', '', word)
words.append(word)
unigrams_ls = words
#bigrams_ls=list(bigrams(words))
for x in list(bigrams(words)):
bigrams_ls.append(x[0] + ' ' + x[1])
for x in list(trigrams(words)):
trigrams_ls.append(x[0] + ' ' + x[1] + ' ' + x[2])
#trigrams_ls=list(trigrams(words))
uni_bi_trigrams = uni_bi_trigrams + unigrams_ls + bigrams_ls + trigrams_ls
return uni_bi_trigrams
def ngrams(self, gram_size=3):
"""Gives ngrams.
Returns a list of ngrams, each ngram represented as a tuple.
Args:
gram_size (:obj:`int`, optional) Size of the ngrams to generate
Returns:
:obj:`list` of :obj:`tuple` Words of each ngram
Example:
>>> text = EnglishText('They hated to think of sample sentences.')
>>> basic_ngrams = text.ngrams()
>>> print(basic_ngrams)
[('They', 'hated', 'to'), ('hated', 'to', 'think'), ('to', 'think', 'of'), ('think', 'of', 'sample'), ('of', 'sample', 'sentences'), ('sample', 'sentences', '.')]
""" # noqa
tokens = self.tokenize()
if gram_size < 2: # pragma: no cover
gram_size = 2
if gram_size == 2: # pragma: no cover
return list(bigrams(tokens))
if gram_size == 3:
return list(trigrams(tokens))
else: # pragma: no cover
return list(ngrams(tokens, gram_size))
def pos_tags(vocab_hash, sentence):
sentence = sentence.split()
unigram_hash = get_pos(vocab_hash, sentence)
bigram_hash = get_pos(vocab_hash, bigrams(sentence))
trigram_hash = get_pos(vocab_hash, trigrams(sentence))
pos_tags = []
ngram_to_tag = {}
ngram_ordering = []
for i in xrange(len(sentence)):
word = sentence[i]
if unigram_hash.has_key(word):
tag = unigram_hash[word]
pos_tags.append(tag)
ngram_to_tag[word] = tag
ngram_ordering.append(word)
elif i < len(sentence) - 1:
bigram = sentence[i] + " " + sentence[i + 1]
if bigram_hash.has_key(bigram):
tag = bigram_hash[bigram]
pos_tags.append(tag)
ngram_to_tag[bigram] = tag
ngram_ordering.append(bigram)
i += 1
elif (i < len(sentence) - 2):
trigram = " ".join(sentence[i:i + 2])
if trigram_hash.has_key(trigram):
tag = trigram_hash[trigram]
pos_tags.append(tag)
ngram_to_tag[trigram] = tag
ngram_ordering.append(trigram)
i += 2
return pos_tags, ngram_to_tag, ngram_ordering
def __init__(self, index: int, sent: str, start: int, end: int):
self.index = index
self.sent = sent
self.words = self.sentToWords()
self.nGrams = list(trigrams(self.words))
self.start = start
self.end = end
def text():
testo = [x for x in request.form.values()]
input_text = testo[0]
input_text = re.sub(r'\n', '' , input_text)
input_text = re.sub(r'\t', '' , input_text)
tok_test = word_tokenize(input_text.lower())
trig = list(trigrams(tok_test))
vocab_file = pd.read_csv('DIZIONARIO.csv', sep=';')
vocab = list(vocab_file['TOKEN'])
model = pickle.load(open('langmod.pickle', 'rb'))
err = 0
for t in trig:
if(t[2] not in vocab):
err +=1
continue
if(model.score(t[2],[t[0],t[1]])==0):
err +=1
if(err<=2):
punteggio = 'Sembra che non ci siano errori ortografici tipici della disortografia :-)'
elif(err>2 and err<=5):
punteggio = 'Forse c\'è qualche errore ortografico che potrebbe far pensare alla presenza di disortografia. Controlla le doppie, lo scambio, l\'inserimento o la traslazione di lettere/sillabe, l\'H nei verbi e gli accenti!'
elif(err>5):
punteggio = 'Gli errori ortografici sembrano essere un po\' tantini e tipici della disortografia! Non preoccuparti, respira e - dopo aver controllato altri testi - consulta uno specialista :-)'
return render_template('elements.html',score=punteggio)
def generate_unibitrigrams(key_score_file):
with open(key_score_file,'rb') as infile:
infile.readline()
key_list=list()
for line in infile:
row=list(line.split(','))
key_list.append(row[0])
uni_bi_trigrams=[]
for phrase in key_list:
words=[]
unigrams_ls=[]
bigrams_ls=[]
trigrams_ls=[]
for word in nltk.word_tokenize(phrase):
word=re.sub('[!"#$%&\'\(\)*+,-./:;<=>?@[\]\^_`{|}~]','',word)
words.append(word)
unigrams_ls=words
#bigrams_ls=list(bigrams(words))
for x in list(bigrams(words)):
bigrams_ls.append(x[0]+' '+x[1] )
for x in list(trigrams(words)):
trigrams_ls.append(x[0]+' '+x[1]+' '+x[2] )
#trigrams_ls=list(trigrams(words))
uni_bi_trigrams=uni_bi_trigrams+unigrams_ls+bigrams_ls+trigrams_ls
return uni_bi_trigrams
def __init__(self, index: int, sent: str, start: int, end: int, lang: int):
self.lang = LangDiff(lang)
self.index = index
self.sent = sent
self.words = self.remove_puncts(self.lang.word_tokenize(sent))
self.nGrams = Counter(trigrams(self.sent_to_words()))
self.start = start
self.end = end
def pre_trigram(texto):
lista = []
for x in texto:
if len(x) < 4:
lista.append((x))
else:
lista.append(tuple(trigrams(x)))
return lista
def wordsToTrigramsWithIndices(self, dictionary):
def getIndexedTuple(word: str):
index = -1
if word in dictionary.wordsToIndices:
index = dictionary.wordsToIndices[word]
return (index, word)
return list(trigrams(list(map(getIndexedTuple, self.words))))
def brown_trigrams(category):
"""Takes as input the name of a brown category, and returns a list of all of the trigrams in the category."""
words = ["<s>"]
words += [
word.lower() for word in brown.words(categories=category)
if word.isalnum()
]
words.append("</s>")
return list(trigrams(words))
def clean_up_sentence(sentence):
# tokenize the pattern
sentence_words = []
w = list(trigrams(sentence))
for x in w:
sentence_words.append(x[0] + x[1] + x[2])
# stem each word
return sentence_words
def getNgrams(self):
"""Get ngrams from the question. Right now only bigrams and trigrams are supported"""
bigram_str = [
bigram[0] + ' ' + bigram[1] for bigram in bigrams(self.tokens)
]
trigram_str = [
trigram[0] + ' ' + trigram[1] + ' ' + trigram[2]
for trigram in trigrams(self.tokens)
]
return (bigram_str, trigram_str)
def ngrams(self, gram_size=3):
tokens = self.tokenize()
if gram_size < 2: # pragma: no cover
gram_size = 2
if gram_size == 2: # pragma: no cover
return list(bigrams(tokens))
if gram_size == 3:
return list(trigrams(tokens))
else: # pragma: no cover
return list(ngrams(tokens, gram_size))
def main():
save_data_from_webpage()
text = get_data_from_file()
#creates a list of the tolkenized words
tt = word_tokenize(text)
pprint(tt)
#creates a new list for the steam words using all of the stemmers
psteam = PorterStemmer()
psteam_list = []
for word in tt:
psteam_list.append(psteam.stem(word))
pprint(psteam_list)
lsteam = LancasterStemmer()
lsteam_list = []
for word in tt:
lsteam_list.append(lsteam.stem(word))
pprint(lsteam_list)
ssteam = SnowballStemmer()
ssteam_list = []
for word in tt:
ssteam_list.append(ssteam.stem(word))
pprint(ssteam_list)
p = set(psteam_list)
l = set(lsteam_list)
s = set(ssteam_list)
#displays the different steams
pprint(s.difference(l.difference(p)))
#pos taging
pos_list = pos_tag(text)
pprint(pos_list)
#creates a new list for the lematized words
lemmatizer = WordNetLemmatizer()
lem = []
for word in tt:
lem.append(lemmatizer.lemmatize(word))
#pprint(lem)
# returns a generator of trigrams using the tokenized list tt
trig = trigrams(tt)
displays the results
print(list(trig))
#ne_chunck finds non overlapping groups
#pos_tag ids how the text is used in speech
NamedEntity = ne_chunk(pos_tag(wordpunct_tokenize(text)))
print(NamedEntity)
def _get_filtered_trigrams(self, words):
# Allow stopword in the middle of trigram
filtered_trigrams = []
for tri in trigrams(words):
leave = True
for i, w in enumerate(tri):
if w in stopwords and i != 1:
leave = False
break
if leave and tri[0] != tri[1] and tri[1] != tri[2]:
filtered_trigrams.append(tri)
return filtered_trigrams
def __init__(self, geo_locations):
'''Initializes the language model by creating the ConditionalFreqDist
and ConditionalProbDist'''
words_count = 0
# will contain all names in a list which preserves their frequenceis as
# they appear in the gazetteer. The frequenceis are going to be used in
# the language model.
gaz_n_grams = list()
self.unigrams = defaultdict(int)
for ln in geo_locations:
number_of_mentions = len(geo_locations[ln])
n_gram = ln.split()
new_list = [n_gram] * number_of_mentions
gaz_n_grams.extend(new_list)
for token in n_gram:
words_count += 1
self.unigrams[token] += 1
self.unigrams = {"words": self.unigrams, "words_count": words_count}
# bigrams +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
train_bigrams = list(chain(*[bigrams(i) for i in gaz_n_grams]))
cfd_bigrams = ConditionalFreqDist()
for bg in train_bigrams:
cfd_bigrams[bg[0]][bg[1]] += 1
# bigrams MLE probabilities
self.cpd_bigrams = ConditionalProbDist(cfd_bigrams, nltk.MLEProbDist)
# trigrams ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
train_trigrams = list(chain(*[trigrams(i) for i in gaz_n_grams]))
cfd_trigrams = ConditionalFreqDist()
for bg in train_trigrams:
bi_gr = " ".join(bg[:-1])
cfd_trigrams[bi_gr][bg[2]] += 1
# trigrams MLE probabilities
self.cpd_trigrams = ConditionalProbDist(cfd_trigrams, nltk.MLEProbDist)
def _eval_sent_entropy(self, sents, human=''):
for i in range(len(sents)):
sent = sents[i]
sent_tokens = sent.split()
unigrams_ = nltk.ngrams(sent_tokens, 1)
bigrams_ = bigrams(sent_tokens)
trigrams_ = trigrams(sent_tokens)
prob_unigrams = [
self.dict.unigram_freq[uni_tok] / self.total_unigrams
for uni_tok in unigrams_
]
prob_bigrams = [
self.dict.bigram_freq[bi_tok] / self.total_bigrams
for bi_tok in bigrams_
]
prob_trigrams = [
self.dict.trigram_freq[tri_tok] / self.total_trigrams
for tri_tok in trigrams_
]
# smoothing zero values
prob_unigrams = np.asarray(
[p if p > 0 else 1 for p in prob_unigrams])
prob_bigrams = np.asarray(
[p if p > 0 else 1 for p in prob_bigrams])
prob_trigrams = np.asarray(
[p if p > 0 else 1 for p in prob_trigrams])
sent_entropy_uni = -np.sum(np.log2(prob_unigrams))
sent_entropy_bi = -np.sum(np.log2(prob_bigrams))
sent_entropy_tri = -np.sum(np.log2(prob_trigrams))
word_entropy_uni = sent_entropy_uni / (len(prob_unigrams) +
sys.float_info.epsilon)
word_entropy_bi = sent_entropy_bi / (len(prob_bigrams) +
sys.float_info.epsilon)
word_entropy_tri = sent_entropy_tri / (len(prob_trigrams) +
sys.float_info.epsilon)
self.metrics[human + 'sent_entropy_uni_cnt'] += 1
self.metrics[human + 'sent_entropy_uni'] += sent_entropy_uni
self.metrics[human + 'sent_entropy_bi_cnt'] += 1
self.metrics[human + 'sent_entropy_bi'] += sent_entropy_bi
self.metrics[human + 'sent_entropy_tri_cnt'] += 1
self.metrics[human + 'sent_entropy_tri'] += sent_entropy_tri
self.metrics[human + 'word_entropy_uni_cnt'] += 1
self.metrics[human + 'word_entropy_uni'] += word_entropy_uni
self.metrics[human + 'word_entropy_bi_cnt'] += 1
self.metrics[human + 'word_entropy_bi'] += word_entropy_bi
self.metrics[human + 'word_entropy_tri_cnt'] += 1
self.metrics[human + 'word_entropy_tri'] += word_entropy_tri
def _create_objects_interface(self):
formated_objects_interface = []
specification = []
for sent in self._exchange_states():
for chunk in sent:
if chunk[0] == 'specification':
specification.append(trigrams([chunk[1], chunk[-2], \
self._convert_to_yakindu_type(type(chunk[-1]).__name__)]))
default_specification = list(OrderedSet(chain(*specification)))
objects_specification = modified_groupby(default_specification, key=lambda obj: obj[0])
for obj, specification_chunks in objects_specification.items():
formated_objects_interface.append('\n\ninterface ' + obj + ':')
for chunk in specification_chunks:
formated_objects_interface.append('\nvar ' + chunk[-2] + ':' + chunk[-1])
return ''.join(formated_objects_interface)
def __call__(self, t):
t = self.reduce_lengthening(t)
tokens = t.split(' ')
cleaned_tokens = []
for token in tokens:
token = self.replace_username(token)
token = self.replace_link(token)
cleaned_tokens.append(token)
rebuild_str = ' '.join(cleaned_tokens)
negated_tokens = mark_negation(list(self.tknzr.tokenize(rebuild_str)))
list_of_trigrams = list([' '.join(s) for s in trigrams(negated_tokens)])
return list_of_trigrams
def trigram_plot(l):
list_trigrams = list(trigrams(l))
dictionary_trigram = {}
for i in range(len(list_trigrams)):
dictionary_trigram[list_trigrams[i]] = 0
for i in range(len(list_trigrams)):
dictionary_trigram[list_trigrams[i]] += 1
plus = 0
for i in dictionary_trigram.values():
plus += i
print("total unique trigram are:", end="")
print(plus)
dictionary_trigram = dict(
sorted(dictionary_trigram.items(), key=lambda x: x[1], reverse=True))
count = 0
pdf = 0.0
for key, i in dictionary_trigram.items():
pdf += i / plus
count += 1
if (pdf > 0.7):
break
print(
"total trigrams are required to cover the 70% of the complete corpus:",
end="")
print(count)
print(pdf)
threshold = 45
for key in dictionary_trigram.copy():
if (dictionary_trigram[key] < threshold):
dictionary_trigram.pop(key)
keys_trigram, values_trigram = dictionary_trigram.keys(
), dictionary_trigram.values()
keys_trigram = list(keys_trigram)
ls = []
for i in keys_trigram:
t = ' '.join(i)
ls.append(t)
print("total trigrams taken for plotting purpose:", end="")
print(len(ls))
plt.loglog(tuple(ls), tuple(values_trigram), color='g')
plt.xticks(range(len(ls)), ls, rotation=90)
plt.xlabel('trigram')
plt.ylabel('trigram count')
plt.xscale('log')
plt.show()
plt.savefig('trigram')
def __call__(self, t):
t = self.reduce_lengthening(t)
tokens = t.split(' ')
cleaned_tokens = []
for token in tokens:
token = self.replace_username(token)
token = self.replace_link(token)
cleaned_tokens.append(token)
rebuild_str = ' '.join(cleaned_tokens)
negated_tokens = mark_negation(list(self.tknzr.tokenize(rebuild_str)))
list_of_trigrams = list(
[' '.join(s) for s in trigrams(negated_tokens)])
return list_of_trigrams
def getTrigramsDistributionFromText(txt):
trigrm = list(trigrams(txt.split()))
# print(bigrm)
trigramWords = ', '.join(' '.join((a, b, c)) for a, b, c in trigrm)
dictResTri = {}
for tri in trigramWords.split(","):
tri = tri.lstrip()
tri = tri.rstrip()
if tri in dictTrigrams:
# print(bi,dictBigrams[bi])
dictResTri[tri] = dictTrigrams[tri]
# else:
# # print("NA")
# dictResBi[bi]= 0
return (sorted(dictResTri.items(), key=lambda x: x[1], reverse=True))
def write_trigrams(words, name, minlength, count):
print("Finding " + name)
stopfilter = lambda w: len(
w) < minlength or w in stop_numbers + stop_common + ['e', 'i']
collocs = Counter(trigrams(words))
collocs = Counter({
key: val
for key, val in collocs.items() if not stopfilter(key[0])
and not stopfilter(key[1]) and not stopfilter(key[2])
})
collocs = collocs.most_common(count)
f = open(name + '.csv', 'w', encoding="utf-8")
for word, val in collocs:
f.write(u'{},{},{},{}\n'.format(word[0], word[1], word[2], val))
f.close
def tuple_to_ngrams(tuple_with_words: tuple[str],
n=2) -> tuple[tuple[str]]:
"""
Take a tuple with words and convert this tuple to a tuple with n grams (bi-/trigrams)
:param tuple_with_words:
:param n:
:return:
"""
# make bigrams and trigrams and store them in dictionary
nGrams: dict[int, tuple[tuple]] = {
2: tuple(bigrams(tuple_with_words)),
3: tuple(trigrams(tuple_with_words))
}
# return a tuple that contains tuples of size n
return nGrams[n]
def processReview_trigram(self,review):
review_text = self.stage2.removePunctuations(review["review"])
if self.const.GENERATE_TRIGRAMS_WITH_STOP_WORDS:
pass
else:
review_text = self.removeStopWordsFromReview(review_text)
review_text = review_text.lower()
tokens = review_text.split(" ")
trigram_list = trigrams(tokens)
lst =[]
for trigram in trigram_list:# if self.string_found(bigram)]
first_word = trigram[0].strip()
second_word = trigram[1].strip()
third_word = trigram[2].strip()
if ""==first_word or ""==second_word or ""==third_word:
pass
else:
lst.append({"word":first_word+" "+second_word+" "+third_word})
return lst
def profile(self, text):
''' Create FreqDist of trigrams within text '''
from nltk import word_tokenize, FreqDist
clean_text = self.remove_punctuation(text)
tokens = word_tokenize(clean_text)
fingerprint = FreqDist()
for t in tokens:
token_trigram_tuples = trigrams(self._START_CHAR + t + self._END_CHAR)
token_trigrams = [''.join(tri) for tri in token_trigram_tuples]
for cur_trigram in token_trigrams:
if cur_trigram in fingerprint:
fingerprint[cur_trigram] += 1
else:
fingerprint[cur_trigram] = 1
return fingerprint
def profile(self, text):
''' Create FreqDist of trigrams within text '''
from nltk import word_tokenize, FreqDist
clean_text = self.remove_punctuation(text)
tokens = word_tokenize(clean_text)
fingerprint = FreqDist()
for t in tokens:
token_trigram_tuples = trigrams(self._START_CHAR + t + self._END_CHAR)
token_trigrams = [''.join(tri) for tri in token_trigram_tuples]
for cur_trigram in token_trigrams:
if cur_trigram in fingerprint:
fingerprint[cur_trigram] += 1
else:
fingerprint[cur_trigram] = 1
return fingerprint
def get_trigram_word_dict(emotion_line):
words = nltk.word_tokenize(emotion_line)
word_bigrams = list(bigrams(words))
word_trigrams = list(trigrams(words))
word_feats = {}
for w in words:
if w not in word_feats:
word_feats[w] = "feature_word"
for w in word_bigrams:
if w not in word_feats:
word_feats[w] = "feature_word"
for w in word_trigrams:
if w not in word_feats:
word_feats[w] = "feature_word"
return word_feats
def buildGraphSentence(self, sentence):
nodes = list()
#unograms
possible_unograms = sentence
possible_unograms = [
uno for uno in possible_unograms
if uno.lower() not in self.stoplist
]
nodes = nodes + possible_unograms
#bigrams
possible_bigrams = list(bigrams(sentence))
possible_bigrams = [
bi for bi in possible_bigrams
if (bi[0].lower() not in self.stoplist
and bi[1].lower() not in self.stoplist)
]
possible_bigrams = [' '.join(bi) for bi in possible_bigrams]
nodes = nodes + possible_bigrams
#trigrams
possible_trigrams = list(trigrams(sentence))
possible_trigrams = [
tri for tri in possible_trigrams
if (tri[0].lower() not in self.stoplist and tri[1].lower() not in
self.stoplist and tri[2].lower() not in self.stoplist)
]
possible_trigrams = [' '.join(tri) for tri in possible_trigrams]
nodes = nodes + possible_trigrams
#print(nodes)
#add nodes
for node in nodes:
self.graph.add_node(node)
#print(self.graph.nodes)
#add edges
for node in nodes:
for node2 in nodes:
if node != node2:
if self.graph.has_edge(node, node2):
if self.NofCooc:
self.graph[node][node2]['weight'] += 1
else:
self.graph.add_edge(node, node2)
self.graph[node][node2]['weight'] = 1
def __init__(self, work_dir):
tokens = []
sentences = []
paragraphs = []
paragraph_sentence_length = []
first_word = []
for path_to_text, _, text_file_names in os.walk(work_dir):
for text_file_name in text_file_names:
text_file = file(os.path.join(path_to_text, text_file_name))
if not os.path.isfile(os.path.join(path_to_text, text_file_name)):
continue
file_content = text_file.read().decode("utf8")
print text_file_name
text_paragraphs = nltk.blankline_tokenize(file_content)
paragraphs += text_paragraphs
self._sent_tokenizer = nltk.tokenize.PunktSentenceTokenizer(file_content)
for paragraph in text_paragraphs:
paragraph_sentence = self._sent_tokenizer.tokenize(paragraph)
paragraph_sentence_length.append(len(paragraph_sentence))
sentences += paragraph_sentence
for sentence in paragraph_sentence:
sentence_tokens = nltk.word_tokenize(sentence)
tokens += sentence_tokens
first_word.append(sentence_tokens[0])
self._trigram_pd = nltk.ConditionalProbDist(
nltk.ConditionalFreqDist([(t[:2], t[2]) for t in trigrams(tokens)]), nltk.probability.ELEProbDist
)
self._bigram_pd = nltk.ConditionalProbDist(
nltk.ConditionalFreqDist([(t[:1], t[1]) for t in bigrams(tokens)]), nltk.probability.ELEProbDist
)
self._sent_begin_pd = nltk.ELEProbDist(nltk.FreqDist(first_word))
self._paragraph_length_pd = nltk.ELEProbDist(nltk.FreqDist(paragraph_sentence_length))
def get_letters():
data_rinat = open('letters_rinat.json')
data_vladimir = open('letters_vladimir.json')
root_data = json.loads(data_rinat.readlines()[0]) + json.loads(
data_vladimir.readlines()[0])
# print (data)
root_letters = list()
root_trigrams = dict()
for i in root_data:
# current_index = i.index(root_data)
get_clean_text = (re.split('\W+',
((i['coverLetter'])))) #очистить текст
# temp.append(list(trigrams(get_clean_text))) #составить триграммы и добавить их в список
# print(temp)
for grams in list(trigrams(get_clean_text)):
root_letters.append(' '.join(grams))
return collections.Counter(root_letters)
def _create_states_specification(self):
states = []
specification = []
states_specification_content = []
for sent in self._exchange_states():
specification.append([list(chain(*trigrams([chunk[1] + '{0}', chunk[-2] + \
' {1} ', str(chunk[-1]).lower() + '{2}{3}']))) \
for chunk in sent if chunk[0] == 'specification'])
while [] in specification:
specification.remove([])
for spec in specification:
spec[0].insert(0, 'entry/\n')
spec[-1][-1] = spec[-1][-1].rstrip('{2}{3}')
states_specification = dict(izip(self._get_states_content(), specification))
for state, specification in states_specification.items():
states_specification[state] = list(chain(*specification))
states_specification[state].insert(0, '"')
states_specification[state].append('"')
for state, specification in states_specification.items():
states_specification[state] = ''.join(specification).format('.', '=', ';', '\n')
return states_specification
def findBM25Terms(self):
allterms = set()
for document in self.dswa:
for sentence in document:
nodes = list()
#unograms
possible_unograms = sentence
possible_unograms = [
uno for uno in possible_unograms
if uno.lower() not in self.stoplist
]
nodes = nodes + possible_unograms
#bigrams
possible_bigrams = list(bigrams(sentence))
possible_bigrams = [
bi for bi in possible_bigrams
if (bi[0].lower() not in self.stoplist
and bi[1].lower() not in self.stoplist)
]
possible_bigrams = [' '.join(bi) for bi in possible_bigrams]
nodes = nodes + possible_bigrams
#trigrams
possible_trigrams = list(trigrams(sentence))
possible_trigrams = [
tri for tri in possible_trigrams
if (tri[0].lower() not in self.stoplist
and tri[1].lower() not in self.stoplist
and tri[2].lower() not in self.stoplist)
]
possible_trigrams = [
' '.join(tri) for tri in possible_trigrams
]
nodes = nodes + possible_trigrams
#print(nodes)
#add nodes
for node in nodes:
allterms.add(node)
return BM25Calculator(self.dswa, allterms)
def get_keywords(sentence, allowed_tags):
sentence = _remove_by_regex(_replace_punct(sentence))
tokens = nltk.word_tokenize(sentence)
tokens = [token.strip("'") for token in tokens]
tagged_tokens = nltk.pos_tag(tokens)
stop_words = get_stop_words('en')
stop_words = {word.decode('utf-8') for word in stop_words}
stop_words |= {'read'}
keywords = []
for word, tag in tagged_tokens:
word = word.lower()
if is_proper_keyword(word, tag, allowed_tags, stop_words):
keywords.append(word)
bigrams_keywords = list(bigrams(keywords))
trigrams_keywords = list(trigrams(keywords))
for k in bigrams_keywords:
keywords.append(' '.join(k))
for k in trigrams_keywords:
keywords.append(' '.join(k))
return keywords
def value_for_text(self, t, rp=default_rp):
syntax_trees = rp.parse_trees(t)
sentence_indices = []
for tree in syntax_trees:
if tree.label() == 'ROOT':
tree = tree[0]
leaves = tree.leaves()
word_indices = [0] * len(leaves)
for i in range(len(leaves)):
ref_vector = tree.leaf_treeposition(i)
j = -2
while j >= -len(ref_vector) and ref_vector[j] == 0:
parent_index = len(ref_vector) + j
parent_node = tree[ref_vector[:parent_index]]
if rp.parser().tagset.is_sentence_node(parent_node):
word_indices[i] += 1.5
else:
word_indices[i] += 1
j -= 1
if len(leaves) < 3:
sentence_index = sum(word_indices)
else:
max_trigrams = 0
for trigram in trigrams(word_indices):
if sum(trigram) > max_trigrams:
max_trigrams = sum(trigram)
sentence_index = max_trigrams
sentence_indices.append(sentence_index)
return sum(sentence_indices) / len(sentence_indices) \
if sentence_indices else 0
def filterPhrases(self):
self.filteredPhrases = []
self.sentences = sent_tokenize(self.text)
for sentence in self.sentences:
bigramList = list(set(bigrams(wordpunct_tokenize(sentence.lower()))))
for bigram in bigramList:
(word1, word2) = bigram
if word1 == "'":
term = word1 + word2
elif re.match(r'\W+', word1) == None and re.match(r'\W+', word2) == None:
term = word1 + ' ' + word2
else:
continue
if self.dbc.execute("SELECT 1 FROM vocabulary WHERE term = %s LIMIT 1;", term) == 0L:
if self.udq.recorded(term):
if len(term) <= 140:
self.dbc.execute("INSERT INTO vocabulary (term) VALUE (%s)", term)
else:
self.filteredPhrases.append(term)
trigramList = list(set(trigrams(wordpunct_tokenize(sentence.lower()))))
for trigram in trigramList:
(word1, word2, word3) = trigram
if word3 == "'":
continue
elif word2 == "'":
term = word1 + word2 + word3
elif re.match(r'\W+', word1) == None and re.match(r'\W+', word2) == None and re.match(r'\W+', word3) == None:
term = word1 + ' ' + word2 + ' ' + word3
else:
continue
if self.dbc.execute("SELECT 1 FROM vocabulary WHERE term = %s LIMIT 1;", term) == 0L:
if self.udq.recorded(term):
if len(term) <= 140:
self.dbc.execute("INSERT INTO vocabulary (term) VALUE (%s)", term)
else:
self.filteredPhrases.append(term)
return self.filteredPhrases
def ngrams(self, value):
for trigram in trigrams(self.tokenize(value)):
yield trigram
def getRecipeInfo(myURL):
### Here the webpage with the recipe is opened ###
driver = webdriver.Chrome('./chromedriver')
# driver = webdriver.Firefox()
# myURL = sys.argv[1] #'http://allrecipes.com/Recipe/Beef-Brisket-My-Way/'
#print myURL
try:
driver.get(myURL)
### Here the recipe name is extracted ###
recipeNameXPath = '//div[@class="detail-right fl-right"]/h1[@id="itemTitle"]'
recipeNameObject = driver.find_elements_by_xpath(recipeNameXPath)
for value in recipeNameObject:
recipeName = value.get_attribute("innerHTML")
#print recipeName
ingredients = []
singleIngredient = {}
ingredientSet1NamesXPath = '//div[@class="ingred-left"]/ul[@class="ingredient-wrap"]/li[@id="liIngredient"]/label/p[@class="fl-ing"]/span[@id="lblIngName"]'
ingredientSet1NamesObjects = driver.find_elements_by_xpath(ingredientSet1NamesXPath)
ingredientSet1AmountsXPath = '//div[@class="ingred-left"]/ul[@class="ingredient-wrap"]/li[@id="liIngredient"]/label/p[@class="fl-ing"]/span[@id="lblIngAmount"]'
ingredientSet1AmountsObjects = driver.find_elements_by_xpath(ingredientSet1AmountsXPath)
except:
driver.quit()
raise
for value in ingredientSet1NamesObjects:
fullSingleIngredient = str(value.get_attribute("innerHTML"))
if string.find(fullSingleIngredient, ', ') > -1:
singleIngredientParts = string.split(fullSingleIngredient, ', ')
singleIngredient['name'] = singleIngredientParts[0]
singleIngredient['descriptor'] = singleIngredientParts[1]
else:
singleIngredient['name'] = fullSingleIngredient
singleIngredient['descriptor'] = ''
singleIngredient['preparation'] = ''
ingredients.append(singleIngredient)
singleIngredient = {}
i = 0
for value in ingredientSet1AmountsObjects:
amount = str(value.get_attribute("innerHTML"))
if string.find(amount, '(') > -1:
actualAmount = string.split(amount, '(')
amount = string.split(actualAmount[1], ')')
amount = amount[0]
#print actualAmount
qty = re.search(r"[a-z]+", amount)
if qty != None:
#print qty.group(0)
ingredients[i]['measurement'] = qty.group(0)
myQty = string.replace(amount, str(qty.group(0)), '')
myQty = myQty.strip()
if string.find(myQty, '/') > -1:
qtyNum = string.split(myQty, '/')
if string.find(qtyNum[0], ' ') > -1:
numerator = string.split(qtyNum[0], ' ')
ingredients[i]['quantity'] = (float(numerator[0])*float(qtyNum[1])+float(numerator[1]))/float(qtyNum[1])
else:
ingredients[i]['quantity'] = float(qtyNum[0])/float(qtyNum[1])
else:
ingredients[i]['quantity'] = myQty
else:
ingredients[i]['measurement'] = 'unit'
ingredients[i]['quantity'] = str(value.get_attribute("innerHTML"))
i += 1
ingredientSet2NamesXPath = '//div[@class="ingred-left"]/ul[@class="ingredient-wrap secondColumn"]/li[@id="liIngredient"]/label/p[@class="fl-ing"]/span[@id="lblIngName"]'
ingredientSet2NamesObjects = driver.find_elements_by_xpath(ingredientSet2NamesXPath)
ingredientSet2AmountsXPath = '//div[@class="ingred-left"]/ul[@class="ingredient-wrap secondColumn"]/li[@id="liIngredient"]/label/p[@class="fl-ing"]/span[@id="lblIngAmount"]'
ingredientSet2AmountsObjects = driver.find_elements_by_xpath(ingredientSet2AmountsXPath)
for value in ingredientSet2NamesObjects:
fullSingleIngredient = str(value.get_attribute("innerHTML"))
if string.find(fullSingleIngredient, ', ') > -1:
singleIngredientParts = string.split(fullSingleIngredient, ', ')
singleIngredient['name'] = singleIngredientParts[0]
singleIngredient['descriptor'] = singleIngredientParts[1]
else:
singleIngredient['name'] = fullSingleIngredient
singleIngredient['descriptor'] = ''
singleIngredient['preparation'] = ''
ingredients.append(singleIngredient)
singleIngredient = {}
for value in ingredientSet2AmountsObjects:
amount = str(value.get_attribute("innerHTML"))
if string.find(amount, '(') > -1:
actualAmount = string.split(amount, '(')
amount = string.split(actualAmount[1], ')')
amount = amount[0]
#print amount
qty = re.search(r"[a-z]+", amount)
if qty != None:
#print qty.group(0)
ingredients[i]['measurement'] = qty.group(0)
myQty = string.replace(amount, str(qty.group(0)), '')
myQty = myQty.strip()
if string.find(myQty, '/') > -1:
qtyNum = string.split(myQty, '/')
if string.find(qtyNum[0], ' ') > -1:
numerator = string.split(qtyNum[0], ' ')
ingredients[i]['quantity'] = (float(numerator[0])*float(qtyNum[1])+float(numerator[1]))/float(qtyNum[1])
else:
ingredients[i]['quantity'] = float(qtyNum[0])/float(qtyNum[1])
else:
ingredients[i]['quantity'] = myQty
else:
ingredients[i]['measurement'] = 'unit'
ingredients[i]['quantity'] = str(value.get_attribute("innerHTML"))
i += 1
#pprint(ingredients)
directions = []
i = 0
directionsXPath = '//div[@class="directLeft"]/ol/li/span'
directionsObjects = driver.find_elements_by_xpath(directionsXPath)
for value in directionsObjects:
directions.append(str(value.get_attribute("innerHTML")))
i += 1
#print directions
driver.quit()
cookingMethods = {}
with open('./text_files/cookingMethods.txt', 'r') as f:
for line in f:
cookingMethods[string.replace(line, '\n', '').strip()] = True
#pprint(cookingMethods)
cookingUtensils = {}
with open('./text_files/cookingUtensils.txt', 'r') as f:
for line in f:
cookingUtensils[string.replace(line, '\n', '').strip()] = True
#pprint(cookingUtensils)
recipeCookingMethods = []
recipeCookingUtensils = []
localPhrase = ''
for step in directions:
for phrase in ngrams(string.split(step), 4):
for word in phrase:
localPhrase += word
localPhrase += ' '
localPhrase = localPhrase.strip()
localPhrase = localPhrase.replace(',', '')
localPhrase = localPhrase.replace('.', '')
for tool in cookingUtensils.keys():
if tool.lower() == localPhrase.lower():
#print localToolPhrase, 'utensil ->', tool
recipeCookingUtensils.append(tool)
localPhrase = ''
#print '4-grams done'
for phrase in trigrams(string.split(step)):
for word in phrase:
localPhrase += word
localPhrase += ' '
localPhrase = localPhrase.strip()
localPhrase = localPhrase.replace(',', '')
localPhrase = localPhrase.replace('.', '')
for tool in cookingUtensils.keys():
if tool.lower() == localPhrase.lower():
#print localToolPhrase, 'utensil ->', tool
#flag = 1
#for myTool in recipeCookingUtensils:
# if string.find(myTool, localPhrase) > -1:
# flag = 0
#if flag == 1:
recipeCookingUtensils.append(tool)
for method in cookingMethods.keys():
if method.lower() == localPhrase.lower():
recipeCookingMethods.append(method)
localPhrase = ''
#print '3-grams done'
for phrase in bigrams(string.split(step)):
for word in phrase:
localPhrase += word
localPhrase += ' '
localPhrase = localPhrase.strip()
localPhrase = localPhrase.replace(',', '')
localPhrase = localPhrase.replace('.', '')
for tool in cookingUtensils.keys():
if tool.lower() == localPhrase.lower():
#print localPhrase, 'utensil ->', tool
#flag = 1
#for myTool in recipeCookingUtensils:
# if string.find(myTool, localPhrase) > -1:
# flag = 0
#if flag == 1:
recipeCookingUtensils.append(tool)
for method in cookingMethods.keys():
if method.lower() == localPhrase.lower():
#flag = 1
#for myMethod in recipeCookingMethods:
# if string.find(myMethod, localPhrase) > -1:
# flag = 0
#if flag == 1:
recipeCookingMethods.append(method)
localPhrase = ''
#print '2-grams done'
for word in string.split(step, ' '):
#print word
if len(word) > 2:
word = word.replace(',', '')
word = word.replace('.', '')
for method in cookingMethods.keys():
#if string.find(method, word) > -1:
if method.lower() == word.lower():
#flag = 1
#for myMethod in recipeCookingMethods:
# if string.find(myMethod, localPhrase) > -1:
# flag = 0
#if flag == 1:
recipeCookingMethods.append(method)
for tool in cookingUtensils.keys():
#if string.find(tool, word) > -1:
if tool.lower() == word.lower():
#print localPhrase, 'utensil ->', tool
#flag = 1
#for myTool in recipeCookingUtensils:
# if string.find(myTool, localPhrase) > -1:
# flag = 0
#if flag == 1:
recipeCookingUtensils.append(tool)
#print '1-grams done'
utensilsSet = set(recipeCookingUtensils)
recipeCookingUtensils = list(utensilsSet)
cookingMethodsSet = set(recipeCookingMethods)
recipeCookingMethods = list(cookingMethodsSet)
recipe = {}
recipe['ingredients'] = ingredients
recipe['cooking method'] = random.choice(recipeCookingMethods)
recipe['cooking tools'] = recipeCookingUtensils
#pprint(recipe)
myInternalRecipe = {}
myInternalRecipe['name'] = str(recipeName)
myInternalRecipe['ingredients'] = []
for item in ingredients:
myInternalRecipe['ingredients'].append(item['name'])
#print myInternalRecipe
f = open('recipeJson.json', 'w')
jobj = json.dumps(recipe)
f.write(jobj)
f.close()
with open('recipeJson.json', 'r') as f:
myJobj = map(json.loads, f)
return myInternalRecipe, recipe
text = gutenberg.raw('austen-emma.txt');
nltk_sents = sent_tokenize(text) # contains the list of sentences detected from the tool
nltk_words = word_tokenize(text)
#print len(nltk_words)
#fnltk_words = [wordnet_lemmatizer.lemmatize(nltk_word) for nltk_word in nltk_words]
#print len(fnltk_words)
dictn=list(set(nltk_words))
tokens = nltk_words
tokens = [token.lower() for token in tokens if len(token) > 1] # same as unigrams
lemma_tokens = [wordnet_lemmatizer.lemmatize(token, wordnet.VERB) for token in tokens]
#print len(lemma_tokens)
tokens = lemma_tokens
bi_tokens = list(bigrams(tokens)) # getting the bigrams
tri_tokens = list(trigrams(tokens))
uni_fdist = nltk.FreqDist(tokens)
bi_fdist = nltk.FreqDist(bi_tokens)
#print len(bi_fdist)
tri_fdist = nltk.FreqDist(tri_tokens)
tri_freq = 0
bi_freq = 0
uni_freq = 0
print "top 15 unigrams with lemma\n\n"
cw=unig[w]
except:
cw=0
cw+=1
unig[w]=cw
bg=bigrams(words)
for b in bg:
try:
cb=big[b]
except:
cb=0
cb+=1
big[b]=cb
tg=trigrams(words)
for t in tg:
try:
ct=trig[t]
except:
ct=0
ct+=1
trig[t]=ct
f.close()
msim=[]
slens={}
for k in models.keys():
m=models[k]
uv=compare_histogram(m["unigrams"], unig)
def getStats2(training_set, test_set, data_type):
#List of ngrams for training list
training_unigram = [unigram for sent in training_set for unigram in sent]
training_bigram = [bigram for sent in training_set for bigram in list(bigrams(sent))]
training_trigram = [trigram for sent in training_set for trigram in list(trigrams(sent))]
#List of ngrams for test list
test_unigram = [unigram for sent in test_set for unigram in sent]
test_bigram = [bigram for sent in test_set for bigram in list(bigrams(sent))]
test_trigram = [trigram for sent in test_set for trigram in list(trigrams(sent))]
#FreqDist for each ngram for training list
fdist_training_unigram = FreqDist(training_unigram)
fdist_training_bigram =FreqDist(training_bigram)
fdist_training_trigram = FreqDist(training_trigram)
#freqDistfor each ngram for test list
fdist_test_unigram = FreqDist(test_unigram)
fdist_test_bigram =FreqDist(test_bigram)
fdist_test_trigram = FreqDist(test_trigram)
#Type freq for ngrams in training list
training_unigram_freq = fdist_training_unigram.N()
training_bigram_freq = fdist_training_bigram.N()
training_trigram_freq = fdist_training_trigram.N()
#Type freq for ngrams in test list
test_unigram_freq = fdist_test_unigram.N()
test_bigram_freq = fdist_test_bigram.N()
test_trigram_freq = fdist_test_trigram.N()
#Types for ngrams in training list
types_training_unigram = set(training_unigram)
types_training_bigram = set(training_bigram)
types_training_trigram = set(training_trigram)
#Types for ngrams in test list
types_test_unigram = set(test_unigram)
types_test_bigram = set(test_bigram)
types_test_trigram = set(test_trigram)
#types of ngrams not in test set that not in training set
unigrams_not_in_tr = types_test_unigram - types_training_unigram
bigrams_not_in_tr = types_test_bigram - types_training_bigram
trigrams_not_in_tr = types_test_trigram - types_training_trigram
perecent_unigrams_not_in_tr = 100 * len(unigrams_not_in_tr)/float(len(types_test_unigram))
perecent_bigrams_not_in_tr = 100 *len(bigrams_not_in_tr)/float(len(types_test_bigram))
perecent_trigrams_not_in_tr = 100 *len(trigrams_not_in_tr)/float(len(types_test_trigram))
overall_percent_unigrams = 100 *len(unigrams_not_in_tr)/ float(test_unigram_freq)
overall_percent_bigrams = 100 * len(bigrams_not_in_tr)/ float(test_bigram_freq )
overall_percent_trigrams = 100* len(trigrams_not_in_tr)/ float(test_trigram_freq)
print("""%s\nnum types in unigram training set:%s\n
num types in bigram training set:%s\nnum types in trigram training set: %s """ % (data_type,len(types_training_unigram), \
len(types_training_bigram), len(types_training_trigram) ))
print ("""\n%s\nnum types in unigram test set: %s\n
num types in bigram test set:%s\nnum types in trigram test set: %s """ % (data_type,len(types_test_unigram),\
len(types_test_bigram), len(types_test_trigram) ))
print ("""\n%s\nunigrams not in training set: %s\n
bigram not in tr set%s\ntrigrams not in training set: %s """ % (data_type,len(unigrams_not_in_tr),\
len(bigrams_not_in_tr), len(trigrams_not_in_tr) ))
print ("""\n%s\percent unigrams: %s\n
percent bigrams:%s\n percent trigrams: %s """ % (data_type, perecent_unigrams_not_in_tr ,\
perecent_bigrams_not_in_tr , perecent_trigrams_not_in_tr))
print ("""\n%s\noverall percent unigrams: %s\n
overall percent bigrams:%s\noverall percent trigrams: %s """ % (data_type,overall_percent_unigrams,\
overall_percent_bigrams, overall_percent_trigrams))
def brown_trigrams(category):
"""Takes as input the name of a brown category, and returns a list of all of the trigrams in the category."""
words = ["<s>"]
words += [word.lower() for word in brown.words(categories=category) if word.isalnum()]
words.append("</s>")
return list(trigrams(words))
def extract(featureList, dir, fileout, n):
tokenizer = RegexpTokenizer(r'\w+')
docPos = {}
docNeg = {}
docFeatures = {}
sentiment = "pos"
for file in os.listdir(dir+sentiment):
if file.endswith(".txt"):
features = {}
sentiment = "pos"
fp = open(dir+sentiment+"/"+file, 'rb')
doc = fp.read()
tokens = [t for t in trigrams(tokenizer.tokenize(doc))]
for word in featureList:
if word in tokens:
features[word] = 1.0
else:
features[word] = 0.0
docPos[file] = ""
docFeatures[file] = features
sentiment = "neg"
for file in os.listdir(dir+sentiment):
if file.endswith(".txt"):
features = {}
sentiment = "neg"
fp = open(dir+sentiment+"/"+file, 'rb')
doc = fp.read()
tokens = [t for t in trigrams(tokenizer.tokenize(doc))]
for word in featureList:
if word in tokens:
features[word] = 1.0
else:
features[word] = 0.0
docNeg[file] = ""
docFeatures[file] = features
f = FreqDist(featureList)
featureList = [x for (x,f) in f.items()[:n]]
allData = []
for doc in docFeatures.keys():
data = []
count = 1
if doc in docNeg.keys():
val =['-1']
if doc in docPos.keys():
val =['1']
for key in featureList:
data.append("%s:%s" %(count, docFeatures[doc][key]))
count +=1
val.extend(data)
allData.append(" ".join(val))
# for doc in docFeaturesPos.keys():
# data =['+1']
# for key in featureList:
# data.append("%s:%s" %(count, docFeaturesPos[doc][key]))
# count +=1
# count = 1
# allData.append(" ".join(data))
fVectorWriter = csv.writer(open(dir+fileout+".txt", 'wb'))
for d in allData:
print d
fVectorWriter.writerow([d])
fp.close()
sentiment = "neg"
for file in os.listdir(dir+sentiment):
if file.endswith(".txt"):
fp = open(dir+sentiment+"/"+file, 'r')
doc = fp.read()
#tokens.extend(word_tokenize(doc))
tokens.extend(tokenizer.tokenize(doc))
fp.close()
return tokens
dir = "/home/jch550/dev/JJboost/data/txt_sentoken/"
print "extracting features..."
featuresRaw = extractFeatures(dir)
# print "cleaning features..."
featuresClean = removeStopwords(featuresRaw)
featuresTrigrams = trigrams(featuresClean)
# print "writing to file..."
# fListWriter = csv.writer(open(dir+"featureTrigramsList.txt", 'w'))
# for f in featuresTrigrams:
# fListWriter.writerow([f])
# features = open(dir+"featureTrigramsList.txt", 'rb')
# featuresList = features.read().split('\r\n')
featuresList = [t for t in featuresTrigrams]
print "extracting features from documents..."
extract(featuresList, dir, "docs_train_trigrams", 500)
print "DONE."
domain_words['HCF LCM']=['The','and','The','a' .......]
'''
domain_words[i] = temp
for s in file_dumps[i].splitlines():
question_collection.append((s,i))
print("++++++++++++++++++++++++++++++++++++++++++++++")
#for random_question in question_collection:
# print(random_question[0])
# print("-----")
tokensPerQuestion = [nltk.word_tokenize(random_question[0]) for random_question in question_collection]
tokensPerQuestion = [[token.lower() for token in t if token.lower() not in symbols and token.lower() not in stoplist] for t in tokensPerQuestion]
print(tokensPerQuestion[:5])
b = list(chain(*[(list(bigrams(tokens))) for tokens in tokensPerQuestion]))
t = list(chain(*[(list(trigrams(tokens))) for tokens in tokensPerQuestion]))
print(b)
fdist = nltk.FreqDist(b)
plt.figure(figsize=(20, 8))
# plot the top 20 bigrams
fdist.plot(30)
fdist = nltk.FreqDist(t)
plt.figure(figsize=(20, 8))
# plot the top 20 trigrams
fdist.plot(30)
total_word_sents = word_training_set + word_test_set
total_pos_sents = pos_training_set + pos_test_set
total_words =[]
total_pos = []
#This flatens the total word and total sentance lists
for (w, p) in zip(total_word_sents, total_pos_sents):
for (word, pos) in zip(w, p):
total_words.append(word)
total_pos.append(pos)
# Calculates total ngram freq for words
total_word_unigrams = len(total_words)
total_word_bigrams = sum([len(list(bigrams(ngrams))) for ngrams in total_word_sents])
total_word_trigrams = sum([len(list(trigrams(ngrams))) for ngrams in total_word_sents])
# Calculates total ngram freq for POS
total_pos_unigrams = len(total_pos)
total_pos_bigrams = sum([len(list(bigrams(ngrams))) for ngrams in total_pos_sents])
total_pos_trigrams = sum([len(list(trigrams(ngrams))) for ngrams in total_pos_sents])
#Prints the info calculated above
print("Words Total:\nUnigrams:%s\nBigrams:%s\nTrigrams:%s\n" % \
(total_word_unigrams, total_word_bigrams, total_word_trigrams))
print("POS Total:\nUnigrams:%s\nBigrams:%s\nTrigrams:%s\n" % \
(total_pos_unigrams, total_pos_bigrams, total_pos_trigrams))
# Calculates all the relevant stats given the training set, testing set,
# and data type
def getStats2(training_set, test_set, data_type):
#List of ngrams for training list
training_unigram = [unigram for sent in training_set for unigram in sent]