def preprocess_one(self, raw_file):
'''
Takes single document (raw_file)
and calculates all neccessary to incorporate
that document into the existing set of documents.
Args:
raw_file: text (string)
'''
logger.info("Preprocessing one...")
# create document
document = preprocess_one(raw_file)
if document.identifier in self.documents:
logger.info("Document already exists.")
return
# update documents
self.documents[document.identifier] = document
self.docs_no = len(self.documents)
self.iterative_docs.append(document.identifier)
# update tokens
for token, occurrence in document.bag.items():
if token not in self.tokens:
self.tokens[token] = self.tokens_no
self.tokens_no += 1
new_tf_shape = (self.docs_no, self.tokens_no)
new_idf_shape = (1, self.tokens_no)
# resize matrices
tf_lil = resize_tf(self.tf, new_tf_shape)
idf_lil = resize_idf(self.idf, new_idf_shape)
# update tf
max_freq = document.bag[max(document.bag, key=document.bag.get)]
for token, freq in document.bag.items():
token_index = self.tokens[token]
doc_index = self.docs_no - 1
tf_lil[doc_index, token_index] = tf(freq, max_freq)
# update idf
for token in document.bag:
index = self.tokens[token]
idf_lil[0, index] += 1
# convert back to csr (faster multiplication)
self.tf = tf_lil.tocsr()
self.idf = idf_lil.tocsr()
# TODO: multiply only last row
self.tf_idf = self.tf.multiply(self.idf)
def preprocess_one(self, raw_file):
'''
Takes single document (raw_file)
and calculates all neccessary to incorporate
that document into the existing set of documents.
Args:
raw_file: text (string)
'''
logger.info("Preprocessing one...")
# create document
document = preprocess_one(raw_file)
if document.identifier in self.documents:
logger.info("Document already exists.")
return
# update documents
self.documents[document.identifier] = document
self.docs_no = len(self.documents)
self.iterative_docs.append(document.identifier)
# update tokens
for token, occurrence in document.bag.items():
if token not in self.tokens:
self.tokens[token] = self.tokens_no
self.tokens_no += 1
new_tf_shape = (self.docs_no, self.tokens_no)
new_idf_shape = (1, self.tokens_no)
# resize matrices
tf_lil = resize_tf(self.tf, new_tf_shape)
idf_lil = resize_idf(self.idf, new_idf_shape)
# update tf
max_freq = document.bag[max(document.bag, key=document.bag.get)]
for token, freq in document.bag.items():
token_index = self.tokens[token]
doc_index = self.docs_no - 1
tf_lil[doc_index, token_index] = tf(freq, max_freq)
# update idf
for token in document.bag:
index = self.tokens[token]
idf_lil[0, index] += 1
# convert back to csr (faster multiplication)
self.tf = tf_lil.tocsr()
self.idf = idf_lil.tocsr()
# TODO: multiply only last row
self.tf_idf = self.tf.multiply(self.idf)
def preprocess_one(self, raw_file):
'''
Takes single document (raw_file)
and calculates all that is neccessary to incorporate
that document into the existing set of documents.
Args:
raw_file: text (string)
'''
logger.info("Preprocessing one...")
# create document
document = preprocess_one(raw_file)
# update docs_bag dict
for token, occurrences in document.bag.items():
docs_number = self.docs_bag.get(token, {})
docs_number[document.identifier] = occurrences
self.docs_bag[token] = docs_number
# add the document to the documents set
self.documents[document.identifier] = document
def preprocess_one(self, raw_file):
'''
Takes single document (raw_file)
and calculates all that is neccessary to incorporate
that document into the existing set of documents.
Args:
raw_file: text (string)
'''
logger.info("Preprocessing one...")
# create document
document = preprocess_one(raw_file)
# update docs_bag dict
for token, occurrences in document.bag.items():
docs_number = self.docs_bag.get(token, {})
docs_number[document.identifier] = occurrences
self.docs_bag[token] = docs_number
# add the document to the documents set
self.documents[document.identifier] = document
