def get_blog(directory):
path = os.getcwd()
amount = 0
corpus = []
index = []
i = 0
for i in range(len(directory)):
with codecs.open(path+'/blogs/'+directory[i], "r",encoding='utf-8', errors='ignore') as file:
blog = file.read().split()
posts = getData.extract_post(blog)
index.append(int(directory[i].split(".")[0]))
corpus.append(posts)
amount += 1
print("blog "+str(amount))
train = []
test = []
p = 0.7
#70 - 30
for i in range(len(corpus)):
size = math.floor(len(corpus[i])*p)
train.append(getData.concat(corpus[i][:size]))
test.append(getData.concat(corpus[i][size:]))
return train, test, index
def write_author_names():
path = os.getcwd()
directory = np.array([x[2] for x in os.walk(path + '/blogs')][0])
author_names = []
with open("author_post.csv", "w") as csv_file:
writer = csv.writer(csv_file, delimiter=',')
writer.writerow(["Author", "Npost"])
for i in range(len(directory)):
with codecs.open(path + '/blogs/' + directory[i], "r", encoding='utf-8', errors='ignore') as file:
blog = file.read().split()
posts = getData.extract_post(blog)
writer.writerow([directory[i], len(posts)])
print(i)