sql = "UPDATE tj_news_clob SET news_group=" + str(
group) + " WHERE news_id='" + newsid + "'"
cursor.execute(sql)
if y_pred[i] in countdict.keys():
countdict[y_pred[i]] = countdict[y_pred[i]] + 1
else:
countdict[y_pred[i]] = 1
print('修改新闻类别成功')
print('km.cluster_centers_.shape[0]:', km.cluster_centers_.shape[0])
for k in range(km.cluster_centers_.shape[0]):
sql1 = "INSERT INTO tj_cluster_centers (group_id, wv,count) VALUES (" + str(
k) + ", '" + str(km.cluster_centers_[k]) + "'," + str(
countdict[k]) + ")"
cursor.execute(sql1)
db.close()
print("修改类别成功")
np.set_printoptions(threshold=np.NaN)
start = DateUtil.nowToStrNormal()
print(start)
c = Cluster(stopword_path='hlt_stop_words.txt')
#分词
# c.wordcut()
#词向量
newsids, tdm = c.wordvector(sql=False)
print('tdm:', tdm.shape)
#kmeans
c.dokmeans(newsids, tdm, 200, 400)
print("start:", start, " end:", DateUtil.nowToStrNormal())
