ePaper-Distrubuted-GPU-Calc.../dealing_dataset.py

49 lines
1.1 KiB
Python
Raw Normal View History

2020-08-31 17:13:29 +00:00
import sqlite3
conn = sqlite3.connect(r"nlpdata.db")\
def create_dataset_ep(table):
cursor = conn.cursor()
sql = "select * from " + table + " LIMIT 20"
cursor.execute(sql)
conn.commit()
dataset = []
for row in cursor:
eid = row[0]
tag = row[1]
content = row[2]
if tag == "5" or tag == "4":
dataset.append([eid, 2, content])
print(eid, 2, content)
elif tag == "1" or tag == "2":
dataset.append([eid, 0, content])
print(eid, 0, content)
else:
dataset.append([eid, 1, content])
print(eid, 1, content)
return dataset
def create_dataset_pdt():
conn_pdt = sqlite3.connect(r".\bptdata.db")
cursor = conn_pdt.cursor()
sql = "select * from " + "predict_data"
cursor.execute(sql)
conn_pdt.commit()
dataset = []
for row in cursor:
stnid = row[0]
text = row[1]
dataset.append([stnid, 0, text])
print(stnid, 0, text)
return dataset
if __name__ == '__main__':
print(create_dataset_ep("amki_test"))