# -*- coding: UTF-8 -*- import pymysql import datetime import pandas as pd def get_yesterday_date(): today = datetime.date.today() yesterday = today - datetime.timedelta(days=1) yesterday = yesterday.strftime("%Y-%m-%d") print(yesterday) return yesterday def get_data(): conn2db = pymysql.connect(host='10.66.157.22', port=4000, user='root', passwd='3SYz54LS9#^9sBvC', db='eagle') cursor = conn2db.cursor() sql = "select distinct device_id from ffm_diary_queue_temp where device_id regexp '[5|6]$'" cursor.execute(sql) result = cursor.fetchall() device = tuple(pd.DataFrame(list(result))[0].values.tolist()) cursor.close() print(device[0:2]) return device def ctr_all(): db = pymysql.connect(host='10.66.157.22', port=4000, user='root', passwd='3SYz54LS9#^9sBvC', db='jerry_prod') cursor = db.cursor() sql_active = "select distinct device_id from data_feed_exposure " \ "where (cid_type = 'diary' or cid_type = 'diary_video') " \ "and device_id regexp'[5|6]$' and stat_date = '{}';".format(date) cursor.execute(sql_active) result = cursor.fetchall() tail56 = pd.DataFrame(list(result))[0].values.tolist() cover = list(set(tail56)&set(device_id)) cid = list() number = list() for i in cover: sql = "select count(cid) from data_feed_exposure " \ "where cid_type = 'diary'" \ "and stat_date = '{}' and device_id = '{}';".format(date,i) cursor.execute(sql) result = cursor.fetchone()[0] cid.append(i) number.append(result) df = pd.DataFrame() df['id'] = cid df['number'] = number df = df.sort_values(by = "number",ascending=False) df.to_csv(DIRECTORY_PATH+"exp.csv",index=None) if __name__ == "__main__": DIRECTORY_PATH = "/data2/ffm/" date = get_yesterday_date() device_id = get_data() ctr_all()