Merge branch 'master' of http://git.wanmeizhensuo.com/ML/ffm-baseline

fix bug

ctr_exp.py

+# -*- 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()
+
+    print("当天尾号5或6活跃用户总数：")
+    print(len(tail56))
+
+    cover = list(set(tail56)&set(device_id))
+    id = list()
+    number = list()
+    for i in cover[0:10]:
+        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]
+        id.append(i)
+        number.append(result)
+    df = pd.DataFrame()
+    df['id'] = id
+    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()
+

ctr_front.py

-# -*- coding: UTF-8 -*-
-import pymysql
-import datetime
-import pandas as pd
-
-DIRECTORY_PATH = "/data2/ffm/"
-
-
-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 device_id from eagle.ffm_diary_queue where device_id regexp '[5|6]$'"
-    cursor.execute(sql)
-    result = cursor.fetchall()
-    df = pd.DataFrame(list(result))
-    df.to_csv("/data2/ffm/10.3-10.9device_id.csv", index=None)
-    device_id = tuple(df[0].values.tolist())
-    cursor.close()
-    return device_id
-
-
-def ctr():
-    device_id = get_data()
-    db = pymysql.connect(host='10.66.157.22', port=4000, user='root', passwd='3SYz54LS9#^9sBvC', db='jerry_prod')
-    cursor = db.cursor()
-    date = []
-    exp_list = []
-    click_list = []
-    rate_list = []
-    a = "2018-10-0"
-    for i in range(1,12):
-        date.append(a+str(i))
-        sql_exp = "select count(cid) from data_feed_exposure " \
-                  "where cid_type = 'diary' and stat_date= '{}' and " \
-                  "device_id in {}".format(a+str(i),device_id)
-        cursor.execute(sql_exp)
-        exp = cursor.fetchone()[0]
-        exp_list.append(exp)
-        sql_click = "select count(cid) from data_feed_click " \
-                  "where cid_type = 'diary' and stat_date= '{}' and " \
-                  "device_id in {}".format(a + str(i), device_id)
-        cursor.execute(sql_click)
-        click = cursor.fetchone()[0]
-        click_list.append(click)
-        print(a+str(i)+"曝光数：" + str(exp))
-        print(a + str(i) + "点击数：" + str(click))
-        if exp != 0:
-            rate = click/exp
-            rate_list.append(rate)
-            print(a+str(i)+"点击率：" + str(rate))
-    db.close()
-    df = pd.DataFrame()
-    df["date"] = date
-    df["exp"] = exp_list
-    df["click"] = click_list
-    df["ctr"] = rate_list
-    print(df)
-    df.to_csv(DIRECTORY_PATH+"10.1-10.11ctr.csv",index=None)
-
-
-if __name__ == "__main__":
-    ctr()
-