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

ctr-56.py

@@ -3,7 +3,6 @@ import pymysql
 import datetime
 import pandas as pd
 
-DIRECTORY_PATH="/data2/ffm/"
 
 def get_yesterday_date():
     today = datetime.date.today()
@@ -16,30 +15,40 @@ def get_yesterday_date():
 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 ffm_diary_queue_temp where device_id regexp '[5|6]$'"
+    sql = "select device_id from ffm_diary_queue where device_id regexp '[5|6]$'"
     cursor.execute(sql)
     result = cursor.fetchall()
-    device_id = tuple(pd.DataFrame(list(result))[0].values.tolist())
+    device = tuple(pd.DataFrame(list(result))[0].values.tolist())
     cursor.close()
-    print(device_id[0:2])
+    print(device[0:2])
 
-    return device_id
+    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]$' and stat_date = '{}';".format(date)
+    sql_active = "select distinct device_id from data_feed_exposure " \
+                 "where cid_type = 'diary' and device_id regexp'[5|6]$' and stat_date = '{}';".format(date)
     cursor.execute(sql_active)
     result = cursor.fetchall()
-    tail56 = tuple(pd.DataFrame(list(result))[0].values.tolist())
-    print("当天尾号5或6活跃用户数：")
-    print(len(tail56))
-    cover = len(set(device_id) & set(tail56))
+    tail56 = pd.DataFrame(list(result))[0].values.tolist()
+    start = "2018-09-05"
+    sql_all = "select distinct device_id from data_feed_click " \
+              "where stat_date >= '{}' and stat_date <= '{}' and cid_type = 'diary'".format(start,date)
+    cursor.execute(sql_all)
+    result_all = cursor.fetchall()
+    df = pd.DataFrame(list(result_all))
+    all_user = df[0].values.tolist()
+    print("老用户总数：")
+    print(len(all_user))
+    day_all = set(all_user) & set(tail56)
+    print("当天老用户总数：")
+    print(len(day_all))
+    cover = len(day_all & set(device_id))
     print("当天尾号5或6活跃用户覆盖数：")
     print(cover)
-    cover_percent = cover / len(set(device_id))
+    cover_percent = format(cover / len(day_all),".3f")
     print("当天尾号5或6活跃用户覆盖率：")
     print(cover_percent)
     sql_click = "select count(cid) from data_feed_click " \
@@ -55,7 +64,7 @@ def ctr_all():
     exp = cursor.fetchone()[0]
     print("曝光数：" + str(exp))
     print("点击率：" + str(click / exp))
-    return len(tail56),cover,cover_percent,click,exp,click / exp
+    return len(day_all),cover,cover_percent,click,exp,format(click/exp,".6f")
 
 
 def ctr():
@@ -75,18 +84,20 @@ def ctr():
     print("实验用户曝光数："+str(exp))
     print("实验用户点击率："+str(click/exp))
 
-    return click,exp,click/exp
+    return click,exp,format(click/exp,".6f")
 
 
 def rate2file():
-    output_path = DIRECTORY_PATH + "56tail.csv"
+    output_path = DIRECTORY_PATH + "56ctr.csv"
     with open(output_path,'a+') as f:
         line = date.replace('-', '')+','+str(temp_data[0])+','+str(temp_data[1])+','+str(temp_data[2])+\
                ","+str(data[0])+","+str(data[1])+","+str(data[2])+","+str(data[3])+","+str(data[4])\
                +","+str(data[5])+'\n'
         f.write(line)
 
+
 if __name__ == "__main__":
+    DIRECTORY_PATH = "/data/ffm/"
     date = get_yesterday_date()
     device_id = get_data()
     temp_data = ctr()