计算匹配度的时候，用户log限制在昨天之前

eda/smart_rank/evaluation_metrics.py

@@ -44,10 +44,11 @@ def get_user_service_portrait_not_alipay(cl_id, all_word_tags, all_tag_tag_type,
         db_jerry_test = pymysql.connect(host='172.16.40.158', port=4000, user='root', passwd='3SYz54LS9#^9sBvC',
                                         db='jerry_test', charset='utf8')
         cur_jerry_test = db_jerry_test.cursor()
+        yesterday = get_yesterday_start_timestamp()
         # 用户的非搜索、支付的行为
         user_df_service_sql = "select time,cl_id,score_type,tag_id,tag_referrer,action from user_new_tag_log " \
-                              "where cl_id ='{}' and action not in " \
-                              "('api/settlement/alipay_callback','do_search')".format(cl_id)
+                              "where cl_id ='{cl_id}' and time < {yesterday} and action not in " \
+                              "('api/settlement/alipay_callback','do_search')".format(cl_id=cl_id, yesterday=yesterday)
         cur_jerry_test.execute(user_df_service_sql)
         data = list(cur_jerry_test.fetchall())
         if data:
@@ -57,7 +58,8 @@ def get_user_service_portrait_not_alipay(cl_id, all_word_tags, all_tag_tag_type,
             user_df_service = pd.DataFrame(columns=["time", "cl_id", "score_type", "tag_id", "tag_referrer", "action"])
         # 用户的搜索行为
         user_df_search_sql = "select time,cl_id,score_type,tag_id,tag_referrer,action from user_new_tag_log " \
-                             "where cl_id ='{}' and action = 'do_search'".format(cl_id)
+                             "where cl_id ='{cl_id}' and time < {yesterday} and " \
+                             "action = 'do_search'".format(cl_id=cl_id, yesterday=yesterday)
         cur_jerry_test.execute(user_df_search_sql)
         data_search = list(cur_jerry_test.fetchall())
         db_jerry_test.close()
@@ -131,10 +133,11 @@ def get_user_service_portrait_not_alipay2(cl_id, all_word_tags, all_tag_tag_type
         db_jerry_test = pymysql.connect(host='172.16.40.158', port=4000, user='root', passwd='3SYz54LS9#^9sBvC',
                                         db='jerry_test', charset='utf8')
         cur_jerry_test = db_jerry_test.cursor()
+        yesterday = get_yesterday_start_timestamp()
         # 用户的非搜索、支付的行为
         user_df_service_sql = "select time,cl_id,score_type,tag_id,tag_referrer,action from user_new_tag_log " \
-                              "where cl_id ='{}' and action not in " \
-                              "('api/settlement/alipay_callback','do_search')".format(cl_id)
+                              "where cl_id ='{cl_id}' and time < {yesterday} and action not in " \
+                              "('api/settlement/alipay_callback','do_search')".format(cl_id=cl_id, yesterday=yesterday)
         cur_jerry_test.execute(user_df_service_sql)
         data = list(cur_jerry_test.fetchall())
         if data:
@@ -144,7 +147,8 @@ def get_user_service_portrait_not_alipay2(cl_id, all_word_tags, all_tag_tag_type
             user_df_service = pd.DataFrame(columns=["time", "cl_id", "score_type", "tag_id", "tag_referrer", "action"])
         # 用户的搜索行为
         user_df_search_sql = "select time,cl_id,score_type,tag_id,tag_referrer,action from user_new_tag_log " \
-                             "where cl_id ='{}' and action = 'do_search'".format(cl_id)
+                             "where cl_id ='{cl_id}' and time < {yesterday} and " \
+                             "action = 'do_search'".format(cl_id=cl_id, yesterday=yesterday)
         cur_jerry_test.execute(user_df_search_sql)
         data_search = list(cur_jerry_test.fetchall())
         db_jerry_test.close()
@@ -207,7 +211,7 @@ def get_2_tags_coincide_rate(device_order_tags, device_portrait_result, portrait
     :param device_portrait_result:
     :param portrait_top_n:
     :param coincide_n:
-    :return: 匹配度：比对的上的用户数/昨天下单了的用户数 * 100%
+    :return: 匹配度：比对的上的用户数/昨天下单了的且有画像的用户数 * 100%
     比对的上：去掉下单和验证行为的画像的前portrait_top_n个tag  与  用户下单的美购的tag 有coincide_n个重合个数
     """
     device_count = len(device_order_tags)  # 总的下单设备数

eda/smart_rank/tool.py

@@ -211,4 +211,11 @@ def get_action_tag_count(df, action_time):
         else:
             return 1
     except Exception as e:
-        print(e)
\ No newline at end of file
+        print(e)
+
+
+def get_yesterday_start_timestamp():
+    today = datetime.date.today()
+    yesterday = today - datetime.timedelta(days=1)
+    yesterday_start_time = int(time.mktime(time.strptime(str(yesterday), '%Y-%m-%d')))
+    return yesterday_start_time