增加首页精选的日记点击和美购首页的美购点击

eda/smart_rank/evaluation_metrics.py

@@ -180,42 +180,7 @@ def get_2_tags_coincide_rate(device_order_tags, device_portrait_result, portrait
     return result
 
 
-if __name__ == '__main__':
-    try:
-        parser = argparse.ArgumentParser(description='画像匹配度的统计')
-        my_yesterday = str(datetime.date.today() - datetime.timedelta(days=1))
-        parser.add_argument("-o", "--order_date", type=str, dest="order_date", default=my_yesterday, help="统计的下单日期")
-        parser.add_argument("-log1", "--log1_file", type=str, dest="portrait_stat_log_path",
-                            default="portrait_stat.log", help="画像统计的日志地址")
-        parser.add_argument("-log2", "--log2_file", type=str, dest="debug_portrait_stat_log_path",
-                            default="debug_portrait_stat.log", help="画像统计的日志地址")
-        parser.add_argument("-t", "--top", type=int, dest="portrait_top_n", default=3, help="选取画像的前n个tag去统计匹配度")
-        parser.add_argument("-c", "--coincide", type=int, dest="coincide_n", default=1, help="选取n个tag重合个数作为判断是否匹配的阈值")
-        parser.add_argument("-v", "--version", type=int, dest="version", default=1, help="选取翔宇(0),英赫(1)版本进行统计")
-        parser.add_argument("-e", "--exponential", type=int, dest="exponential", default=0, help="是否采用指数衰减")
-        parser.add_argument("-n", "--normalization_size", type=int, dest="normalization_size", default=7,
-                            help="天数差归一化的区间")
-        parser.add_argument("-d", "--decay_days", type=int, dest="decay_days", default=180, help="分数衰减的天数")
-        args = parser.parse_args()
-        order_date = args.order_date
-        order_date_tomorrow = str(datetime.datetime.strptime(order_date, '%Y-%m-%d') + datetime.timedelta(days=1))
-        portrait_stat_log_path = args.portrait_stat_log_path
-        debug_portrait_stat_log_path = args.debug_portrait_stat_log_path
-        cmd_portrait_top_n = args.portrait_top_n
-        cmd_coincide_n = args.coincide_n
-        version = args.version
-        exponential = args.exponential
-        normalization_size = args.normalization_size
-        decay_days = args.decay_days
-
-        LOG_DIR = "/home/gmuser/gyz/log/"
-        my_today = str(datetime.date.today())
-
-        setup_logger("log1", LOG_DIR + portrait_stat_log_path)
-        setup_logger("log2", LOG_DIR + debug_portrait_stat_log_path)
-        log1 = logging.getLogger('log1')
-        log2 = logging.getLogger('log2')
-
+def get_user_order_info_yesterday():
     # 获取昨天下单的用户设备id，下单的美购，美购对应的tag
     # api_order只有用户的user_id，一个user_id对应多个device_id
     # 用户一次可以下多个订单(美购)，一个美购对应多个tag
@@ -262,8 +227,56 @@ if __name__ == '__main__':
         """.format(order_date=order_date, order_date_tomorrow=order_date_tomorrow)
     mysql_results = get_data_by_mysql('172.16.30.141', 3306, 'work', 'BJQaT9VzDcuPBqkd', 'zhengxing',
                                       sql_order_device_info_yesterday)
-        device_ids_lst = [(i["device_id"], int(i["pay_time"])) for i in mysql_results]
+    device_ids_info = [(i["device_id"], int(i["pay_time"])) for i in mysql_results]
     all_device_order_tags = {i["device_id"]: [int(tag) for tag in i["tag_ids"].split(",")] for i in mysql_results}
+    return device_ids_info, all_device_order_tags
+
+
+def get_user_diary_click_info_yesterday():
+    pass
+
+
+def get_user_service_click_info_yesterday():
+    pass
+
+
+if __name__ == '__main__':
+    try:
+        parser = argparse.ArgumentParser(description='画像匹配度的统计')
+        my_yesterday = str(datetime.date.today() - datetime.timedelta(days=1))
+        parser.add_argument("-o", "--order_date", type=str, dest="order_date", default=my_yesterday, help="统计的下单日期")
+        parser.add_argument("-log1", "--log1_file", type=str, dest="portrait_stat_log_path",
+                            default="portrait_stat.log", help="画像统计的日志地址")
+        parser.add_argument("-log2", "--log2_file", type=str, dest="debug_portrait_stat_log_path",
+                            default="debug_portrait_stat.log", help="画像统计的日志地址")
+        parser.add_argument("-t", "--top", type=int, dest="portrait_top_n", default=3, help="选取画像的前n个tag去统计匹配度")
+        parser.add_argument("-c", "--coincide", type=int, dest="coincide_n", default=1, help="选取n个tag重合个数作为判断是否匹配的阈值")
+        parser.add_argument("-v", "--version", type=int, dest="version", default=1, help="选取翔宇(0),英赫(1)版本进行统计")
+        parser.add_argument("-e", "--exponential", type=int, dest="exponential", default=0, help="是否采用指数衰减")
+        parser.add_argument("-n", "--normalization_size", type=int, dest="normalization_size", default=7,
+                            help="天数差归一化的区间")
+        parser.add_argument("-d", "--decay_days", type=int, dest="decay_days", default=180, help="分数衰减的天数")
+        parser.add_argument("-a", "--action_type", type=list, dest="action_type", default=["order"], help="计算匹配度的行为")
+        args = parser.parse_args()
+        order_date = args.order_date
+        order_date_tomorrow = str(datetime.datetime.strptime(order_date, '%Y-%m-%d') + datetime.timedelta(days=1))
+        portrait_stat_log_path = args.portrait_stat_log_path
+        debug_portrait_stat_log_path = args.debug_portrait_stat_log_path
+        cmd_portrait_top_n = args.portrait_top_n
+        cmd_coincide_n = args.coincide_n
+        version = args.version
+        exponential = args.exponential
+        normalization_size = args.normalization_size
+        decay_days = args.decay_days
+        action_type = args.action_type
+
+        LOG_DIR = "/home/gmuser/gyz/log/"
+        my_today = str(datetime.date.today())
+
+        setup_logger("log1", LOG_DIR + portrait_stat_log_path)
+        setup_logger("log2", LOG_DIR + debug_portrait_stat_log_path)
+        log1 = logging.getLogger('log1')
+        log2 = logging.getLogger('log2')
 
         # 获取搜索词及其近义词对应的tag
         all_word_tags = get_all_word_tags()
@@ -272,7 +285,20 @@ if __name__ == '__main__':
         # 3级tag对应的2级tag
         all_3tag_2tag = get_all_3tag_2tag()
 
-        # 昨天下单了的用户的美购tags(转成2级tags)
+        for action in action_type:
+            # 获取昨天产生行为的设备id、以及行为对应的tag
+            device_ids_lst = list()
+            all_device_order_tags = dict()
+            if "order" in action_type:
+                device_ids_lst, all_device_order_tags = get_user_order_info_yesterday()
+            elif "diary" in action_type:
+                device_ids_lst, all_device_order_tags = get_user_diary_click_info_yesterday()
+            elif "service" in action_type:
+                device_ids_lst, all_device_order_tags = get_user_service_click_info_yesterday()
+            else:
+                break
+
+            # tags扩展2级tags
             all_device_order_tags2 = dict()
             for device in all_device_order_tags:
                 tags = all_device_order_tags[device]
@@ -281,7 +307,7 @@ if __name__ == '__main__':
                     tags += tags2
                 all_device_order_tags2[device] = tags
 
-        # 昨天下单了的用户的去除支付行为的画像
+            # 用户的去除支付行为的画像
             all_device_portrait_result = dict()
             debug_all_device_portrait_result = dict()
             for order_info in device_ids_lst:
@@ -308,6 +334,16 @@ if __name__ == '__main__':
             log2.info({"统计日期": my_today})
             log2.info({"参数信息": args})
             log2.info({"版本": "英赫版" if version == 1 else "翔宇版"})
+            action_type_detail = ""
+            if action_type == "order":
+                action_type_detail = "昨天下单了的用户"
+            elif action_type == "diary":
+                action_type_detail = "昨天在首页精选点击了日记的用户"
+            elif action_type == "service":
+                action_type_detail = "昨天在美购首页点击了美购的用户"
+            else:
+                pass
+            log2.info({"统计用户": action_type_detail})
             for device in no_coincide_devices:
                 no_coincide_devices_debug = dict()
                 device_portrait_n = all_device_portrait_result[device][:args.portrait_top_n]
@@ -337,6 +373,7 @@ if __name__ == '__main__':
             log1.info({"画像信息统计日期": my_today})
             log1.info({"参数信息": args})
             log1.info({"版本": "英赫版" if version == 1 else "翔宇版"})
+            log1.info({"统计用户": action_type_detail})
             log1.info({"画像更新耗时(分钟)": time_consuming})
             log1.info({"画像更新的设备数": portrait_device_count[0]["count(*)"]})
             log1.info("")