修改模型训练数据日期

diaryUpdateOnlineOffline.py

@@ -174,7 +174,6 @@ def update_dairy_queue(score_df,predict_score_df,total_video_id):
 
 
 def update_sql_dairy_queue(queue_name, diary_id,device_id, city_id):
-    begin = time.time()
     if flag:
         db = pymysql.connect(host=QUEUE_ONLINE_HOST, port=3306, user='doris', passwd='o5gbA27hXHHm',
                              db='doris_prod')
@@ -190,8 +189,6 @@ def update_sql_dairy_queue(queue_name, diary_id,device_id, city_id):
     cursor.execute(sql)
     db.commit()
     db.close()
-    over = time.time()
-    print("写入日记队列耗时{}秒".format(over-begin))
     print("成功写入diary_id")
 
 
@@ -219,7 +216,6 @@ def queue_compare(old_list, new_list):
 
 
 def get_queue(device_id, city_id,queue_name):
-    begin = time.time()
     if flag:
         db = pymysql.connect(host=QUEUE_ONLINE_HOST, port=3306, user='doris',passwd='o5gbA27hXHHm',
                              db='doris_prod')
@@ -232,8 +228,6 @@ def get_queue(device_id, city_id,queue_name):
           "where device_id = '{}' and city_id = '{}';".format(queue_name,device_id, city_id)
     cursor.execute(sql)
     result = cursor.fetchall()
-    over = time.time()
-    print("获取日记队列耗时{}秒".format(over-begin))
     df = pd.DataFrame(list(result))
     if df.empty:
         print("该用户对应的日记为空")

local/testCases.py

@@ -40,7 +40,7 @@ def get_local_device():
     print(1)
 
 
-def save_queue(id,device_id,city_id):
+def save_queue():
     queue_name_list = ["native_queue", "nearby_queue", "nation_queue", "megacity_queue"]
     for i in queue_name_list:
         sql = "select {} from device_diary_queue " \
@@ -65,10 +65,10 @@ def delete():
     sql = "delete from device_diary_queue where device_id = '{}' and city_id = '{}';".format(device_id, city_id)
     cursor.execute(sql)
     db.close()
-    print("shan")
+    print("删除成功")
 
 
-def insert(id,device_id,city_id):
+def insert():
     queue_name_list = ["native_queue", "nearby_queue", "nation_queue", "megacity_queue"]
     a = {}
     for i in queue_name_list:
@@ -87,14 +87,14 @@ def insert(id,device_id,city_id):
     print("end")
 
 
-
-
 if __name__=="__main__":
-    # 先把数据保存下来，调用上面的save函数，
+    # 先把数据保存下来，调用上面的save函数，然后调上面delete函数删除数据，然后调insert函数插入数据
     id = 334
     device_id = '00CA20EB-2719-4518-85CC-60E765AC526F'
     city_id = 'beijing'
-    insert(id, device_id, city_id)
+    save_queue()
+    delete()
+    insert()
 
 
 

train.py

@@ -19,7 +19,7 @@ if __name__ == "__main__":
     print("训练模型耗时{}分".format((end_train-start_train)/60))
     move_file()
     #TODO 如果用自己写的keepProcess文件守护进程，下面在这个函数里删掉重新启动进程那行代码，因为可能会多启动一次进程
-    # restart_process()
+    restart_process()
 
 
 

utils.py

@@ -29,11 +29,11 @@ def get_date():
     day = now.day
     date = datetime(year,month,day)
     data_start_date = "2018-07-15"
-    data_end_date = "2018-09-01"
-    validation_date = "2018-08-31"
+    # data_end_date = "2018-09-02"
+    # validation_date = "2018-09-01"
     # data_start_date = (date - timedelta(days=3)).strftime("%Y-%m-%d")
-    # data_end_date = (date - timedelta(days=1)).strftime("%Y-%m-%d")
-    # validation_date = (date - timedelta(days=2)).strftime("%Y-%m-%d")
+    data_end_date = (date - timedelta(days=1)).strftime("%Y-%m-%d")
+    validation_date = (date - timedelta(days=2)).strftime("%Y-%m-%d")
     # 验证集和测试集的日期必须相差一天，否则切割数据集时会报错
     test_date = data_end_date
     print("data_start_date,data_end_date,validation_date,test_date：")