add function save dataset_device_id,change main for train

predictDiary.py

@@ -94,17 +94,16 @@ def router(device_id):
 
 if __name__ == "__main__":
     # TODO 如果耗时小于一分钟，下一次取到的device_id和上一次相同
-    router("00BBE073-ACE3-4588-9ED8-B35F13D282EB")
-    # while True:
-    #     start = time.time()
-    #     empty,device_id_list = get_active_users()
-    #     if empty:
-    #         time.sleep(10)
-    #     else:
-    #         for device_id in device_id_list:
-    #             router(device_id)
-    #
-    #         end = time.time()
-    #         time_cost = (end - start)
-    #         print("耗时{}秒".format(time_cost))
+    while True:
+        start = time.time()
+        empty,device_id_list = get_active_users()
+        if empty:
+            time.sleep(10)
+        else:
+            for device_id in device_id_list:
+                router(device_id)
+
+            end = time.time()
+            time_cost = (end - start)
+            print("耗时{}秒".format(time_cost))
 

processData.py

@@ -60,6 +60,13 @@ def feature_en():
     print(cid_df.head(2))
     cid_df.to_csv(DIRECTORY_PATH + "data_set_cid.csv", index=False)
 
+    # 将device_id 保存。目的是为了判断预测的device_id是否在这个集合里,如果不在，不需要预测
+    data_set_device_id = data["device_id"].unique()
+    device_id_df = pd.DataFrame()
+    device_id_df['device_id'] = data_set_device_id
+    print("data_set_device_id :")
+    print(device_id_df.head(2))
+    device_id_df.to_csv(DIRECTORY_PATH + "device_id.csv", index=False)
     return data, test_number, validation_number
 
 

train.py

@@ -8,6 +8,7 @@ if __name__ == "__main__":
     data_fe = feature_en()
     ffm_transform(data_fe)
     train()
-    print('---------------prepare candidates--------------')
-    get_eachCityDiaryTop3000()
+    print("end")
+    # print('---------------prepare candidates--------------')
+    # get_eachCityDiaryTop3000()