调试

tools/generate_content_user_feature.py

@@ -225,62 +225,70 @@ if __name__ == "__main__":
 
     linucb_matrix_redis_name = "strategy:linucb:matrix:content_type:diary"
 
-    # diary_click_file = "/data/log/duan_test/feed_query_data/feed_click_info.txt"
-    # diary_fd = open(diary_click_file,"r")
-    #
-    # for line in diary_fd.readlines():
-    #     line = line.strip()
-    #     line = line.strip("\n")
-    #     line = line.strip("\r")
-    #     line = line.strip(" ")
-    #
-    #     item_list = line.split(",")
-    #     device_id = item_list[3]
-    #     diary_id = item_list[4]
+    diary_click_file = "/data/log/duan_test/feed_query_data/feed_click_info.txt"
+    diary_fd = open(diary_click_file,"r")
+
+    diary_tag_label_num_dict = dict()
+    for line in diary_fd.readlines():
+        line = line.strip()
+        line = line.strip("\n")
+        line = line.strip("\r")
+        line = line.strip(" ")
+
+        item_list = line.split(",")
+        device_id = item_list[3]
+        diary_id = item_list[4]
         # Generate_Feature_Info.generate_user_feature_to_redis(device_id, label_encoder,onehot_encoder)
-    #
-    #     user_feature_list = Generate_Feature_Info.get_user_feature_by_device_id(device_id)
-    #     content_feature_list = Generate_Feature_Info.get_content_feature(diary_id)
-    #
-    #     offi_context_feature_list = []
-    #     if len(user_feature_list)>0 and len(content_feature_list)>0:
-    #         offi_context_feature_list.append(user_feature_list[0])
-    #         offi_context_feature_list.append(content_feature_list[0])
-    #         diary_content_level = str(content_feature_list[1])
-    #
-    #         print(offi_context_feature_list)
-    #         print("*"*50)
-    #         LinUCB.update_linucb_info(user_features=offi_context_feature_list,reward=1,content_id=diary_id,
-    #                                   redis_name_linucb_matrix=linucb_matrix_redis_name,redis_cli=redis_client,
-    #                                   content_level_str=diary_content_level)
-    # diary_fd.close()
 
-
-    # device_id="868771031984211"
-    device_id="F0E142E6-EF98-482F-A470-34A6CE442D88"
         user_feature_list = Generate_Feature_Info.get_user_feature_by_device_id(device_id)
-
-    test_count = 50
-    while test_count:
-        begin=time.time()
-        test_feature = [user_feature_list[0],user_feature_list[0]]
-        print(test_feature)
-        all_diary_content_redis_dict = redis_client.hgetall("strategy:linucb:matrix:content_type:diary")
-        top_tag_dict, top_tag_set = LinUCB.linucb_recommend_tag(all_diary_content_redis_dict,test_feature,list(all_diary_content_redis_dict.keys()))
-        print(time.time()-begin)
-        print(top_tag_dict)
-
-        for diary_id in top_tag_set:
-            update_feature_list = [user_feature_list[0]]
         content_feature_list = Generate_Feature_Info.get_content_feature(diary_id)
-            if len(content_feature_list)>0:
-                update_feature_list.append(content_feature_list[0])
-            print(update_feature_list)
-            LinUCB.update_linucb_info(user_features=update_feature_list,reward=0,content_id=diary_id,redis_name_linucb_matrix=linucb_matrix_redis_name,redis_cli=redis_client)
 
-        test_count -= 1
+        offi_context_feature_list = []
+        if len(user_feature_list)>0 and len(content_feature_list)>0:
+            offi_context_feature_list.append(user_feature_list[0])
+            offi_context_feature_list.append(content_feature_list[0])
+            diary_content_level = str(content_feature_list[1])
+
+            if content_feature_list[0] not in diary_tag_label_num_dict:
+                diary_tag_label_num_dict[content_feature_list[0]] = 0
+            diary_tag_label_num_dict[content_feature_list[0]] += 1
+
+            print(offi_context_feature_list)
             print("*"*50)
-        time.sleep(2)
+            LinUCB.update_linucb_info(user_features=offi_context_feature_list,reward=1,content_id=diary_id,
+                                      redis_name_linucb_matrix=linucb_matrix_redis_name,redis_cli=redis_client,
+                                      content_level_str=diary_content_level)
+
+    diary_tag_label_num_list = sorted(diary_tag_label_num_dict.items(),key=lambda item:item[1],reverse=True)
+    print(diary_tag_label_num_list)
+    diary_fd.close()
+
+
+    # device_id="868771031984211"
+    # device_id="F0E142E6-EF98-482F-A470-34A6CE442D88"
+    # user_feature_list = Generate_Feature_Info.get_user_feature_by_device_id(device_id)
+    #
+    # test_count = 50
+    # while test_count:
+    #     begin=time.time()
+    #     test_feature = [user_feature_list[0],user_feature_list[0]]
+    #     print(test_feature)
+    #     all_diary_content_redis_dict = redis_client.hgetall("strategy:linucb:matrix:content_type:diary")
+    #     top_tag_dict, top_tag_set = LinUCB.linucb_recommend_tag(all_diary_content_redis_dict,test_feature,list(all_diary_content_redis_dict.keys()))
+    #     print(time.time()-begin)
+    #     print(top_tag_dict)
+    #
+    #     for diary_id in top_tag_set:
+    #         update_feature_list = [user_feature_list[0]]
+    #         content_feature_list = Generate_Feature_Info.get_content_feature(diary_id)
+    #         if len(content_feature_list)>0:
+    #             update_feature_list.append(content_feature_list[0])
+    #         print(update_feature_list)
+    #         LinUCB.update_linucb_info(user_features=update_feature_list,reward=0,content_id=diary_id,redis_name_linucb_matrix=linucb_matrix_redis_name,redis_cli=redis_client)
+    #
+    #     test_count -= 1
+    #     print("*"*50)
+    #     time.sleep(2)
 
     # begin=time.time()
     # test_feature = [1,2]