调试

linucb/core/Linucb.py

@@ -155,8 +155,8 @@ class LinUCB:
             for tag_id in tag_list:
                 init_dict = {
                     "Aa": np.identity(cls.d),
-                    "theta": np.zeros((cls.d, 491)),
-                    "ba": np.zeros((cls.d, 491)),
+                    "theta": np.zeros((cls.d, 1)),
+                    "ba": np.zeros((cls.d, 1)),
                     "AaI": np.identity(cls.d)
                 }
                 pickle_data = pickle.dumps(init_dict)

tools/generate_content_user_feature.py

@@ -223,43 +223,43 @@ 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")
+
+    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 = 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])
-    #
-    #     print(offi_context_feature_list)
-    #     LinUCB.update_linucb_info(user_features_array=offi_context_feature_list,reward=1,content_id=diary_id,
-    #                               redis_name_linucb_matrix=linucb_matrix_redis_name,redis_cli=redis_client)
-    # diary_fd.close()
 
+        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 = 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])
 
-    duan_device_id="868771031984211"
-    user_feature_list = Generate_Feature_Info.get_user_feature_by_device_id(duan_device_id)
+        print(offi_context_feature_list)
+        LinUCB.update_linucb_info(user_features_array=offi_context_feature_list,reward=1,content_id=diary_id,
+                                  redis_name_linucb_matrix=linucb_matrix_redis_name,redis_cli=redis_client)
+    diary_fd.close()
 
-    begin=time.time()
-    test_feature = [user_feature_list[0],user_feature_list[0]]
-    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)
+
+    # duan_device_id="868771031984211"
+    # user_feature_list = Generate_Feature_Info.get_user_feature_by_device_id(duan_device_id)
+    #
+    # begin=time.time()
+    # test_feature = [user_feature_list[0],user_feature_list[0]]
+    # 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)
 
     # test_val_list = ["切开双眼皮"]
     #