recover data

eda/smart_rank/tag3_update_user_portrait_offline.py

@@ -197,6 +197,9 @@ def update_tag3_user_portrait(cl_id):
         user_df["second_positions"] = list(zip(user_df["second_positions"].apply(lambda x: x.split(",")), user_df["tag_score"]))
         user_df["projects"] = list(zip(user_df["projects"].apply(lambda x: x.split(",")), user_df["tag_score"]))
         user_df["business_tags"] = list(zip(user_df["business_tags"].apply(lambda x: x.split(",")), user_df["tag_score"]))
+        user_df["selected_stars"] = list(zip(user_df["selected_stars"].apply(lambda x: x.split(",")), user_df["tag_score"]))
+        user_df["selected_internet_celebrity"] = list(
+            zip(user_df["selected_internet_celebrity"].apply(lambda x: x.split(",")), user_df["tag_score"]))
 
         user_df["first_solutions_dict"] = user_df["first_solutions"].apply(lambda x: make_dict_from_pair(x))
         user_df["second_solutions_dict"] = user_df["second_solutions"].apply(lambda x: make_dict_from_pair(x))
@@ -206,6 +209,9 @@ def update_tag3_user_portrait(cl_id):
         user_df["second_positions_dict"] = user_df["second_positions"].apply(lambda x: make_dict_from_pair(x))
         user_df["projects_dict"] = user_df["projects"].apply(lambda x: make_dict_from_pair(x))
         user_df["business_tags_dict"] = user_df["business_tags"].apply(lambda x: make_dict_from_pair(x))
+        user_df["selected_stars_dict"] = user_df["selected_stars"].apply(lambda x: make_dict_from_pair(x))
+        user_df["selected_internet_celebrity_dict"] = user_df["selected_internet_celebrity"].apply(
+            lambda x: make_dict_from_pair(x))
 
         first_solutions_list = user_df["first_solutions_dict"].tolist()
         second_solutions_list = user_df["second_solutions_dict"].tolist()
@@ -215,6 +221,8 @@ def update_tag3_user_portrait(cl_id):
         second_positions_list = user_df["second_positions_dict"].tolist()
         projects_list = user_df["projects_dict"].tolist()
         business_tags_list = user_df["business_tags_dict"].tolist()
+        selected_stars_list = user_df["selected_stars_dict"].tolist()
+        selected_internet_celebrity_list = user_df["selected_internet_celebrity_dict"].tolist()
 
         first_demands_score = merge_values(first_demands_list)
         second_demands_score = merge_values(second_demands_list)
@@ -225,6 +233,8 @@ def update_tag3_user_portrait(cl_id):
         # projects_score = merge_values(projects_list)
         projects_score_tmp = merge_values(projects_list)
         business_tags_score = merge_values(business_tags_list)
+        selected_stars_score = merge_values(selected_stars_list)
+        selected_internet_celebrity_score = merge_values(selected_internet_celebrity_list)
 
         tmp = [
             "不感兴趣", "没有想法", "八卦来了", "颜值速报", "医美审美干货", "其他项目", "网红颜值", "少年之名", "郑人予", "热点课代表", "私密", "陈瑞泽", "符仁杰", "祖力亚尔", "刘泽旭",
@@ -275,6 +285,8 @@ def update_tag3_user_portrait(cl_id):
             "second_positions": second_positions_score,
             "projects": projects_score,
             "business_tags": business_tags_score,
+            "selected_stars": selected_stars_score,
+            "selected_internet_celebrity": selected_internet_celebrity_score,
             "paid_business_tags": paid_business_tags_score,
             "validate_business_tags": validate_business_tags_score,
             "update_time": str(datetime.datetime.now())
@@ -298,7 +310,8 @@ def update_tag3_user_portrait(cl_id):
         if (len(first_demands_score.keys()) > 0) or (len(second_demands_score.keys()) > 0) or \
            (len(first_solutions_score.keys()) > 0) or (len(second_solutions_score.keys()) > 0) or \
            (len(first_positions_score.keys()) > 0) or (len(second_positions_score.keys()) > 0) or \
-           (len(projects_score.keys()) > 0) or (len(business_tags_score.keys()) > 0):
+           (len(projects_score.keys()) > 0) or (len(business_tags_score.keys()) > 0) or \
+           (len(selected_stars_score.keys() > 0)) or (len(selected_internet_celebrity_score.keys() > 0)):
             redis_client.set(key, json.dumps(res))
             redis_client.expire(key, 60 * 60 * 24 * 180)
 
@@ -312,6 +325,8 @@ def update_tag3_user_portrait(cl_id):
                 ",".join(second_positions_score.keys()),
                 ",".join(projects_score.keys()),
                 ",".join(business_tags_score.keys()),
+                ",".join(selected_stars_score.keys()),
+                ",".join(selected_internet_celebrity_score.keys()),
             )
 
             # body = {}
@@ -383,10 +398,10 @@ def consume_kafka():
 if __name__ == "__main__":
     start = datetime.datetime.now()
 
-    # cl_id = "867617044159377"
-    # print(update_tag3_user_portrait(cl_id))
+    cl_id = "androidid_0db761c3cd10c5a9"
+    update_tag3_user_portrait(cl_id)
 
-    consume_kafka()
+    # consume_kafka()
     end = datetime.datetime.now()
     print(end - start)
     print("done")

eda/smart_rank/tool.py

@@ -444,10 +444,12 @@ def get_tag3_user_log(cl_id):
     columns = [
         "log_time", "score_type", "event_cn", "first_solutions", "second_solutions", "first_demands", "second_demands",
         "first_positions", "second_positions", "projects", "business_tags"
+        "selected_stars", "selected_internet_celebrity"
     ]
     try:
         sql = """select log_time, score_type, event_cn, first_solutions, second_solutions, first_demands,
-                 second_demands, first_positions, second_positions, projects, business_tags
+                 second_demands, first_positions, second_positions, projects,
+                 business_tags, selected_stars, selected_internet_celebrity
                  from kafka_tag3_log where cl_id = '{}'""".format(cl_id)
 
         db, cursor = get_doris_prod()
@@ -465,6 +467,8 @@ def get_tag3_user_log(cl_id):
         user_df["days_diff_now"] = round((int(time.time()) - user_df["log_time"].astype(float)) / (24 * 60 * 60))
         user_df["tag_score"] = user_df.apply(lambda x: compute_tag3_score(x), axis=1)
         user_df["business_tags"] = user_df["business_tags"].fillna("")
+        user_df["selected_stars"] = user_df["selected_stars"].fillna("")
+        user_df["selected_internet_celebrity"] = user_df["selected_internet_celebrity"].fillna("")
         return user_df
     except Exception as e:
         print(e)
@@ -513,15 +517,17 @@ def get_tag3_user_order_log(cl_id):
 #    PRIMARY KEY(`id`)
 # )
 # ALTER TABLE `user_tag3_portrait` ADD COLUMN business_tags text COMMENT '商业标签';
+# ALTER TABLE `user_tag3_portrait` ADD COLUMN selected_internet_celebrity text COMMENT '网红';
+# ALTER TABLE `user_tag3_portrait` ADD COLUMN selected_stars text COMMENT '明星';
 def write_user_portrait(cl_id, first_solutions, second_solutions, first_demands, second_demands, first_positions,
-                        second_positions, projects, business_tags):
+                        second_positions, projects, business_tags, selected_stars, selected_internet_celebrity):
     try:
         today = datetime.date.today()
         oneday = datetime.timedelta(days=1)
         yesterday = today - oneday
-        sql = """insert into user_tag3_portrait values(null, '{}', '{}', '{}', '{}', '{}', '{}', '{}', '{}', '{}', '{}')""".format(
+        sql = """insert into user_tag3_portrait values(null, '{}', '{}', '{}', '{}', '{}', '{}', '{}', '{}', '{}', '{}', '{}', '{}')""".format(
             yesterday, cl_id, first_solutions, second_solutions, first_demands, second_demands, first_positions, second_positions,
-            projects, business_tags)
+            projects, business_tags, selected_stars, selected_internet_celebrity)
 
         db, cursor = get_jerry_test()
         cursor.execute(sql)