get user log

eda/smart_rank/tag3_update_user_portrait_offline.py

 from tool import get_tag3_user_log
 
+
+def update_tag3_user_portrait(cl_id):
+    user_df = get_tag3_user_log(cl_id)
+    if not user_df.empty:
+        pass
+
+
 if __name__ == "__main__":
     cl_id = "866017030837899"
     a = get_tag3_user_log(cl_id)

eda/smart_rank/tool.py

@@ -365,20 +365,38 @@ def get_jerry_test_cursor():
     return db.cursor()
 
 
+def compute_tag3_score(x):
+    if x.score_type == "henqiang":
+        return compute_henqiang(x.days_diff_now, exponential=1)
+    elif x.score_type == "jiaoqiang":
+        return compute_jiaoqiang(x.days_diff_now, exponential=1)
+    elif x.score_type == "ai_scan":
+        return compute_ai_scan(x.days_diff_now, exponential=1)
+    elif x.score_type == "ruoyixiang":
+        return compute_ruoyixiang(x.days_diff_now, exponential=1)
+    else:
+        return compute_validate(x.days_diff_now, exponential=1)
+
+
 def get_tag3_user_log(cl_id):
     columns = [
-        "log_time", "cl_id", "tag_referrer", "score_type", "action", "event", "event_cn", "first_solutions",
-        "second_solutions", "first_demands", "second_demands", "first_positions", "second_positions",
-        "projects", "card_type"
+        "log_time", "score_type", "event_cn", "first_solutions", "second_solutions", "first_demands",
+        "second_demands", "first_positions", "second_positions", "projects"
     ]
-    sql = """select log_time, cl_id, tag_referrer, score_type, action, event, event_cn, first_solutions,
-        second_solutions, first_demands, second_demands, first_positions, second_positions,
-        projects, card_type from kafka_tag3_log where cl_id = '{}'""".format(cl_id)
+    sql = """select log_time, score_type, event_cn, first_solutions, second_solutions, first_demands,
+             second_demands, first_positions, second_positions, projects
+             from kafka_tag3_log where cl_id = '{}'""".format(cl_id)
 
     cursor = get_jerry_test_cursor()
     cursor.execute(sql)
     data = list(cursor.fetchall())
-    user_df = pd.DataFrame(data)
-    user_df.columns = columns
+    if data:
+        user_df = pd.DataFrame(data)
+        user_df.columns = columns
+    else:
+        return pd.DataFrame(columns=columns)
+
+    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)
 
     return user_df