调试

tools/generate_content_user_feature.py

+# -*- coding: UTF-8 -*-
+# !/usr/bin/env python
+
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+import numpy as np
+import redis
+# from libs.cache import redis_client
+import logging
+import traceback
+import json
+import pickle
+import pymysql
+import random
+import time
+import pymysql
+from sklearn.preprocessing import LabelEncoder
+
+
+
+class Generate_Feature_Info(object):
+
+    host = "172.16.30.141"
+    user = "work"
+    password = "BJQaT9VzDcuPBqkd"
+    database = "zhengxing"
+
+    # 新项目标签encode对象
+    tagv3_label_encoder = None
+
+    @classmethod
+    def get_tagv3_word_list(cls):
+        try:
+            tagv3_name_set = set()
+            zhengxing_conn = pymysql.connect(
+                host=cls.host,
+                user=cls.user,
+                password=cls.password,
+                database=cls.database,
+                charset="utf8")
+
+            zhengxing_cursor = zhengxing_conn.cursor()
+
+            tag_v3_project_sql = """
+                    select name from api_tag_3_0 where is_online=1 and tag_type=1;
+            """
+
+            zhengxing_cursor.execute(tag_v3_project_sql)
+            sql_tag_results = zhengxing_cursor.fetchall()
+            for item in sql_tag_results:
+                encode_item = item[0].encode("utf-8")
+                tagv3_name_set.add(encode_item)
+
+            return tagv3_name_set
+        except:
+            logging.error("catch exception,err_msg:%s" % traceback.format_exc())
+            return set()
+
+    @classmethod
+    def get_tagv3_label_encode(cls,tagv3_name_list):
+        try:
+            label_encoder = LabelEncoder().fit(tagv3_name_list)
+            return label_encoder
+        except:
+            logging.error("catch exception,err_msg:%s" % traceback.format_exc())
+            return None
+
+
+
+
+
+
+
+if __name__ == "__main__":
+
+    tagv3_name_set = Generate_Feature_Info.get_tagv3_word_list()
+
+    label_encoder = Generate_Feature_Info.get_tagv3_label_encode(tagv3_name_list=list(tagv3_name_set))
+
+    test_val = "切开双眼皮"
+
+    label_encoder.transform(test_val)