update

new_user_analysis/user_behavior_path.py

@@ -5,3 +5,175 @@
 # @author : litao
 
 
+# -*- coding:UTF-8 -*-
+# @Time  : 2020/9/8 13:39
+# @File  : spark_test.py
+# @email : litao@igengmei.com
+# @author : litao
+
+# -*- coding:UTF-8 -*-
+# @Time  : 2020/9/4 17:07
+# @File  : search_meigou_ctr.py
+# @email : litao@igengmei.com
+# @author : litao
+
+import hashlib
+import json
+
+import pymysql
+import xlwt, datetime
+import redis
+# from pyhive import hive
+from maintenance.func_send_email_with_file import send_file_email
+from typing import Dict, List
+from elasticsearch_7 import Elasticsearch
+from elasticsearch_7.helpers import scan
+import sys
+import time
+from pyspark import SparkConf
+from pyspark.sql import SparkSession, DataFrame
+
+
+# from pyspark.sql.functions import lit
+# import pytispark.pytispark as pti
+
+
+def con_sql(sql):
+    # 从数据库的表里获取数据
+
+    db = pymysql.connect(host='172.16.40.158', port=4000, user='st_user', passwd='aqpuBLYzEV7tML5RPsN1pntUzFy',
+                         db='jerry_prod')
+    cursor = db.cursor()
+    cursor.execute(sql)
+    result = cursor.fetchall()
+    db.close()
+    return result
+
+
+startTime = time.time()
+sparkConf = SparkConf()
+sparkConf.set("spark.sql.crossJoin.enabled", True)
+sparkConf.set("spark.debug.maxToStringFields", "100")
+sparkConf.set("spark.tispark.plan.allow_index_double_read", False)
+sparkConf.set("spark.tispark.plan.allow_index_read", True)
+sparkConf.set("spark.hive.mapred.supports.subdirectories", True)
+sparkConf.set("spark.hadoop.mapreduce.input.fileinputformat.input.dir.recursive", True)
+sparkConf.set("spark.serializer", "org.apache.spark.serializer.KryoSerializer")
+sparkConf.set("mapreduce.output.fileoutputformat.compress", False)
+sparkConf.set("mapreduce.map.output.compress", False)
+sparkConf.set("prod.gold.jdbcuri",
+              "jdbc:mysql://172.16.30.136/doris_prod?user=doris&password=o5gbA27hXHHm&rewriteBatchedStatements=true")
+sparkConf.set("prod.mimas.jdbcuri",
+              "jdbc:mysql://172.16.30.138/mimas_prod?user=mimas&password=GJL3UJe1Ck9ggL6aKnZCq4cRvM&rewriteBatchedStatements=true")
+sparkConf.set("prod.gaia.jdbcuri",
+              "jdbc:mysql://172.16.30.143/zhengxing?user=work&password=BJQaT9VzDcuPBqkd&rewriteBatchedStatements=true")
+sparkConf.set("prod.tidb.jdbcuri",
+              "jdbc:mysql://172.16.40.158:4000/eagle?user=st_user&password=aqpuBLYzEV7tML5RPsN1pntUzFy&rewriteBatchedStatements=true")
+sparkConf.set("prod.jerry.jdbcuri",
+              "jdbc:mysql://172.16.40.158:4000/jerry_prod?user=st_user&password=aqpuBLYzEV7tML5RPsN1pntUzFy&rewriteBatchedStatements=true")
+sparkConf.set("prod.tispark.pd.addresses", "172.16.40.158:2379")
+sparkConf.set("prod.tispark.pd.addresses", "172.16.40.170:4000")
+sparkConf.set("prod.tidb.database", "jerry_prod")
+sparkConf.setAppName("test")
+
+spark = (SparkSession.builder.config(conf=sparkConf).config("spark.sql.extensions", "org.apache.spark.sql.TiExtensions")
+         .config("spark.tispark.pd.addresses", "172.16.40.170:2379").enableHiveSupport().getOrCreate())
+
+spark.sql("ADD JAR hdfs:///user/hive/share/lib/udf/brickhouse-0.7.1-SNAPSHOT.jar")
+spark.sql("ADD JAR hdfs:///user/hive/share/lib/udf/hive-udf-1.0-SNAPSHOT.jar")
+spark.sql("CREATE TEMPORARY FUNCTION json_map AS 'brickhouse.udf.json.JsonMapUDF'")
+spark.sql("CREATE TEMPORARY FUNCTION is_json AS 'com.gmei.hive.common.udf.UDFJsonFormatCheck'")
+spark.sql("CREATE TEMPORARY FUNCTION arrayMerge AS 'com.gmei.hive.common.udf.UDFArryMerge'")
+
+
+task_list = []
+task_days = 2
+
+for t in range(1, task_days):
+    day_num = 0 - t
+    now = (datetime.datetime.now() + datetime.timedelta(days=day_num))
+    last_30_day_str = (now + datetime.timedelta(days=-30)).strftime("%Y%m%d")
+    today_str = now.strftime("%Y%m%d")
+    yesterday_str = (now + datetime.timedelta(days=-1)).strftime("%Y%m%d")
+    one_week_age_str = (now + datetime.timedelta(days=-7)).strftime("%Y%m%d")
+    new_urser_device_id_sql = r"""
+    select t2.device_id from 
+    (select distinct(first_device) as device_id from ml_user_history_detail where partition_date = '{today_str}' and date_joined >= '{yesterday_str}') t2
+    LEFT JOIN
+    (
+        select distinct device_id
+        from ml.ml_d_ct_dv_devicespam_d  --去除机构刷单设备，即作弊设备（浏览和曝光事件去除）
+        WHERE partition_day=-'{today_str}'
+
+        union all
+
+        select distinct device_id
+        from dim.dim_device_user_staff   --去除内网用户
+    )spam_pv
+    on spam_pv.device_id=t2.device_id
+        LEFT JOIN
+    (
+        SELECT partition_date,device_id
+        FROM
+            (--找出user_id当天活跃的第一个设备id
+          SELECT user_id,partition_date,
+                  if(size(device_list) > 0, device_list [ 0 ], '') AS device_id
+            FROM online.ml_user_updates
+            WHERE partition_date=-'{today_str}'
+        )t1
+                JOIN
+            (  --医生账号
+          SELECT distinct user_id
+          FROM online.tl_hdfs_doctor_view
+          WHERE partition_date = -'{today_str}'
+
+          --马甲账号/模特用户
+          UNION ALL
+          SELECT user_id
+          FROM ml.ml_c_ct_ui_user_dimen_d
+          WHERE partition_day = -'{today_str}'
+          AND (is_puppet = 'true' or is_classifyuser = 'true')
+
+          UNION ALL
+          --公司内网覆盖用户
+          select distinct user_id
+          from dim.dim_device_user_staff
+
+          UNION ALL
+          --登陆过医生设备
+          SELECT distinct t1.user_id
+            FROM
+            (
+                SELECT user_id, v.device_id as device_id
+                FROM online.ml_user_history_detail
+                    LATERAL VIEW EXPLODE(device_history_list) v AS device_id
+                WHERE partition_date = -'{today_str}'
+            ) t1
+            JOIN
+            (
+                SELECT device_id
+                FROM online.ml_device_history_detail
+                WHERE partition_date = -'{today_str}'
+                AND is_login_doctor = '1'
+            ) t2
+                ON t1.device_id = t2.device_id
+        )t2
+            on t1.user_id=t2.user_id
+        group by partition_date,device_id
+    )dev
+    on t2.device_id=dev.device_id
+WHERE spam_pv.device_id IS NULL
+  and dev.device_id is null
+    """.format(today_str=today_str,yesterday_str=yesterday_str)
+
+    print(new_urser_device_id_sql)
+    new_urser_device_id_df = spark.sql(new_urser_device_id_sql)
+    # spam_pv_df.createOrReplaceTempView("dev_view")
+    new_urser_device_id_df.show(1)
+    sql_res = new_urser_device_id_df.collect()
+
+
+print("-------------------------------")
+for count,res in enumerate(sql_res):
+    print(count,res)
+