Merge branch 'master' of git.wanmeizhensuo.com:ML/ffm-baseline

修改测试文件

eda/feededa/src/main/scala/com/gmei/GmeiConfig.scala

@@ -34,6 +34,7 @@ object GmeiConfig extends Serializable {
     val sparkConf = new SparkConf
     sparkConf.set("spark.sql.crossJoin.enabled", "true")
     sparkConf.set("spark.debug.maxToStringFields", "100")
+    sparkConf.set("spark.sql.broadcastTimeout",  "1000")
 
     if (!sparkConf.contains("spark.master")) {
       sparkConf.setMaster("local[3]")

eda/feededa/src/main/scala/com/gmei/Recommendation_strategy_all.scala

+package com.gmei
+
+import java.io.Serializable
+
+import com.gmei.WeafareStat.{defaultParams, parser}
+import org.apache.spark.sql.{SaveMode, TiContext}
+import org.apache.log4j.{Level, Logger}
+import scopt.OptionParser
+import com.gmei.lib.AbstractParams
+
+object Recommendation_strategy_all {
+
+  Logger.getLogger("org.apache.spark").setLevel(Level.WARN)
+  Logger.getLogger("org.apache.eclipse.jetty.server").setLevel(Level.OFF)
+
+  case class Params(env: String = "dev",
+                    date: String = "2018-08-01"
+                   ) extends AbstractParams[Params] with Serializable
+
+  val defaultParams = Params()
+
+  val parser = new OptionParser[Params]("Feed_EDA") {
+    head("WeafareStat")
+    opt[String]("env")
+      .text(s"the databases environment you used")
+      .action((x, c) => c.copy(env = x))
+    opt[String] ("date")
+      .text(s"the date you used")
+      .action((x,c) => c.copy(date = x))
+    note(
+      """
+        |For example, the following command runs this app on a tidb dataset:
+        |
+        | spark-submit --class com.gmei.WeafareStat ./target/scala-2.11/feededa-assembly-0.1.jar \
+      """.stripMargin +
+        s"|   --env ${defaultParams.env}"
+    )
+  }
+
+  def main(args: Array[String]): Unit = {
+    parser.parse(args, defaultParams).map { param =>
+      GmeiConfig.setup(param.env)
+      val spark_env = GmeiConfig.getSparkSession()
+      val sc = spark_env._2
+
+      val ti = new TiContext(sc)
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "diary_video")
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "data_feed_click")
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "blacklist")
+      ti.tidbMapTable(dbName = "jerry_test", tableName = "bl_device_list")
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "data_feed_exposure")
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "merge_queue_table")
+
+
+      import sc.implicits._
+      //val stat_date = GmeiConfig.getMinusNDate(1)
+      //println(param.date)
+      val partition_date = param.date.replace("-","")
+      val decive_id_oldUser = sc.sql(
+        s"""
+           |select distinct(device_id) as decive_id
+           |from online.ml_device_day_active_status
+           |where active_type = '4'
+           |and first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
+           |        ,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
+           |        ,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
+           |        ,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
+           |        ,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
+           |        ,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
+           |        ,'promotion_shike','promotion_julang_jl03')
+           |and partition_date ='${partition_date}'
+         """.stripMargin
+      )
+      decive_id_oldUser.createOrReplaceTempView("device_id_old")
+
+
+      val clk_count_oldUser_Contrast = sc.sql(
+        s"""
+           |select '${param.date}' as stat_date, count(cid_id) as clk_count_oldUser_Contrast
+           |from data_feed_click jd inner join device_id_old
+           |on jd.device_id = device_id_old.decive_id
+           |where  (jd.cid_type = 'diary' or jd.cid_type = 'diary_video')
+           |and jd.device_id regexp'1$$'
+           |and jd.device_id not in (select device_id from bl_device_list)
+           |and jd.device_id not in (select device_id from blacklist)
+           |and jd.stat_date ='${param.date}'
+         """.stripMargin
+        )
+
+      val imp_count_oldUser_Contrast = sc.sql(
+        s"""
+           |select '${param.date}' as stat_date, count(cid_id) as imp_count_oldUser_Contrast
+           |from data_feed_exposure je inner join device_id_old
+           |on je.device_id = device_id_old.decive_id
+           |where je.cid_type = 'diary'
+           |and je.device_id regexp'1$$'
+           |and je.device_id not in (select device_id from bl_device_list)
+           |and je.device_id not in (select device_id from blacklist)
+           |and je.stat_date ='${param.date}'
+         """.stripMargin
+        )
+
+      val clk_count_oldUser_all = sc.sql(
+        s"""
+           |select '${param.date}' as stat_date, count(cid_id) as clk_count_oldUser_all
+           |from data_feed_click jd inner join device_id_old
+           |on jd.device_id = device_id_old.decive_id
+           |where  (jd.cid_type = 'diary' or jd.cid_type = 'diary_video')
+           |and jd.device_id not in (select device_id from bl_device_list)
+           |and jd.device_id not in (select device_id from blacklist)
+           |and jd.stat_date ='${param.date}'
+         """.stripMargin
+      )
+
+      val imp_count_oldUser_all = sc.sql(
+        s"""
+           |select '${param.date}' as stat_date, count(cid_id) as imp_count_oldUser_all
+           |from data_feed_exposure je inner join device_id_old
+           |on je.device_id = device_id_old.decive_id
+           |where je.cid_type = 'diary'
+           |and je.device_id not in (select device_id from bl_device_list)
+           |and je.device_id not in (select device_id from blacklist)
+           |and je.stat_date ='${param.date}'
+         """.stripMargin
+      )
+
+//获取策略命中用户device_id
+      val device_id_cover = sc.sql(
+        s"""
+           |select distinct(device_id) as device_id
+           |from merge_queue_table
+         """.stripMargin
+      )
+      device_id_cover.createOrReplaceTempView("device_id_cover_older")
+
+
+      val clk_count_oldUser_Cover = sc.sql(
+        s"""
+           |select '${param.date}' as stat_date, count(cid_id) as clk_count_oldUser_Cover
+           |from data_feed_click jd inner join device_id_cover_older
+           |on jd.device_id = device_id_cover_older.device_id
+           |where  (jd.cid_type = 'diary' or jd.cid_type = 'diary_video')
+           |and jd.device_id not in (select device_id from bl_device_list)
+           |and jd.device_id not in (select device_id from blacklist)
+           |and jd.stat_date ='${param.date}'
+         """.stripMargin
+      )
+
+      val imp_count_oldUser_Cover = sc.sql(
+        s"""
+           |select '${param.date}' as stat_date, count(cid_id) as imp_count_oldUser_Cover
+           |from data_feed_exposure je inner join device_id_cover_older
+           |on je.device_id = device_id_cover_older.device_id
+           |where je.cid_type = 'diary'
+           |and je.device_id not in (select device_id from bl_device_list)
+           |and je.device_id not in (select device_id from blacklist)
+           |and je.stat_date ='${param.date}'
+         """.stripMargin
+      )
+
+
+
+
+      val result = clk_count_oldUser_Contrast.join(imp_count_oldUser_Contrast,"stat_date")
+          .join(clk_count_oldUser_all,"stat_date")
+          .join(imp_count_oldUser_all,"stat_date")
+          .join(clk_count_oldUser_Cover,"stat_date")
+          .join(imp_count_oldUser_Cover,"stat_date")
+      result.show()
+
+      GmeiConfig.writeToJDBCTable(result, "Recommendation_strategy_temp", SaveMode.Append)
+
+
+    }
+
+
+  }
+
+}

eda/feededa/src/main/scala/com/gmei/WeafareStat.scala

@@ -40,7 +40,7 @@ object WeafareStat {
       val sc = spark_env._2
 
       val ti = new TiContext(sc)
-      ti.tidbMapTable(dbName = "jerry_prod",tableName = "diary_video")
+      ti.tidbMapTable(dbName = "eagle",tableName = "diary_video")
       ti.tidbMapTable(dbName = "jerry_prod",tableName = "data_feed_click")
       ti.tidbMapTable(dbName = "eagle",tableName = "feed_diary_boost")
 

eda/feededa/src/main/scala/com/gmei/strategy_clk_imp_oldUser.scala

@@ -60,6 +60,13 @@ object strategy_clk_imp_oldUser {
            |select distinct(device_id) as decive_id
            |from online.ml_device_day_active_status
            |where active_type = '4'
+           |and first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
+           |        ,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
+           |        ,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
+           |        ,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
+           |        ,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
+           |        ,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
+           |        ,'promotion_shike','promotion_julang_jl03')
            |and partition_date ='${partition_date}'
          """.stripMargin
       )
@@ -71,6 +78,13 @@ object strategy_clk_imp_oldUser {
            |select distinct(device_id) as decive_id
            |from online.ml_device_day_active_status
            |where active_type != '4'
+           |and first_channel_source_type not in ('yqxiu1','yqxiu2','yqxiu3','yqxiu4','yqxiu5','mxyc1','mxyc2','mxyc3'
+           |        ,'wanpu','jinshan','jx','maimai','zhuoyi','huatian','suopingjingling','mocha','mizhe','meika','lamabang'
+           |        ,'js-az1','js-az2','js-az3','js-az4','js-az5','jfq-az1','jfq-az2','jfq-az3','jfq-az4','jfq-az5','toufang1'
+           |        ,'toufang2','toufang3','toufang4','toufang5','toufang6','TF-toufang1','TF-toufang2','TF-toufang3','TF-toufang4'
+           |        ,'TF-toufang5','tf-toufang1','tf-toufang2','tf-toufang3','tf-toufang4','tf-toufang5','benzhan','promotion_aso100'
+           |        ,'promotion_qianka','promotion_xiaoyu','promotion_dianru','promotion_malioaso','promotion_malioaso-shequ'
+           |        ,'promotion_shike','promotion_julang_jl03')
            |and partition_date ='${partition_date}'
          """.stripMargin
       )

eda/feededa/src/main/scala/com/gmei/temp_analysis.scala

+package com.gmei
+
+import java.io.Serializable
+
+import com.gmei.WeafareStat.{defaultParams, parser}
+import org.apache.spark.sql.{SaveMode, TiContext}
+import org.apache.log4j.{Level, Logger}
+import scopt.OptionParser
+import com.gmei.lib.AbstractParams
+import java.io._
+
+object temp_analysis {
+
+  Logger.getLogger("org.apache.spark").setLevel(Level.WARN)
+  Logger.getLogger("org.apache.eclipse.jetty.server").setLevel(Level.OFF)
+
+  case class Params(env: String = "dev",
+                    date: String = "2018-08-01"
+                   ) extends AbstractParams[Params] with Serializable
+
+  val defaultParams = Params()
+
+  val parser = new OptionParser[Params]("Feed_EDA") {
+    head("WeafareStat")
+    opt[String]("env")
+      .text(s"the databases environment you used")
+      .action((x, c) => c.copy(env = x))
+    opt[String] ("date")
+      .text(s"the date you used")
+      .action((x,c) => c.copy(date = x))
+    note(
+      """
+        |For example, the following command runs this app on a tidb dataset:
+        |
+        | spark-submit --class com.gmei.WeafareStat ./target/scala-2.11/feededa-assembly-0.1.jar \
+      """.stripMargin +
+        s"|   --env ${defaultParams.env}"
+    )
+  }
+
+  def main(args: Array[String]): Unit = {
+    parser.parse(args, defaultParams).map { param =>
+      GmeiConfig.setup(param.env)
+      val spark_env = GmeiConfig.getSparkSession()
+      val sc = spark_env._2
+
+      val ti = new TiContext(sc)
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "diary_video")
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "data_feed_click")
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "blacklist")
+      ti.tidbMapTable(dbName = "jerry_test", tableName = "bl_device_list")
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "data_feed_exposure")
+      ti.tidbMapTable(dbName = "jerry_prod", tableName = "merge_queue_table")
+
+
+      import sc.implicits._
+      val stat_date = GmeiConfig.getMinusNDate(1)
+      //println(param.date)
+      val partition_date = stat_date.replace("-","")
+
+      //获取策略命中用户device_id
+      val device_id_cover = sc.sql(
+        s"""
+           |select distinct(device_id) as device_id
+           |from merge_queue_table
+         """.stripMargin
+      )
+      device_id_cover.createOrReplaceTempView("device_id_cover_older")
+
+
+      val diary_pv = sc.sql(
+        s"""
+           |select '${stat_date}' as stat_date,count(params["business_id"]) as diary_pv,count(distinct(cl_id)) as device_num_diary
+           |from online.tl_hdfs_maidian_view
+           |where action="page_view"
+           |and params["page_name"]="diary_detail"
+           |and (params["out"]-params["in"])<7200
+           |and partition_date ='${partition_date}'
+         """.stripMargin
+      )
+
+      val meigou_pv = sc.sql(
+        s"""
+           |select '${stat_date}' as stat_date,count(params["business_id"]) as meigou_pv,count(distinct(cl_id)) as device_num_meigou
+           |from online.tl_hdfs_maidian_view
+           |where action="page_view"
+           |and params["page_name"]="welfare_detail"
+           |and (params["out"]-params["in"])<7200
+           |and partition_date ='${partition_date}'
+         """.stripMargin
+      )
+
+
+      val result = diary_pv.join(meigou_pv,"stat_date")
+      result.show()
+      GmeiConfig.writeToJDBCTable(result, "diary_pv", SaveMode.Append)
+
+//      result.select("stat_date","diary_pv","device_num_diary","meigou_pv","device_num_meigou").write.mode(SaveMode.Append).save("/data2/test.txt")
+
+
+
+    }
+
+
+  }
+
+}

eda/feededa/src/main/scala/com/gmei/testt.scala

@@ -7,29 +7,28 @@ import org.apache.spark.sql.{SaveMode, TiContext}
 import org.apache.log4j.{Level, Logger}
 import scopt.OptionParser
 import com.gmei.lib.AbstractParams
+import com.gmei.GmeiConfig.{writeToJDBCTable,getMinusNDate}
 
 object testt {
 
   Logger.getLogger("org.apache.spark").setLevel(Level.WARN)
   Logger.getLogger("org.apache.eclipse.jetty.server").setLevel(Level.OFF)
 
-  case class Params(env: String = "dev") extends AbstractParams[Params] with Serializable
+  case class Params(env: String = "dev"
+                   ) extends AbstractParams[Params] with Serializable
 
   val defaultParams = Params()
 
   val parser = new OptionParser[Params]("Feed_EDA") {
-    head("testt")
+    head("WeafareStat")
     opt[String]("env")
       .text(s"the databases environment you used")
       .action((x, c) => c.copy(env = x))
-    //opt[String] ("date")
-     // .text(s"the date you used")
-     // .action((x,c) => c.copy(date = x))
     note(
       """
         |For example, the following command runs this app on a tidb dataset:
         |
-        | spark-submit --class com.gmei.testt ./target/scala-2.11/feededa-assembly-0.1.jar \
+        | spark-submit --class com.gmei.WeafareStat ./target/scala-2.11/feededa-assembly-0.1.jar \
       """.stripMargin +
         s"|   --env ${defaultParams.env}"
     )
@@ -48,25 +47,34 @@ object testt {
       ti.tidbMapTable(dbName = "jerry_test", tableName = "bl_device_list")
       ti.tidbMapTable(dbName = "jerry_prod", tableName = "data_feed_exposure")
 
-      val strategies = Seq("3$","4$","5$","6$","7$","8$","c$","d$","e$","A$","B$","C$","D$")
-      for (strategy <- strategies){
-        println(strategy)
-        val get_data_dura = sc.sql(
+
+      val view_count = sc.sql(
           s"""
-             |select partition_date, sum(params['duration']) as total_dur,count(distinct(cl_id)) as num
+             |select params["business_id"] as diary_id,(params["out"]-params["in"]) as dur_time
              |from online.tl_hdfs_maidian_view
-             |where action="on_app_session_over"
+             |where action="page_view"
+             |and params["page_name"]="diary_detail"
+             |and (params["out"]-params["in"])<7200
+             |and partition_date >='20180901'
          """.stripMargin
         )
-        get_data_dura.show()
+      view_count.show()
+      view_count.createOrReplaceTempView("temp")
+
+      GmeiConfig.writeToJDBCTable(view_count, "avg", SaveMode.Overwrite)
+
+
+      val result = view_count
+      result.show()
+
 
 
       }
 
 
-    }
 
 
-  }
+    }
+
 
 }

eda/gray_stat/Recommendation_strategy_all.py

+#该文件统计ffm\node2vec\搜索推荐三个策略merge后的全量用户的ctr
+#device_id尾号为1的用户为对照组
+#该表中未能去除每天新用户
+
+
+# -*- coding: UTF-8 -*-
+from utils import con_sql,get_yesterday_date,get_between_day
+import time
+import sys
+
+OUTPUT_PATH = "/data2/models/eda/gray_stat/"
+
+#获取过滤机构用户和黑用户名单后的曝光量
+def get_imp_count_all(stragety,sta_date):
+    sql = "select count(cid_id) from jerry_prod.data_feed_exposure where " \
+          "cid_type = 'diary' " \
+          "and device_id regexp'{}$' " \
+          "and device_id not in (select distinct(device_id) from jerry_test.bl_device_list) " \
+          "and device_id not in (select device_id from jerry_prod.blacklist) and stat_date='{}'".format(stragety,sta_date)
+    imp_count_all = con_sql(sql)[0][0]
+    return imp_count_all
+
+def get_clk_count_all(stragety,sta_date):
+    sql = "select count(cid_id) from jerry_prod.data_feed_click " \
+          "where (cid_type = 'diary' or cid_type = 'diary_video') " \
+          "and device_id regexp'{}$'  " \
+          "and device_id not in (select device_id from jerry_test.bl_device_list) " \
+          "and device_id not in (select device_id from jerry_prod.blacklist) " \
+          "and stat_date='{}'".format(stragety,sta_date)
+    clk_count_all = con_sql(sql)[0][0]
+    return clk_count_all
+
+
+if __name__ == '__main__':
+
+    yesterday= get_yesterday_date()
+    if len(sys.argv) != 2:
+        print("usage: python recommendation_strategy_indicator.py date")
+    date1 = sys.argv[1]
+    date_list = get_between_day(date1, yesterday)
+    stragety_list = ['[1|2]', '[3|4]', '[5|6]', '[7|8]']
+    stragety_l = ['1', '3', '5', '7']
+    start_time = time.time()
+
+    for my_date in date_list:
+        result1_imp=[]
+        result2_imp=[]
+        result1_clk = []
+        result2_clk = []
+
+        print ("开始获取{}数据".format(my_date))
+        for i in stragety_list:
+            result1 = get_clk_count_all(i, my_date)
+            result1_clk.append(result1)
+
+            result1_all=get_imp_count_all(i,my_date)
+            result1_imp.append(result1_all)
+
+        for j in stragety_l:
+            result2 = get_clk_count(j, my_date)
+            result2_clk.append(result2)
+
+            result2_all=get_imp_count(j,my_date)
+            result2_imp.append(result2_all)
+        num_click_2=result1_clk[0]-result2_clk[0]
+        num_click_4=result1_clk[1]-result2_clk[1]
+        num_click_6 =result1_clk[2]-result2_clk[2]
+        num_click_8 =result1_clk[3]-result2_clk[3]
+
+        num_imp_2 = result1_imp[0] - result2_imp[0]
+        num_imp_4 = result1_imp[1] - result2_imp[1]
+        num_imp_6 = result1_imp[2] - result2_imp[2]
+        num_imp_8 = result1_imp[3] - result2_imp[3]
+        ctr_12 = round(result1_clk[0]/result1_imp[0],6)
+        ctr_34 = round(result1_clk[1]/result1_imp[1],6)
+        ctr_56 = round(result1_clk[2]/result1_imp[2],6)
+        ctr_78 = round(result1_clk[3]/result1_imp[3],6)
+        ctr_1 =round(result2_clk[0]/result2_imp[0],6)
+        ctr_2 =round(num_click_2/num_imp_2,6)
+        ctr_3 =round(result2_clk[1]/result2_imp[1],6)
+        ctr_4 =round(num_click_4/num_imp_4,6)
+        ctr_5 =round(result2_clk[2]/result2_imp[2],6)
+        ctr_6 =round(num_click_6/num_imp_6,6)
+        ctr_7 =round(result2_clk[3]/result2_imp[3],6)
+        ctr_8 =round(num_click_8/num_imp_8,6)
+        print(result1_clk[0])
+        print(result1_imp[0])
+        print(ctr_12)
+        print(ctr_34)
+
+
+
+        output_path = OUTPUT_PATH + "recommendation.csv"
+        with open(output_path, 'a+') as f:
+            line = my_date.replace('-','')+','+str(result1_clk[0])+','+str(result1_clk[1])+','+str(result1_clk[2])+','+str(result1_clk[3])\
+                   +','+str(result1_imp[0])+','+str(result1_imp[1])+','+str(result1_imp[2])+','+str(result1_imp[3])+','\
+                   +str(result2_clk[0])+','+str(result2_clk[1])+','+str(result2_clk[2])+','+str(result2_clk[3])\
+                   +','+str(result2_imp[0])+','+str(result2_imp[1])+','+str(result2_imp[2])+','+str(result2_imp[3]) \
+                   + ',' + str(num_click_2)+','+str(num_click_4)+','+str(num_click_6)+','+str(num_click_8) \
+                   + ',' + str(num_imp_2)+','+str(num_imp_4)+','+str(num_imp_6)+','+str(num_imp_8) \
+                   + ',' + str(ctr_12)+ ',' + str(ctr_34)+ ',' + str(ctr_56)+ ',' + str(ctr_78) \
+                   + ',' + str(ctr_1)+ ',' + str(ctr_2)+ ',' + str(ctr_3)+ ',' + str(ctr_4) \
+                   + ',' + str(ctr_5)+ ',' + str(ctr_6)+ ',' + str(ctr_7)+ ',' + str(ctr_8)+'\n'
+            f.write(line)
+    end_time = time.time()
+    print("程序执行时间：{}s".format(end_time - start_time))
\ No newline at end of file