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

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

+package com.gmei
+
+
+import java.io.Serializable
+import java.time.LocalDate
+
+import com.gmei.lib.AbstractParams
+import org.apache.log4j.{Level, Logger}
+import org.apache.spark.sql.{DataFrame, SaveMode, SparkSession, TiContext}
+import scopt.OptionParser
+
+import scala.util.parsing.json.JSON
+
+
+object esmm_feature {
+
+  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_test",tableName = "device_app_list")
+      ti.tidbMapTable(dbName = "jerry_test",tableName = "user_feature")
+
+      user_feature(sc)
+      get_applist(sc)
+
+    sc.stop()
+  }}
+
+  def get_applist(spark:SparkSession): Unit ={
+    val yesterday = LocalDate.now().minusDays(1).toString.replace("-","")
+    println(yesterday)
+    val df = spark.sql(
+      s"""
+         |select device["device_id"] as device_id,cl_type,params["installed_app_info"]
+         |from online.tl_hdfs_maidian_view where partition_date = $yesterday
+         |and action = 'user_installed_all_app_info'
+       """.stripMargin).dropDuplicates("device_id")
+    df.persist()
+
+    val old = spark.sql("select device_id from device_app_list").collect().map(x => x(0).toString)
+
+    import spark.implicits._
+    val android = df.rdd.map(x => (x(0).toString,x(1).toString,x(2).toString))
+      .filter(x => x._2 == "android").map(x => (x._1,x._2,parse_json(x._3),yesterday))
+
+    val ios = df.rdd.map(x => (x(0).toString,x(1).toString,x(2).toString))
+      .filter(x => x._2 == "ios").map(x => (x._1,x._2,x._3,yesterday))
+
+    val rdd = android.union(ios)
+
+    val new_user = rdd.filter(x => old.indexOf(x._1)== -1)
+      .toDF("device_id","os","app_list","update_date")
+    if (new_user.take(1).nonEmpty){
+      val jdbc = "jdbc:mysql://10.66.157.22:4000/jerry_test?user=root&password=3SYz54LS9#^9sBvC&rewriteBatchedStatements=true"
+      GmeiConfig.writeToJDBCTable(jdbc, new_user,"device_app_list", SaveMode.Append)
+      val tecent_jdbc = "jdbc:mysql://152.136.44.138:4000/jerry_test?user=root&password=3SYz54LS9#^9sBvC&rewriteBatchedStatements=true"
+      GmeiConfig.writeToJDBCTable(tecent_jdbc, new_user,"device_app_list", SaveMode.Append)
+
+    }else{
+      println("没有新用户需要写入")
+    }
+    df.unpersist()
+
+
+
+
+  }
+  def parse_json(str:String): String ={
+    var t = List[Map[String, Any]]()
+    val result = JSON.parseFull(str)
+    result match {
+      case Some(b: List[Map[String, Any]]) => t = t ++ b
+      case None => println("Parsing failed")
+      case other => println("Unknown data structure: " + other)
+    }
+    var x = List[String]()
+    if (t.nonEmpty){
+      for (i <- t){
+        x = x:+i("appName").toString
+      }
+    }
+    x.mkString(",")
+  }
+  def user_feature(spark:SparkSession): Unit ={
+    val yesterday = LocalDate.now().minusDays(1).toString.replace("-","")
+    println(yesterday)
+    val sql_exist = "select device_id from user_feature"
+    val old = spark.sql(sql_exist)
+      .collect().map(x => x(0).toString)
+    val sql_yesterday =
+      s"""
+         |select device["device_id"] as id,device["device_type"],device["manufacturer"],city_id,channel,
+         |partition_date from online.tl_hdfs_maidian_view where partition_date = $yesterday
+       """.stripMargin
+
+    val rdd = spark.sql(sql_yesterday).repartition(200).na.drop().dropDuplicates("id").rdd
+      .map(x =>(x(0).toString,x(1).toString,x(2).toString,x(3).toString,
+        x(4).toString,x(5).toString))
+    import spark.implicits._
+    val df_new = rdd.filter(x => old.indexOf(x._1)== -1)
+      .toDF("device_id","device_type","manufacturer","city_id","channel","date")
+    if (df_new.take(1).nonEmpty){
+      df_new.persist()
+      val jdbcuri = "jdbc:mysql://10.66.157.22:4000/jerry_test?user=root&password=3SYz54LS9#^9sBvC&rewriteBatchedStatements=true"
+      GmeiConfig.writeToJDBCTable(jdbcuri, df_new, "user_feature", SaveMode.Append)
+      val tecent_jdbc = "jdbc:mysql://152.136.44.138:4000/jerry_test?user=root&password=3SYz54LS9#^9sBvC&rewriteBatchedStatements=true"
+      GmeiConfig.writeToJDBCTable(tecent_jdbc, df_new, "user_feature", SaveMode.Append)
+      df_new.unpersist()
+    }else {
+      println("no need to insert into user feature")
+    }
+  }
+}
+

tensnsorflow/es/feature.py

@@ -37,7 +37,7 @@ def get_data():
     validate_date = con_sql(db, sql)[0].values.tolist()[0]
     print("validate_date:" + validate_date)
     temp = datetime.datetime.strptime(validate_date, "%Y-%m-%d")
-    start = (temp - datetime.timedelta(days=30)).strftime("%Y-%m-%d")
+    start = (temp - datetime.timedelta(days=300)).strftime("%Y-%m-%d")
     print(start)
     db = pymysql.connect(host='10.66.157.22', port=4000, user='root', passwd='3SYz54LS9#^9sBvC', db='jerry_test')
     sql = "select e.y,e.z,e.stat_date,e.ucity_id,feat.level2_ids,e.ccity_name,u.device_type,u.manufacturer," \
@@ -143,7 +143,7 @@ def get_predict(date,value_map,app_list_map,level2_map,level3_map):
           "left join cid_type_top c on e.device_id = c.device_id " \
           "left join cid_time_cut cut on e.cid_id = cut.cid " \
           "left join device_app_list dl on e.device_id = dl.device_id " \
-          "left join diary_feat feat on e.cid_id = feat.diary_id"
+          "left join diary_feat feat on e.cid_id = feat.diary_id limit 600"
     df = con_sql(db, sql)
     df = df.rename(columns={0: "y", 1: "z", 2: "label", 3: "ucity_id", 4: "clevel2_id", 5: "ccity_name",
                             6: "device_type", 7: "manufacturer", 8: "channel", 9: "top",10: "device_id",

tensnsorflow/es/submit.sh

@@ -32,7 +32,7 @@ rm ${DATA_PATH}/nearby/nearby_*
 
 
 echo "train..."
-${PYTHON_PATH} ${MODEL_PATH}/train.py --ctr_task_wgt=0.5 --learning_rate=0.0001 --deep_layers=512,256,128,64,32 --dropout=0.3,0.3,0.3,0.3,0.3 --optimizer=Adam --num_epochs=1 --embedding_size=16 --batch_size=1024 --field_size=8 --feature_size=300000 --l2_reg=0.005 --log_steps=100 --num_threads=36 --model_dir=${DATA_PATH}/model_ckpt/DeepCvrMTL/ --data_dir=${DATA_PATH} --task_type=train
+${PYTHON_PATH} ${MODEL_PATH}/train.py --ctr_task_wgt=0.5 --learning_rate=0.0001 --deep_layers=512,256,128,64,32 --dropout=0.3,0.3,0.3,0.3,0.3 --optimizer=Adam --num_epochs=1 --embedding_size=16 --batch_size=1024 --field_size=9 --feature_size=300000 --l2_reg=0.005 --log_steps=100 --num_threads=36 --model_dir=${DATA_PATH}/model_ckpt/DeepCvrMTL/ --data_dir=${DATA_PATH} --task_type=train
 
 
 echo "infer native..."

tensnsorflow/es/to_tfrecord.py

@@ -29,18 +29,20 @@ def gen_tfrecords(in_file):
 
     for i in range(df.shape[0]):
         feats = ["ucity_id", "ccity_name", "device_type", "manufacturer",
-                "channel", "top", "time", "stat_date"]
+                "channel", "top", "time", "stat_date","hospital_id"]
         id = np.array([])
         for j in feats:
             id = np.append(id,df[j][i])
         app_list = np.array(str(df["app_list"][i]).split(","))
         level2_list = np.array(str(df["clevel2_id"][i]).split(","))
+        level3_list = np.array(str(df["level3_ids"][i]).split(","))
         features = tf.train.Features(feature={
             "y": tf.train.Feature(float_list=tf.train.FloatList(value=[df["y"][i]])),
             "z": tf.train.Feature(float_list=tf.train.FloatList(value=[df["z"][i]])),
             "ids": tf.train.Feature(int64_list=tf.train.Int64List(value=id.astype(np.int))),
             "app_list":tf.train.Feature(int64_list=tf.train.Int64List(value=app_list.astype(np.int))),
-            "level2_list": tf.train.Feature(int64_list=tf.train.Int64List(value=level2_list.astype(np.int)))
+            "level2_list": tf.train.Feature(int64_list=tf.train.Int64List(value=level2_list.astype(np.int))),
+            "level3_list": tf.train.Feature(int64_list=tf.train.Int64List(value=level3_list.astype(np.int)))
             })
 
         example = tf.train.Example(features = features)

tensnsorflow/es/train.py

@@ -55,7 +55,8 @@ def input_fn(filenames, batch_size=32, num_epochs=1, perform_shuffle=False):
             "z": tf.FixedLenFeature([], tf.float32),
             "ids": tf.FixedLenFeature([FLAGS.field_size], tf.int64),
             "app_list": tf.VarLenFeature(tf.int64),
-            "level2_list": tf.VarLenFeature(tf.int64)
+            "level2_list": tf.VarLenFeature(tf.int64),
+            "level3_list": tf.VarLenFeature(tf.int64)
         }
 
         parsed = tf.parse_single_example(record, features)

tensnsorflow/test.py

@@ -10,7 +10,6 @@ from pyspark.sql import SparkSession
 import pandas as pd
 import pymysql
 
-
 def con_sql(db,sql):
     cursor = db.cursor()
     try: