客户主题
一、背景介绍
客户主题主要是通过分析用户的下单情况构建用户画像
二、指标明细
指标列表 |
---|
总客户数 |
今日新增客户数 |
留存率(超过180天未下单表示已流失,否则表示留存) |
活跃用户数(近10天内有发件的客户表示活跃用户) |
月度新老用户数(应该是月度新用户!) |
沉睡用户数(3个月~6个月之间的用户表示已沉睡) |
流失用户数(9个月未下单表示已流失) |
客单数 |
客单价 |
平均客单数 |
普通用户数 |
三、表关联关系
1、事实表
表名 | 描述 |
---|---|
tbl_customer | 用户表 |
2、维度表
表名 | 描述 |
---|---|
tbl_codes | 物流系统码表 |
tbl_consumer_sender_info | 客户寄件信息表 |
tbl_express_package | 快递包裹表 |
3、关联关系
用户表与维度表的关联关系如下:
四、客户数据拉宽开发
1、拉宽后的字段
表 | 字段名 | 别名 | 字段描述 |
---|---|---|---|
tbl_customer | id | id | 客户ID |
tbl_customer | name | name | 客户姓名 |
tbl_customer | tel | tel | 客户电话 |
tbl_customer | mobile | mobile | 客户手机 |
tbl_customer | 客户邮箱 | ||
tbl_customer | type | type | 客户类型ID |
tbl_codes | codeDesc | type_name | 客户类型名称 |
tbl_customer | isownreg | is_own_reg | 是否自行注册 |
tbl_customer | regdt | regdt | 注册时间 |
tbl_customer | regchannelid | reg_channel_id | 注册渠道ID |
tbl_customer | state | state | 客户状态ID |
tbl_customer | cdt | cdt | 创建时间 |
tbl_customer | udt | udt | 修改时间 |
tbl_customer | lastlogindt | last_login_dt | 最后登录时间 |
tbl_customer | remark | remark | 备注 |
tbl_consumer_sender_info | cdt | first_cdt | 首次下单时间 |
tbl_consumer_sender_info | cdt | last_cdt | 尾次下单时间 |
tbl_express_package | billCount | billCount | 下单总数 |
tbl_express_package | totalAmount | totalAmount | 累计下单金额 |
tbl_customer | yyyyMMdd(cdt) | day | 创建时间 年月日格式 |
2、SQL语句
代码语言:javascript复制SELECT
TC."id" ,
TC."name" ,
TC."tel",
TC."mobile",
TC."email",
TC."type",
TC."is_own_reg",
TC."reg_dt",
TC."reg_channel_id",
TC."state",
TC."cdt",
TC."udt",
TC."last_login_dt",
TC."remark",
customercodes."code_desc",
sender_info.first_cdt AS first_sender_cdt ,
sender_info.last_cdt AS last_sender_cdt,
sender_info.billCount AS billCount,
sender_info.totalAmount AS totalAmount
FROM "tbl_customer" tc
LEFT JOIN (
SELECT
"ciid", min(sender_info."id") first_id, max(sender_info."id") last_id, min(sender_info."cdt") first_cdt, max(sender_info."cdt") last_cdt,COUNT(sender_info."id" ) billCount,sum(express_package."actual_amount") totalAmount
FROM "tbl_consumer_sender_info" sender_info
LEFT JOIN "tbl_express_package" express_package
ON SENDER_INFO."pkg_id" =express_package."id"
GROUP BY sender_info."ciid"
) sender_info
ON tc."id" = sender_info."ciid"
LEFT JOIN "tbl_codes" customercodes ON customercodes."type" =16 AND tc."type" =customercodes."code"
3、Spark实现
实现步骤:
- 在dwd目录下创建 CustomerDWD 单例对象,继承自OfflineApp特质
- 初始化环境的参数,创建SparkSession对象
- 获取客户表(tbl_customer)数据,并缓存数据
- 判断是否是首次运行,如果是首次运行的话,则全量装载数据(含历史数据)
- 获取客户寄件信息表(tbl_consumer_sender_info)数据,并缓存数据
- 获取客户包裹表(tbl_express_package)数据,并缓存数据
- 获取物流字典码表(tbl_codes)数据,并缓存数据
- 根据以下方式拉宽仓库车辆明细数据
- 根据客户id,在客户表中获取客户数据
- 根据包裹id,在包裹表中获取包裹数据
- 根据客户类型id,在物流字典码表中获取客户类型名称数据
- 创建客户明细宽表(若存在则不创建)
- 将客户明细宽表数据写入到kudu数据表中
- 删除缓存数据
3.1、初始化环境变量
初始化客户明细拉宽作业的环境变量
代码语言:javascript复制package cn.it.logistics.offline.dwd
import cn.it.logistics.common.{CodeTypeMapping, Configuration, OfflineTableDefine, SparkUtils}
import cn.it.logistics.offline.OfflineApp
import org.apache.spark.SparkConf
import org.apache.spark.sql.{DataFrame, SparkSession}
import org.apache.spark.storage.StorageLevel
import org.apache.spark.sql.functions._
import org.apache.spark.sql.types.IntegerType
/**
* 客户主题数据的拉宽操作
*/
object CustomerDWD extends OfflineApp {
//定义应用的名称
val appName = this.getClass.getSimpleName
def main(args: Array[String]): Unit = {
/**
* 实现步骤:
* 1)初始化sparkConf对象
* 2)创建sparkSession对象
* 3)加载kudu中的事实表和维度表的数据(将加载后的数据进行缓存)
* 4)定义维度表与事实表的关联
* 5)将拉宽后的数据再次写回到kudu数据库中(DWD明细层)
* 5.1:创建车辆明细宽表的schema表结构
* 5.2:创建车辆宽表(判断宽表是否存在,如果不存在则创建)
* 5.3:将数据写入到kudu中
* 6)将缓存的数据删除掉
* 7)停止任务
*/
//1)初始化sparkConf对象
val sparkConf: SparkConf = SparkUtils.autoSettingEnv(
SparkUtils.sparkConf(appName)
)
//2)创建sparkSession对象
val sparkSession: SparkSession = SparkUtils.getSparkSession(sparkConf)
sparkSession.sparkContext.setLogLevel(Configuration.LOG_OFF)
//数据处理
execute(sparkSession)
}
/**
* 数据处理
*
* @param sparkSession
*/
override def execute(sparkSession: SparkSession): Unit = {
sparkSession.stop()
}
}
3.2、加载客户相关的表并缓存
- 加载客户表的时候,需要指定日期条件,因为客户主题最终需要Azkaban定时调度执行,每天执行一次增量数据,因此需要指定日期。
- 判断是否是首次运行,如果是首次运行的话,则全量装载数据(含历史数据)
//导入隐士转换
import sparkSession.implicits._
val customerSenderInfoDF: DataFrame = getKuduSource(sparkSession, TableMapping.consumerSenderInfo, Configuration.isFirstRunnable).persist(StorageLevel.DISK_ONLY_2)
val customerDF = getKuduSource(sparkSession, TableMapping.customer, true).persist(StorageLevel.DISK_ONLY_2)
val expressPageageDF = getKuduSource(sparkSession, TableMapping.expressPackage, true).persist(StorageLevel.DISK_ONLY_2)
val codesDF: DataFrame = getKuduSource(sparkSession, TableMapping.codes, true).persist(StorageLevel.DISK_ONLY_2)
val customerTypeDF = codesDF.where($"type" === CodeTypeMapping.CustomType)
3.3、定义表的关联关系
- 为了在DWS层任务中方便的获取每日增量客户表数据(根据日期),因此在DataFrame基础上动态增加列(day),指定日期格式为yyyyMMdd
代码如下:
代码语言:javascript复制//TODO 4)定义维度表与事实表的关联关系
val left_outer = "left_outer"
/**
* 获取每个用户的首尾单发货信息及发货件数和总金额
*/
val customerSenderDetailInfoDF: DataFrame = customerSenderInfoDF.join(expressPageageDF, expressPageageDF("id") === customerSenderInfoDF("pkgId"), left_outer)
.groupBy(customerSenderInfoDF("ciid"))
.agg(min(customerSenderInfoDF("id")).alias("first_id"),
max(customerSenderInfoDF("id")).alias("last_id"),
min(expressPageageDF("cdt")).alias("first_cdt"),
max(expressPageageDF("cdt")).alias("last_cdt"),
count(customerSenderInfoDF("id")).alias("totalCount"),
sum(expressPageageDF("actualAmount")).alias("totalAmount")
)
val customerDetailDF: DataFrame = customerDF
.join(customerSenderDetailInfoDF, customerDF("id") === customerSenderInfoDF("ciid"), left_outer)
.join(customerTypeDF, customerDF("type") === customerTypeDF("code").cast(IntegerType), left_outer)
.sort(customerDF("cdt").asc)
.select(
customerDF("id"),
customerDF("name"),
customerDF("tel"),
customerDF("mobile"),
customerDF("type").cast(IntegerType),
customerTypeDF("codeDesc").as("type_name"),
customerDF("isownreg").as("is_own_reg"),
customerDF("regdt").as("regdt"),
customerDF("regchannelid").as("reg_channel_id"),
customerDF("state"),
customerDF("cdt"),
customerDF("udt"),
customerDF("lastlogindt").as("last_login_dt"),
customerDF("remark"),
customerSenderDetailInfoDF("first_id").as("first_sender_id"), //首次寄件id
customerSenderDetailInfoDF("last_id").as("last_sender_id"), //尾次寄件id
customerSenderDetailInfoDF("first_cdt").as("first_sender_cdt"), //首次寄件时间
customerSenderDetailInfoDF("last_cdt").as("last_sender_cdt"), //尾次寄件时间
customerSenderDetailInfoDF("totalCount"), //寄件总次数
customerSenderDetailInfoDF("totalAmount") //总金额
)
3.4、创建客户明细宽表并将客户明细数据写入到kudu数据表中
客户明细宽表数据需要保存到kudu中,因此在第一次执行客户明细拉宽操作时,客户明细宽表是不存在的,因此需要实现自动判断宽表是否存在,如果不存在则创建
实现步骤:
- 在CustomerDWD 单例对象中调用save方法
实实现过程:
- 在CustomerDWD 单例对象Main方法中调用save方法
save(customerDetailDF, OfflineTableDefine.customerDetail)
3.5、删除缓存数据
为了释放资源,客户明细宽表数据计算完成以后,需要将缓存的源表数据删除。
代码语言:javascript复制//移除缓存
customerDetailDF.unpersist
codesDF.unpersist
expressPackageDF.unpersist
customerSenderDF.unpersist
customerDF.unpersist
3.6、完整代码
代码语言:javascript复制package cn.it.logistics.offline.dwd
import cn.it.logistics.common.{CodeTypeMapping, Configuration, OfflineTableDefine, SparkUtils, TableMapping}
import cn.it.logistics.offline.OfflineApp
import org.apache.spark.SparkConf
import org.apache.spark.sql.{DataFrame, SparkSession}
import org.apache.spark.storage.StorageLevel
import org.apache.spark.sql.functions._
import org.apache.spark.sql.types.IntegerType
/**
* 客户主题数据的拉宽操作
*/
object CustomerDWD extends OfflineApp {
//定义应用的名称
val appName = this.getClass.getSimpleName
def main(args: Array[String]): Unit = {
/**
* 实现步骤:
* 1)初始化sparkConf对象
* 2)创建sparkSession对象
* 3)加载kudu中的事实表和维度表的数据(将加载后的数据进行缓存)
* 4)定义维度表与事实表的关联
* 5)将拉宽后的数据再次写回到kudu数据库中(DWD明细层)
* 5.1:创建车辆明细宽表的schema表结构
* 5.2:创建车辆宽表(判断宽表是否存在,如果不存在则创建)
* 5.3:将数据写入到kudu中
* 6)将缓存的数据删除掉
* 7)停止任务
*/
//1)初始化sparkConf对象
val sparkConf: SparkConf = SparkUtils.autoSettingEnv(
SparkUtils.sparkConf(appName)
)
//2)创建sparkSession对象
val sparkSession: SparkSession = SparkUtils.getSparkSession(sparkConf)
sparkSession.sparkContext.setLogLevel(Configuration.LOG_OFF)
//数据处理
execute(sparkSession)
}
/**
* 数据处理
*
* @param sparkSession
*/
override def execute(sparkSession: SparkSession): Unit = {
//导入隐士转换
import sparkSession.implicits._
val customerSenderInfoDF: DataFrame = getKuduSource(sparkSession, TableMapping.consumerSenderInfo, Configuration.isFirstRunnable).persist(StorageLevel.DISK_ONLY_2)
val customerDF = getKuduSource(sparkSession, TableMapping.customer, true).persist(StorageLevel.DISK_ONLY_2)
val expressPageageDF = getKuduSource(sparkSession, TableMapping.expressPackage, true).persist(StorageLevel.DISK_ONLY_2)
val codesDF: DataFrame = getKuduSource(sparkSession, TableMapping.codes, true).persist(StorageLevel.DISK_ONLY_2)
val customerTypeDF = codesDF.where($"type" === CodeTypeMapping.CustomType)
//TODO 4)定义维度表与事实表的关联关系
val left_outer = "left_outer"
/**
* 获取每个用户的首尾单发货信息及发货件数和总金额
*/
val customerSenderDetailInfoDF: DataFrame = customerSenderInfoDF.join(expressPageageDF, expressPageageDF("id") === customerSenderInfoDF("pkgId"), left_outer)
.groupBy(customerSenderInfoDF("ciid"))
.agg(min(customerSenderInfoDF("id")).alias("first_id"),
max(customerSenderInfoDF("id")).alias("last_id"),
min(expressPageageDF("cdt")).alias("first_cdt"),
max(expressPageageDF("cdt")).alias("last_cdt"),
count(customerSenderInfoDF("id")).alias("totalCount"),
sum(expressPageageDF("actualAmount")).alias("totalAmount")
)
val customerDetailDF: DataFrame = customerDF
.join(customerSenderDetailInfoDF, customerDF("id") === customerSenderInfoDF("ciid"), left_outer)
.join(customerTypeDF, customerDF("type") === customerTypeDF("code").cast(IntegerType), left_outer)
.sort(customerDF("cdt").asc)
.select(
customerDF("id"),
customerDF("name"),
customerDF("tel"),
customerDF("mobile"),
customerDF("type").cast(IntegerType),
customerTypeDF("codeDesc").as("type_name"),
customerDF("isownreg").as("is_own_reg"),
customerDF("regdt").as("regdt"),
customerDF("regchannelid").as("reg_channel_id"),
customerDF("state"),
customerDF("cdt"),
customerDF("udt"),
customerDF("lastlogindt").as("last_login_dt"),
customerDF("remark"),
customerSenderDetailInfoDF("first_id").as("first_sender_id"), //首次寄件id
customerSenderDetailInfoDF("last_id").as("last_sender_id"), //尾次寄件id
customerSenderDetailInfoDF("first_cdt").as("first_sender_cdt"), //首次寄件时间
customerSenderDetailInfoDF("last_cdt").as("last_sender_cdt"), //尾次寄件时间
customerSenderDetailInfoDF("totalCount"), //寄件总次数
customerSenderDetailInfoDF("totalAmount") //总金额
)
save(customerDetailDF, OfflineTableDefine.customerDetail)
// 5.4:将缓存的数据删除掉
customerDF.unpersist()
customerSenderInfoDF.unpersist()
expressPageageDF.unpersist()
customerTypeDF.unpersist()
sparkSession.stop()
}
}
五、客户数据指标开发
1、计算的字段
字段名 | 字段描述 |
---|---|
id | 主键id(数据产生时间) |
customerTotalCount | 总客户数 |
addtionTotalCount | 今日新增客户数(注册时间为今天) |
lostCustomerTotalCount | 留存数(超过180天未下单表示已流失,否则表示留存) |
lostRate | 留存率 |
activeCount | 活跃用户数(近10天内有发件的客户表示活跃用户) |
monthOfNewCustomerCount | 月度新老用户数(应该是月度新用户!) |
sleepCustomerCount | 沉睡用户数(3个月~6个月之间的用户表示已沉睡) |
loseCustomerCount | 流失用户数(9个月未下单表示已流失) |
customerBillCount | 客单数 |
customerAvgAmount | 客单价 |
avgCustomerBillCount | 平均客单数 |
2、Spark实现
实现步骤:
- 在dws目录下创建 ConsumerDWS 单例对象,继承自OfflineApp特质
- 初始化环境的参数,创建SparkSession对象
- 根据指定的日期获取拉宽后的用户宽表(tbl_customer_detail)增量数据,并缓存数据
- 判断是否是首次运行,如果是首次运行的话,则全量装载数据(含历史数据)
- 指标计算
- 总客户数
- 今日新增客户数(注册时间为今天)
- 留存数(超过180天未下单表示已流失,否则表示留存)
- 留存率
- 活跃用户数(近10天内有发件的客户表示活跃用户)
- 月度新老用户数(应该是月度新用户!)
- 沉睡用户数(3个月~6个月之间的用户表示已沉睡)
- 流失用户数(9个月未下单表示已流失)
- 客单数
- 客单价
- 平均客单数
- 普通用户数
- 获取当前时间yyyyMMddHH
- 构建要持久化的指标数据(需要判断计算的指标是否有值,若没有需要赋值默认值)
- 通过StructType构建指定Schema
- 创建客户指标数据表(若存在则不创建)
- 持久化指标数据到kudu表
2.1、初始化环境变量
代码语言:javascript复制package cn.it.logistics.offline.dws
import cn.it.logistics.common.{Configuration, DateHelper, OfflineTableDefine, SparkUtils}
import cn.it.logistics.offline.OfflineApp
import org.apache.spark.SparkConf
import org.apache.spark.rdd.RDD
import org.apache.spark.sql.{DataFrame, Row, SparkSession}
import org.apache.spark.sql.types.{DoubleType, LongType, Metadata, StringType, StructField, StructType}
import org.apache.spark.sql.functions._
import scala.collection.mutable.ArrayBuffer
/**
* 客户主题指标计算
*/
object CustomerDWS extends OfflineApp {
//定义应用程序的名称
val appName = this.getClass.getSimpleName
def main(args: Array[String]): Unit = {
/**
* 实现步骤:
* 1)创建SparkConf对象
* 2)创建SparkSession对象
* 3)读取客户明细宽表的数据
* 4)对客户明细宽表的数据进行指标的计算
* 5)将计算好的指标数据写入到kudu数据库中
* 5.1:定义指标结果表的schema信息
* 5.2:组织需要写入到kudu表的数据
* 5.3:判断指标结果表是否存在,如果不存在则创建
* 5.4:将数据写入到kudu表中
* 6)删除缓存数据
* 7)停止任务,退出sparksession
*/
//TODO 1)创建SparkConf对象
val sparkConf: SparkConf = SparkUtils.autoSettingEnv(
SparkUtils.sparkConf(appName)
)
//TODO 2)创建SparkSession对象
val sparkSession: SparkSession = SparkUtils.getSparkSession(sparkConf)
sparkSession.sparkContext.setLogLevel(Configuration.LOG_OFF)
//处理数据
execute(sparkSession)
}
/**
* 数据处理
*
* @param sparkSession
*/
override def execute(sparkSession: SparkSession): Unit = {
sparkSession.stop()
}
}
2.2、加载客户宽表增量数据并缓存
加载客户宽表的时候,需要指定日期条件,因为客户主题最终需要Azkaban定时调度执行,每天执行一次增量数据,因此需要指定日期。
代码语言:javascript复制//TODO 3)读取客户明细宽表的数据(用户主题的数据不需要按照天进行增量更新,而是每天全量运行)
val customerDetailDF = getKuduSource(sparkSession, OfflineTableDefine.customerDetail, Configuration.isFirstRunnable)
2.3、指标计算
代码语言:javascript复制//定义数据集合
val rows: ArrayBuffer[Row] = ArrayBuffer[Row]()
//TODO 4)对客户明细宽表的数据进行指标的计算
val customerTotalCount: Row = customerDetailDF.agg(count($"id").alias("total_count")).first()
//今日新增客户数
val addTotalCount: Long = customerDetailDF.where(date_format($"regDt", "yyyy-MM-dd").equalTo(DateHelper.getyesterday("yyyy-MM-dd"))).agg(count($"id")).first().getLong(0)
//留存率(超过180天未下单表示已经流失,否则表示留存)
//留存用户数
//val lostCustomerTotalCount: Long = customerDetailDF.join(customerSenderInfoDF.where("cdt >= date_sub(now(), 180)"), customerDetailDF("id") === customerSenderInfoDF("ciid")).count()
val lostCustomerTotalCount: Long = customerDetailDF.where("last_sender_cdt >= date_sub(now(), 180)").count()
println(lostCustomerTotalCount)
//留存率,超过180天未下单的用户数/所有的用户数
val lostRate: Double = (lostCustomerTotalCount / (if (customerTotalCount.isNullAt(0)) 1D else customerTotalCount.getLong(0))).asInstanceOf[Number].doubleValue()
println(lostRate)
// 活跃用户数(近10天内有发件的客户表示活跃用户)
val activeCount = customerDetailDF.where("last_sender_cdt>=date_sub(now(), 10)").count
// 月度新老用户数(应该是月度新用户!)
val monthOfNewCustomerCount = customerDetailDF.where($"regDt".between(trunc($"regDt", "MM"), date_format(current_date(), "yyyy-MM-dd"))).count
// 沉睡用户数(3个月~6个月之间的用户表示已沉睡)
val sleepCustomerCount = customerDetailDF.where("last_sender_cdt>=date_sub(now(), 180) and last_sender_cdt<=date_sub(now(), 90)").count
println(sleepCustomerCount)
// 流失用户数(9个月未下单表示已流失)
val loseCustomerCount = customerDetailDF.where("last_sender_cdt>=date_sub(now(), 270)").count
println(loseCustomerCount)
// 客单数
val customerSendInfoDF = customerDetailDF.where("first_sender_id is not null")
val customerBillCountAndAmount: Row = customerSendInfoDF.agg(sum("totalCount").alias("totalCount"), sum("totalAmount").alias("totalAmount")).first()
// 客单价
val customerAvgAmount = customerBillCountAndAmount.get(1).toString.toDouble / customerBillCountAndAmount.get(0).toString.toDouble //总金额/总件数
println(customerAvgAmount)
// 平均客单数
val avgCustomerBillCount = customerSendInfoDF.count / customerDetailDF.count
// 获取昨天时间yyyyMMdd
val cdt = DateHelper.getyesterday("yyyyMMdd")
// 构建要持久化的指标数据
val rowInfo = Row(
cdt,
if (customerTotalCount.isNullAt(0)) 0L else customerTotalCount.get(0).asInstanceOf[Number].longValue(),
addTotalCount,
lostCustomerTotalCount,
lostRate,
activeCount,
monthOfNewCustomerCount,
sleepCustomerCount,
loseCustomerCount,
if (customerBillCountAndAmount.isNullAt(0)) 0L else customerBillCountAndAmount.get(0).asInstanceOf[Number].longValue(),
customerAvgAmount,
avgCustomerBillCount
)
rows.append(rowInfo)
2.4、通过StructType构建指定Schema
代码语言:javascript复制import sparkSession.implicits._
val schema = StructType(Array(
StructField("id", StringType, true, Metadata.empty),
StructField("customerTotalCount", LongType, true, Metadata.empty),
StructField("addtionTotalCount", LongType, true, Metadata.empty),
StructField("lostCustomerTotalCount", LongType, true, Metadata.empty),
StructField("lostRate", DoubleType, true, Metadata.empty),
StructField("activeCount", LongType, true, Metadata.empty),
StructField("monthOfNewCustomerCount", LongType, true, Metadata.empty),
StructField("sleepCustomerCount", LongType, true, Metadata.empty),
StructField("loseCustomerCount", LongType, true, Metadata.empty),
StructField("customerBillCount", LongType, true, Metadata.empty),
StructField("customerAvgAmount", DoubleType, true, Metadata.empty),
StructField("avgCustomerBillCount", LongType, true, Metadata.empty)
))
2.5、持久化指标数据到kudu表
代码语言:javascript复制// 5.2:组织要写入到kudu表的数据
val data: RDD[Row] = sparkSession.sparkContext.makeRDD(rows)
val quotaDF: DataFrame = sparkSession.createDataFrame(data, schema)
save(quotaDF, OfflineTableDefine.customerSummery)
2.6、完整代码
代码语言:javascript复制package cn.it.logistics.offline.dws
import cn.it.logistics.common.{Configure, DateHelper, OfflineTableDefine, SparkUtils}
import cn.it.logistics.offline.OfflineApp
import cn.it.logistics.offline.dws.ExpressBillDWS.{appName, execute}
import org.apache.spark.SparkConf
import org.apache.spark.rdd.RDD
import org.apache.spark.sql.{DataFrame, Row, SparkSession}
import org.apache.spark.sql.functions._
import org.apache.spark.sql.types.{DoubleType, LongType, Metadata, StringType, StructField, StructType}
import scala.collection.mutable.ArrayBuffer
/**
* 客户主题开发
* 读取客户明细宽表的数据,然后进行指标开发,将结果存储到kudu表中(DWS层)
*/
object ConsumerDWS extends OfflineApp{
//定义应用的名称
val appName: String = this.getClass.getSimpleName
/**
* 入口函数
* @param args
*/
def main(args: Array[String]): Unit = {
/**
* 实现步骤:
* 1)创建sparkConf对象
* 2)创建SparkSession对象
* 3)读取客户宽表数据(判断是全量装载还是增量装载),将加载的数据进行缓存
* 4)对客户明细表的数据进行指标计算
* 5)将计算好的数写入到kudu表中
* 5.1)定义写入kudu表的schema结构信息
* 5.2)将组织好的指标结果集合转换成RDD对象
* 5.3)创建表,写入数据
* 6)删除缓存,释放资源
* 7)停止作业,退出sparkSession
*/
//TODO 1)创建sparkConf对象
val sparkConf: SparkConf = SparkUtils.autoSettingEnv(
SparkUtils.sparkConf(appName),
SparkUtils.parameterParser(args)
)
//TODO 2)创建SparkSession对象
val sparkSession: SparkSession = SparkUtils.getSparkSession(sparkConf)
sparkSession.sparkContext.setLogLevel(Configure.LOG_OFF)
//执行数据处理的逻辑
execute(sparkSession)
}
/**
* 数据处理
*
* @param sparkSession
*/
override def execute(sparkSession: SparkSession): Unit = {
//TODO 3)读取客户明细宽表的数据(用户主题的数据不需要按照天进行增量更新,而是每天全量运行)
val customerDetailDF: DataFrame = getKuduSource(sparkSession, OfflineTableDefine.customerDetail, true)
import sparkSession.implicits._
val schema = StructType(Array(
StructField("id", StringType, true, Metadata.empty),
StructField("customerTotalCount", LongType, true, Metadata.empty),
StructField("addtionTotalCount", LongType, true, Metadata.empty),
StructField("lostCustomerTotalCount", LongType, true, Metadata.empty),
StructField("lostRate", DoubleType, true, Metadata.empty),
StructField("activeCount", LongType, true, Metadata.empty),
StructField("monthOfNewCustomerCount", LongType, true, Metadata.empty),
StructField("sleepCustomerCount", LongType, true, Metadata.empty),
StructField("loseCustomerCount", LongType, true, Metadata.empty),
StructField("customerBillCount", LongType, true, Metadata.empty),
StructField("customerAvgAmount", DoubleType, true, Metadata.empty),
StructField("avgCustomerBillCount", LongType, true, Metadata.empty),
StructField("normalCustomerCount", LongType, true, Metadata.empty)
))
//定义数据集合
val rows: ArrayBuffer[Row] = ArrayBuffer[Row]()
//TODO 4)对客户明细宽表的数据进行指标的计算
val customerTotalCount: Row = customerDetailDF.agg(count($"id").alias("total_count")).first()
//今日新增客户数
val addTotalCount: Long = customerDetailDF.where(date_format($"regDt", "yyyy-MM-dd").equalTo(DateHelper.getyestday("yyyy-MM-dd"))).agg(count($"id")).first().getLong(0)
//留存率(超过180天未下单表示已经流失,否则表示留存)
//留存用户数
//val lostCustomerTotalCount: Long = customerDetailDF.join(customerSenderInfoDF.where("cdt >= date_sub(now(), 180)"), customerDetailDF("id") === customerSenderInfoDF("ciid")).count()
val lostCustomerTotalCount: Long = customerDetailDF.where("last_sender_cdt >= date_sub(now(), 180)").count()
println(lostCustomerTotalCount)
//留存率,超过180天未下单的用户数/所有的用户数
val lostRate: Double = (lostCustomerTotalCount / (if (customerTotalCount.isNullAt(0)) 1D else customerTotalCount.getLong(0))).asInstanceOf[Number].doubleValue()
println(lostRate)
// 活跃用户数(近10天内有发件的客户表示活跃用户)
val activeCount = customerDetailDF.where("last_sender_cdt>=date_sub(now(), 10)").count
// 月度新老用户数(应该是月度新用户!)
val monthOfNewCustomerCount = customerDetailDF.where($"regDt".between(trunc($"regDt", "MM"), date_format(current_date(), "yyyy-MM-dd"))).count
// 沉睡用户数(3个月~6个月之间的用户表示已沉睡)
val sleepCustomerCount = customerDetailDF.where("last_sender_cdt>=date_sub(now(), 180) and last_sender_cdt<=date_sub(now(), 90)").count
println(sleepCustomerCount)
// 流失用户数(9个月未下单表示已流失)
val loseCustomerCount = customerDetailDF.where("last_sender_cdt>=date_sub(now(), 270)").count
println(loseCustomerCount)
// 客单数
val customerSendInfoDF = customerDetailDF.where("first_sender_id is not null")
val customerBillCountAndAmount: Row = customerSendInfoDF.agg(sum("totalCount").alias("totalCount"), sum("totalAmount").alias("totalAmount")).first()
// 客单价
val customerAvgAmount = customerBillCountAndAmount.get(1).toString.toDouble / customerBillCountAndAmount.get(0).toString.toDouble //总金额/总件数
println(customerAvgAmount)
// 平均客单数
val avgCustomerBillCount = customerSendInfoDF.count / customerDetailDF.count
// 普通用户数
val normalCustomerRow: Row = customerDetailDF.where("type=1").agg(count($"id").alias("total_count")).first()
println(normalCustomerRow)
val normalCustomerCount: Long = if (normalCustomerRow.isNullAt(0)) 0L else normalCustomerRow.get(0).asInstanceOf[Number].longValue()
// 获取昨天时间yyyyMMdd
val cdt = DateHelper.getyestday("yyyyMMdd")
// 构建要持久化的指标数据
val rowInfo = Row(
cdt,
if (customerTotalCount.isNullAt(0)) 0L else customerTotalCount.get(0).asInstanceOf[Number].longValue(),
addTotalCount,
lostCustomerTotalCount,
lostRate,
activeCount,
monthOfNewCustomerCount,
sleepCustomerCount,
loseCustomerCount,
if (customerBillCountAndAmount.isNullAt(0)) 0L else customerBillCountAndAmount.get(0).asInstanceOf[Number].longValue(),
customerAvgAmount,
avgCustomerBillCount,
normalCustomerCount
)
rows.append(rowInfo)
// 5.2:组织要写入到kudu表的数据
val data: RDD[Row] = sparkSession.sparkContext.makeRDD(rows)
val quotaDF: DataFrame = sparkSession.createDataFrame(data, schema)
save(quotaDF, OfflineTableDefine.customerSummery)
//删除缓存,释放资源
customerDetailDF.unpersist()
sparkSession.stop()
}
}