5.道路卡口摄像头的异常状态统计

package camera

import org.apache.spark.SparkConf
import org.apache.spark.sql.SparkSession

// 摄像异常状态的功能代码
object CameraAbnormality {
  def main(args: Array[String]): Unit = {
    val conf = new SparkConf().setAppName("camera").setMaster("local[*]")
    val session = SparkSession.builder().config(conf).getOrCreate()

    val map = Map[String, String]("mode" -> "FAILFAST", "inferSchema" -> "true")

    val frame1 = session.read.options(map).csv("hdfs://HadoopCluster/cameras.csv")
    // 卡口摄像头数据
    // 卡口id: 区域id-街道id-卡口编号
    // 摄像头id: 1-8
    val cameras = frame1.toDF("monitorId", "cameraId")

    val frame2 = session.read.options(map).csv("hdfs://HadoopCluster/project/20220913/*")
    // 车流量信息
    // 车辆通过时间  卡口编号  摄像头编号  通过车辆车牌  某个摄像头拍摄时间 单位:秒  通过卡口的速度  道路id  区域id
    // 2022-09-13,  5-23-63,        3,      晋KIFWNQ,      2022-09-13 15:47:47,        17,        23,      5
    // todayDate  monitorId  camera_id    chepai           actionTime             speed      roadId  areaId
    val flow = frame2.toDF("todayDate", "monitorId", "cameraId", "chepai", "actionTime", "speed", "roadId", "areaId")
    cameras.createTempView("cameras")
    flow.createTempView("flow")

//    cameras.show(10)
//    flow.show(10)

    val frame = session.sql(
      "select a.monitorId, a.cameraId " +
        "from cameras as a " +
        "left join flow as b " +
        "on a.monitorId = b.monitorId and a.cameraId = b.cameraId " +
        "where b.chepai is null " +
        "order by a.monitorId")
    frame.show(200)

    session.stop()
  }
}