spark StreamingTestExample 源码

spark StreamingTestExample 代码

文件路径：/examples/src/main/scala/org/apache/spark/examples/mllib/StreamingTestExample.scala

/*
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *    http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package org.apache.spark.examples.mllib

import org.apache.spark.SparkConf
import org.apache.spark.mllib.stat.test.{BinarySample, StreamingTest}
import org.apache.spark.streaming.{Seconds, StreamingContext}
import org.apache.spark.util.Utils

/**
 * Perform streaming testing using Welch's 2-sample t-test on a stream of data, where the data
 * stream arrives as text files in a directory. Stops when the two groups are statistically
 * significant (p-value < 0.05) or after a user-specified timeout in number of batches is exceeded.
 *
 * The rows of the text files must be in the form `Boolean, Double`. For example:
 *   false, -3.92
 *   true, 99.32
 *
 * Usage:
 *   StreamingTestExample <dataDir> <batchDuration> <numBatchesTimeout>
 *
 * To run on your local machine using the directory `dataDir` with 5 seconds between each batch and
 * a timeout after 100 insignificant batches, call:
 *    $ bin/run-example mllib.StreamingTestExample dataDir 5 100
 *
 * As you add text files to `dataDir` the significance test wil continually update every
 * `batchDuration` seconds until the test becomes significant (p-value < 0.05) or the number of
 * batches processed exceeds `numBatchesTimeout`.
 */
object StreamingTestExample {

  def main(args: Array[String]): Unit = {
    if (args.length != 3) {
      // scalastyle:off println
      System.err.println(
        "Usage: StreamingTestExample " +
          "<dataDir> <batchDuration> <numBatchesTimeout>")
      // scalastyle:on println
      System.exit(1)
    }
    val dataDir = args(0)
    val batchDuration = Seconds(args(1).toLong)
    val numBatchesTimeout = args(2).toInt

    val conf = new SparkConf().setMaster("local").setAppName("StreamingTestExample")
    val ssc = new StreamingContext(conf, batchDuration)
    ssc.checkpoint {
      val dir = Utils.createTempDir()
      dir.toString
    }

    // $example on$
    val data = ssc.textFileStream(dataDir).map(line => line.split(",") match {
      case Array(label, value) => BinarySample(label.toBoolean, value.toDouble)
    })

    val streamingTest = new StreamingTest()
      .setPeacePeriod(0)
      .setWindowSize(0)
      .setTestMethod("welch")

    val out = streamingTest.registerStream(data)
    out.print()
    // $example off$

    // Stop processing if test becomes significant or we time out
    var timeoutCounter = numBatchesTimeout
    out.foreachRDD { rdd =>
      timeoutCounter -= 1
      val anySignificant = rdd.map(_.pValue < 0.05).fold(false)(_ || _)
      if (timeoutCounter == 0 || anySignificant) rdd.context.stop()
    }

    ssc.start()
    ssc.awaitTermination()
  }
}

相关信息

spark 源码目录

相关文章

spark AbstractParams 源码

spark AssociationRulesExample 源码

spark BinaryClassification 源码

spark BinaryClassificationMetricsExample 源码

spark BisectingKMeansExample 源码

spark ChiSqSelectorExample 源码

spark Correlations 源码

spark CorrelationsExample 源码

spark CosineSimilarity 源码

spark DecisionTreeClassificationExample 源码