com.intel.analytics.zoo.models.anomalydetection.Utils.scala Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of analytics-zoo-bigdl_0.8.0-spark_2.2.1 Show documentation
The newest version!
/*
 * Copyright 2018 Analytics Zoo Authors.
 *
 * Licensed 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 com.intel.analytics.zoo.models.anomalydetection

import com.intel.analytics.bigdl.dataset.Sample
import org.apache.spark.rdd.RDD
import org.apache.spark.sql.DataFrame
import org.apache.spark.sql.functions.{avg, col, udf}

object Utils {

  def standardScaleHelper(df: DataFrame, colName: String): DataFrame = {

    val mean = df.select(colName).agg(avg(col(colName))).collect()(0).getDouble(0)

    val stddevUdf = udf((num: Float) => (num - mean) * (num - mean))

    val stddev = Math.sqrt(df.withColumn("stddev", stddevUdf(col(colName)))
      .agg(avg(col("stddev"))).collect()(0).getDouble(0))

    val scaledUdf = udf((num: Float) => ((num - mean) / stddev).toFloat)

    df.withColumn(colName, scaledUdf(col(colName)))
  }

  def standardScale(df: DataFrame, fields: Seq[String], index: Int = 0): DataFrame = {

    if (index == fields.length) {
      df
    } else {
      val colDf = standardScaleHelper(df, fields(index))
      standardScale(colDf, fields, index + 1)
    }
  }

  def trainTestSplit(unrolled: RDD[FeatureLabelIndex[Float]], testSize: Int = 1000):
  (RDD[Sample[Float]], RDD[Sample[Float]]) = {

    val cutPoint = unrolled.count() - testSize

    val train = AnomalyDetector.toSampleRdd(unrolled.filter(x => x.index < cutPoint))
    val test = AnomalyDetector.toSampleRdd(unrolled.filter(x => x.index >= cutPoint))

    (train, test)
  }

  def trainTestSplit(unrolled: RDD[FeatureLabelIndex[Float]], testSize: Float)
  : (RDD[Sample[Float]], RDD[Sample[Float]]) = {

    val totalSize = unrolled.count()
    val testSizeInt = (totalSize * testSize).toInt

    trainTestSplit(unrolled, testSizeInt)
  }

}