spark-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From yanboliang <...@git.apache.org>
Subject [GitHub] spark pull request #12819: [SPARK-14077][ML] Refactor NaiveBayes to support ...
Date Tue, 20 Sep 2016 09:34:04 GMT
Github user yanboliang commented on a diff in the pull request:

    https://github.com/apache/spark/pull/12819#discussion_r79567686
  
    --- Diff: mllib/src/main/scala/org/apache/spark/ml/classification/NaiveBayes.scala ---
    @@ -109,10 +119,88 @@ class NaiveBayes @Since("1.5.0") (
             s" numClasses=$numClasses, but thresholds has length ${$(thresholds).length}")
         }
     
    -    val oldDataset: RDD[OldLabeledPoint] =
    -      extractLabeledPoints(dataset).map(OldLabeledPoint.fromML)
    -    val oldModel = OldNaiveBayes.train(oldDataset, $(smoothing), $(modelType))
    -    NaiveBayesModel.fromOld(oldModel, this)
    +    val numFeatures = dataset.select(col($(featuresCol))).head().getAs[Vector](0).size
    +
    +    val requireNonnegativeValues: Vector => Unit = (v: Vector) => {
    +      val values = v match {
    +        case sv: SparseVector => sv.values
    +        case dv: DenseVector => dv.values
    +      }
    +      if (!values.forall(_ >= 0.0)) {
    +        throw new SparkException(s"Naive Bayes requires nonnegative feature values but
found $v.")
    +      }
    +    }
    +
    +    val requireZeroOneBernoulliValues: Vector => Unit = (v: Vector) => {
    +      val values = v match {
    +        case sv: SparseVector => sv.values
    +        case dv: DenseVector => dv.values
    +      }
    +      if (!values.forall(v => v == 0.0 || v == 1.0)) {
    +        throw new SparkException(
    +          s"Bernoulli naive Bayes requires 0 or 1 feature values but found $v.")
    +      }
    +    }
    +
    +    val requireValues: Vector => Unit = {
    +      $(modelType) match {
    +        case Multinomial =>
    +          requireNonnegativeValues
    +        case Bernoulli =>
    +          requireZeroOneBernoulliValues
    +        case _ =>
    +          // This should never happen.
    +          throw new UnknownError(s"Invalid modelType: ${$(modelType)}.")
    +      }
    +    }
    +
    +    val w = if (!isDefined(weightCol) || $(weightCol).isEmpty) lit(1.0) else col($(weightCol))
    +
    +    val aggregated = dataset.select(col($(labelCol)).cast(DoubleType), w, col($(featuresCol))).rdd
    +      .map { row => (row.getDouble(0), (row.getDouble(1), row.getAs[Vector](2)))
    +      }.aggregateByKey[(Double, DenseVector)]((0.0, Vectors.zeros(numFeatures).toDense))(
    +      seqOp = {
    +         case (agg, (weight, features)) =>
    +           requireValues(features)
    +           BLAS.axpy(weight, features, agg._2)
    +           (agg._1 + weight, agg._2)
    +      },
    +      combOp = {
    +         case (agg1, agg2) =>
    +           BLAS.axpy(1.0, agg2._2, agg1._2)
    +           (agg1._1 + agg2._1, agg1._2)
    +      }).collect().sortBy(_._1)
    +
    +    val numLabels = aggregated.length
    +    val numDocuments = aggregated.map(_._2._1).sum
    +
    +    val pi = Array.fill[Double](numLabels)(0.0)
    +    val theta = Array.fill[Double](numLabels, numFeatures)(0.0)
    +
    +    val lambda = $(smoothing)
    +    val piLogDenom = math.log(numDocuments + numLabels * lambda)
    +    var i = 0
    +    aggregated.foreach { case (label, (n, sumTermFreqs)) =>
    +      pi(i) = math.log(n + lambda) - piLogDenom
    +      val thetaLogDenom = $(modelType) match {
    +        case Multinomial => math.log(sumTermFreqs.values.sum + numFeatures * lambda)
    +        case Bernoulli => math.log(n + 2.0 * lambda)
    +        case _ =>
    +          // This should never happen.
    +          throw new UnknownError(s"Invalid modelType: ${$(modelType)}.")
    +      }
    +      var j = 0
    +      while (j < numFeatures) {
    +        theta(i)(j) = math.log(sumTermFreqs(j) + lambda) - thetaLogDenom
    +        j += 1
    +      }
    +      i += 1
    +    }
    +
    +    val uid = Identifiable.randomUID("nb")
    +    val piVector = Vectors.dense(pi)
    +    val thetaMatrix = new DenseMatrix(numLabels, theta(0).length, theta.flatten, true)
    --- End diff --
    
    ```pi -> piArray, piVector -> pi, theta -> thetaArrays, thetaMatrix -> theta```
should be better?


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org


Mime
View raw message