SPARK-2341 [MLLIB] loadLibSVMFile doesn't handle regression datasets

Per discussion at https://issues.apache.org/jira/browse/SPARK-2341 , this is a look at deprecating the multiclass parameter. Thoughts welcome of course.

Author: Sean Owen <srowen@gmail.com>

Closes #1663 from srowen/SPARK-2341 and squashes the following commits:

8a3abd7 [Sean Owen] Suppress MIMA error for removed package private classes
18a8c8e [Sean Owen] Updates from review
83d0092 [Sean Owen] Deprecated methods with multiclass, and instead always parse target as a double (ie. multiclass = true)

2 files changed, 16 insertions, 92 deletions

diff --git a/mllib/src/main/scala/org/apache/spark/mllib/util/LabelParsers.scala b/mllib/src/main/scala/org/apache/spark/mllib/util/LabelParsers.scala
deleted file mode 100644
index e25bf18b78..0000000000
--- a/mllib/src/main/scala/org/apache/spark/mllib/util/LabelParsers.scala
+++ /dev/null
@@ -1,56 +0,0 @@
-/*
- * 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.mllib.util
-
-/** Trait for label parsers. */
-private trait LabelParser extends Serializable {
-  /** Parses a string label into a double label. */
-  def parse(labelString: String): Double
-}
-
-/** Factory methods for label parsers. */
-private object LabelParser {
-  def getInstance(multiclass: Boolean): LabelParser = {
-    if (multiclass) MulticlassLabelParser else BinaryLabelParser
-  }
-}
-
-/**
- * Label parser for binary labels, which outputs 1.0 (positive) if the value is greater than 0.5,
- * or 0.0 (negative) otherwise. So it works with +1/-1 labeling and +1/0 labeling.
- */
-private object BinaryLabelParser extends LabelParser {
-  /** Gets the default instance of BinaryLabelParser. */
-  def getInstance(): LabelParser = this
-
-  /**
-   * Parses the input label into positive (1.0) if the value is greater than 0.5,
-   * or negative (0.0) otherwise.
-   */
-  override def parse(labelString: String): Double = if (labelString.toDouble > 0.5) 1.0 else 0.0
-}
-
-/**
- * Label parser for multiclass labels, which converts the input label to double.
- */
-private object MulticlassLabelParser extends LabelParser {
-  /** Gets the default instance of MulticlassLabelParser. */
-  def getInstance(): LabelParser = this
-
-  override def parse(labelString: String): Double =  labelString.toDouble
-}
diff --git a/mllib/src/main/scala/org/apache/spark/mllib/util/MLUtils.scala b/mllib/src/main/scala/org/apache/spark/mllib/util/MLUtils.scala
index 30de24ad89..dc10a19478 100644
--- a/mllib/src/main/scala/org/apache/spark/mllib/util/MLUtils.scala
+++ b/mllib/src/main/scala/org/apache/spark/mllib/util/MLUtils.scala
@@ -55,7 +55,6 @@ object MLUtils {
    *
    * @param sc Spark context
    * @param path file or directory path in any Hadoop-supported file system URI
-   * @param labelParser parser for labels
    * @param numFeatures number of features, which will be determined from the input data if a
    *                    nonpositive value is given. This is useful when the dataset is already split
    *                    into multiple files and you want to load them separately, because some
@@ -64,10 +63,9 @@ object MLUtils {
    * @param minPartitions min number of partitions
    * @return labeled data stored as an RDD[LabeledPoint]
    */
-  private def loadLibSVMFile(
+  def loadLibSVMFile(
       sc: SparkContext,
       path: String,
-      labelParser: LabelParser,
       numFeatures: Int,
       minPartitions: Int): RDD[LabeledPoint] = {
     val parsed = sc.textFile(path, minPartitions)
@@ -75,7 +73,7 @@ object MLUtils {
       .filter(line => !(line.isEmpty || line.startsWith("#")))
       .map { line =>
         val items = line.split(' ')
-        val label = labelParser.parse(items.head)
+        val label = items.head.toDouble
         val (indices, values) = items.tail.map { item =>
           val indexAndValue = item.split(':')
           val index = indexAndValue(0).toInt - 1 // Convert 1-based indices to 0-based.
@@ -102,36 +100,14 @@ object MLUtils {
 
   // Convenient methods for `loadLibSVMFile`.
 
-  /**
-   * Loads labeled data in the LIBSVM format into an RDD[LabeledPoint].
-   * The LIBSVM format is a text-based format used by LIBSVM and LIBLINEAR.
-   * Each line represents a labeled sparse feature vector using the following format:
-   * {{{label index1:value1 index2:value2 ...}}}
-   * where the indices are one-based and in ascending order.
-   * This method parses each line into a [[org.apache.spark.mllib.regression.LabeledPoint]],
-   * where the feature indices are converted to zero-based.
-   *
-   * @param sc Spark context
-   * @param path file or directory path in any Hadoop-supported file system URI
-   * @param multiclass whether the input labels contain more than two classes. If false, any label
-   *                   with value greater than 0.5 will be mapped to 1.0, or 0.0 otherwise. So it
-   *                   works for both +1/-1 and 1/0 cases. If true, the double value parsed directly
-   *                   from the label string will be used as the label value.
-   * @param numFeatures number of features, which will be determined from the input data if a
-   *                    nonpositive value is given. This is useful when the dataset is already split
-   *                    into multiple files and you want to load them separately, because some
-   *                    features may not present in certain files, which leads to inconsistent
-   *                    feature dimensions.
-   * @param minPartitions min number of partitions
-   * @return labeled data stored as an RDD[LabeledPoint]
-   */
-   def loadLibSVMFile(
+  @deprecated("use method without multiclass argument, which no longer has effect", "1.1.0")
+  def loadLibSVMFile(
       sc: SparkContext,
       path: String,
       multiclass: Boolean,
       numFeatures: Int,
       minPartitions: Int): RDD[LabeledPoint] =
-    loadLibSVMFile(sc, path, LabelParser.getInstance(multiclass), numFeatures, minPartitions)
+    loadLibSVMFile(sc, path, numFeatures, minPartitions)
 
   /**
    * Loads labeled data in the LIBSVM format into an RDD[LabeledPoint], with the default number of
@@ -140,26 +116,30 @@ object MLUtils {
   def loadLibSVMFile(
       sc: SparkContext,
       path: String,
+      numFeatures: Int): RDD[LabeledPoint] =
+    loadLibSVMFile(sc, path, numFeatures, sc.defaultMinPartitions)
+
+  @deprecated("use method without multiclass argument, which no longer has effect", "1.1.0")
+  def loadLibSVMFile(
+      sc: SparkContext,
+      path: String,
       multiclass: Boolean,
       numFeatures: Int): RDD[LabeledPoint] =
-    loadLibSVMFile(sc, path, multiclass, numFeatures, sc.defaultMinPartitions)
+    loadLibSVMFile(sc, path, numFeatures)
 
-  /**
-   * Loads labeled data in the LIBSVM format into an RDD[LabeledPoint], with the number of features
-   * determined automatically and the default number of partitions.
-   */
+  @deprecated("use method without multiclass argument, which no longer has effect", "1.1.0")
   def loadLibSVMFile(
       sc: SparkContext,
       path: String,
       multiclass: Boolean): RDD[LabeledPoint] =
-    loadLibSVMFile(sc, path, multiclass, -1, sc.defaultMinPartitions)
+    loadLibSVMFile(sc, path)
 
   /**
    * Loads binary labeled data in the LIBSVM format into an RDD[LabeledPoint], with number of
    * features determined automatically and the default number of partitions.
    */
   def loadLibSVMFile(sc: SparkContext, path: String): RDD[LabeledPoint] =
-    loadLibSVMFile(sc, path, multiclass = false, -1, sc.defaultMinPartitions)
+    loadLibSVMFile(sc, path, -1)
 
   /**
    * Save labeled data in LIBSVM format.