Code cleanups and warnings

samoa-api/src/main/java/org/apache/samoa/learners/classifiers/ensemble/AdaptiveBagging.java

@@ -66,9 +66,6 @@ public class AdaptiveBagging implements Learner, Configurable {
   /** The distributor processor. */
   private BaggingDistributorProcessor distributorP;
 
-  /** The input streams for the ensemble, one per member. */
-  private Stream[] ensembleStreams;
-
   /** The result stream. */
   protected Stream resultStream;
 
@@ -116,7 +113,8 @@ protected void setLayout() {
       }
     }
 
-    ensembleStreams = new Stream[ensembleSize];
+    /* The input streams for the ensemble, one per member. */
+    Stream[] ensembleStreams = new Stream[ensembleSize];
     for (int i = 0; i < ensembleSize; i++) {
       ensembleStreams[i] = builder.createStream(distributorP);
       builder.connectInputShuffleStream(ensembleStreams[i], ensemble[i].getInputProcessor()); // connect streams one-to-one with ensemble members (the type of connection does not matter)

samoa-api/src/main/java/org/apache/samoa/learners/classifiers/ensemble/Boosting.java

@@ -103,9 +103,6 @@ protected void setLayout() {
     Stream predictionStream = this.builder.createStream(distributorP);
     this.builder.connectInputKeyStream(predictionStream, classifier.getInputProcessor());
 
-//    distributorP.setOutputStream(testingStream);
-//    distributorP.setPredictionStream(predictionStream);
-
     // Addition to Bagging: stream to train
     /* The training stream. */
     Stream trainingStream = this.builder.createStream(predictionCombinerP);

samoa-api/src/main/java/org/apache/samoa/learners/classifiers/rules/centralized/AMRulesRegressorProcessor.java

@@ -163,7 +163,7 @@ private double[] getVotesForInstance(Instance instance) {
     int numberOfRulesCovering = 0;
 
     for (ActiveRule rule : ruleSet) {
-      if (rule.isCovering(instance) == true) {
+      if (rule.isCovering(instance)) {
         numberOfRulesCovering++;
         double[] vote = rule.getPrediction(instance);
         double error = rule.getCurrentError();
@@ -179,9 +179,8 @@ private double[] getVotesForInstance(Instance instance) {
       double error = defaultRule.getCurrentError();
       errorWeightedVote.addVote(vote, error);
     }
-    double[] weightedVote = errorWeightedVote.computeWeightedVote();
 
-    return weightedVote;
+    return errorWeightedVote.computeWeightedVote();
   }
 
   public ErrorWeightedVote newErrorWeightedVote() {
@@ -223,13 +222,13 @@ public void trainOnInstanceImpl(Instance instance) {
     Iterator<ActiveRule> ruleIterator = this.ruleSet.iterator();
     while (ruleIterator.hasNext()) {
       ActiveRule rule = ruleIterator.next();
-      if (rule.isCovering(instance) == true) {
+      if (rule.isCovering(instance)) {
         rulesCoveringInstance = true;
-        if (isAnomaly(instance, rule) == false) {
+        if (!isAnomaly(instance, rule)) {
           // Update Change Detection Tests
           double error = rule.computeError(instance); // Use adaptive mode error
           boolean changeDetected = ((RuleActiveRegressionNode) rule.getLearningNode()).updateChangeDetection(error);
-          if (changeDetected == true) {
+          if (changeDetected) {
             ruleIterator.remove();
           } else {
             rule.updateStatistics(instance);
@@ -245,10 +244,10 @@ public void trainOnInstanceImpl(Instance instance) {
       }
     }
 
-    if (rulesCoveringInstance == false) {
+    if (!rulesCoveringInstance) {
       defaultRule.updateStatistics(instance);
       if (defaultRule.getInstancesSeen() % this.gracePeriod == 0.0) {
-        if (defaultRule.tryToExpand(this.splitConfidence, this.tieThreshold) == true) {
+        if (defaultRule.tryToExpand(this.splitConfidence, this.tieThreshold)) {
           ActiveRule newDefaultRule = newRule(defaultRule.getRuleNumberID(),
               (RuleActiveRegressionNode) defaultRule.getLearningNode(),
               ((RuleActiveRegressionNode) defaultRule.getLearningNode()).getStatisticsOtherBranchSplit()); // other branch
@@ -274,7 +273,7 @@ private boolean isAnomaly(Instance instance, ActiveRule rule) {
     // AMRUles is equipped with anomaly detection. If on, compute the anomaly
     // value.
     boolean isAnomaly = false;
-    if (this.noAnomalyDetection == false) {
+    if (!this.noAnomalyDetection) {
       if (rule.getInstancesSeen() >= this.anomalyNumInstThreshold) {
         isAnomaly = rule.isAnomaly(instance,
             this.univariateAnomalyprobabilityThreshold,
@@ -308,19 +307,19 @@ private ActiveRule newRule(int ID, RuleActiveRegressionNode node, double[] stati
         }
       }
     }
-    if (statistics != null && ((RuleActiveRegressionNode) r.getLearningNode()).getTargetMean() != null)
+    if (statistics != null && (r.getLearningNode()).getTargetMean() != null)
     {
       double mean;
       if (statistics[0] > 0) {
         mean = statistics[1] / statistics[0];
-        ((RuleActiveRegressionNode) r.getLearningNode()).getTargetMean().reset(mean, (long) statistics[0]);
+        (r.getLearningNode()).getTargetMean().reset(mean, (long) statistics[0]);
       }
     }
     return r;
   }
 
   private ActiveRule newRule(int ID) {
-    ActiveRule r = new ActiveRule.Builder().
+    return new ActiveRule.Builder().
         threshold(this.pageHinckleyThreshold).
         alpha(this.pageHinckleyAlpha).
         changeDetection(this.driftDetection).
@@ -329,7 +328,6 @@ private ActiveRule newRule(int ID) {
         learningRatio(this.learningRatio).
         numericObserver(numericObserver).
         id(ID).build();
-    return r;
   }
 
   /*
@@ -342,7 +340,7 @@ public void onCreate(int id) {
     this.ruleNumberID = 0;
     this.defaultRule = newRule(++this.ruleNumberID);
 
-    this.ruleSet = new LinkedList<ActiveRule>();
+    this.ruleSet = new LinkedList<>();
   }
 
   /*

samoa-api/src/main/java/org/apache/samoa/learners/classifiers/rules/common/ActiveRule.java

@@ -38,8 +38,6 @@ public class ActiveRule extends LearningRule {
 
   private static final long serialVersionUID = 1L;
 
-  private double[] statisticsOtherBranchSplit;
-
   private Builder builder;
 
   private RuleActiveRegressionNode learningNode;
@@ -89,10 +87,6 @@ public void setLearningNode(RuleRegressionNode learningNode) {
     this.learningNode = (RuleActiveRegressionNode) learningNode;
   }
 
-  public double[] statisticsOtherBranchSplit() {
-    return this.statisticsOtherBranchSplit;
-  }
-
   public RuleSplitNode getLastUpdatedRuleSplitNode() {
     return this.lastUpdatedRuleSplitNode;
   }
@@ -104,7 +98,6 @@ public static class Builder implements Serializable {
 
     private static final long serialVersionUID = 1712887264918475622L;
     protected boolean changeDetection;
-    protected boolean usePerceptron;
     protected double threshold;
     protected double alpha;
     protected int predictionFunction;
@@ -115,8 +108,6 @@ public static class Builder implements Serializable {
 
     protected FIMTDDNumericAttributeClassLimitObserver numericObserver;
 
-    protected double lastTargetMean;
-
     public int id;
 
     public Builder() {
@@ -182,8 +173,7 @@ public ActiveRule build() {
    */
   public boolean tryToExpand(double splitConfidence, double tieThreshold) {
 
-    boolean shouldSplit = this.learningNode.tryToExpand(splitConfidence, tieThreshold);
-    return shouldSplit;
+    return this.learningNode.tryToExpand(splitConfidence, tieThreshold);
 
   }
 

samoa-api/src/main/java/org/apache/samoa/learners/classifiers/rules/distributed/AMRLearnerProcessor.java

@@ -50,10 +50,7 @@ public class AMRLearnerProcessor implements Processor {
 	 */
   private static final long serialVersionUID = -2302897295090248013L;
 
-  private static final Logger logger =
-      LoggerFactory.getLogger(AMRLearnerProcessor.class);
-
-  private int processorId;
+  private static final Logger logger = LoggerFactory.getLogger(AMRLearnerProcessor.class);
 
   private transient List<ActiveRule> ruleSet;
 
@@ -105,10 +102,10 @@ private void trainRuleOnInstance(int ruleID, Instance instance) {
       ActiveRule rule = ruleIterator.next();
       if (rule.getRuleNumberID() == ruleID) {
         // Check (again) for coverage
-        if (rule.isCovering(instance) == true) {
+        if (rule.isCovering(instance)) {
           double error = rule.computeError(instance); // Use adaptive mode error
           boolean changeDetected = ((RuleActiveRegressionNode) rule.getLearningNode()).updateChangeDetection(error);
-          if (changeDetected == true) {
+          if (changeDetected) {
             ruleIterator.remove();
 
             this.sendRemoveRuleEvent(ruleID);
@@ -137,7 +134,7 @@ private boolean isAnomaly(Instance instance, LearningRule rule) {
     // AMRUles is equipped with anomaly detection. If on, compute the anomaly
     // value.
     boolean isAnomaly = false;
-    if (this.noAnomalyDetection == false) {
+    if (!this.noAnomalyDetection) {
       if (rule.getInstancesSeen() >= this.anomalyNumInstThreshold) {
         isAnomaly = rule.isAnomaly(instance,
             this.univariateAnomalyprobabilityThreshold,
@@ -167,8 +164,7 @@ private boolean addRule(ActiveRule rule) {
 
   @Override
   public void onCreate(int id) {
-    this.processorId = id;
-    this.ruleSet = new LinkedList<ActiveRule>();
+    this.ruleSet = new LinkedList<>();
   }
 
   @Override

samoa-api/src/main/java/org/apache/samoa/learners/classifiers/rules/distributed/AMRulesStatisticsProcessor.java

@@ -49,10 +49,7 @@ public class AMRulesStatisticsProcessor implements Processor {
 	 */
   private static final long serialVersionUID = 5268933189695395573L;
 
-  private static final Logger logger =
-      LoggerFactory.getLogger(AMRulesStatisticsProcessor.class);
-
-  private int processorId;
+  private static final Logger logger = LoggerFactory.getLogger(AMRulesStatisticsProcessor.class);
 
   private transient List<ActiveRule> ruleSet;
 
@@ -98,10 +95,10 @@ private void trainRuleOnInstance(int ruleID, Instance instance) {
       if (rule.getRuleNumberID() == ruleID) {
         // Check (again) for coverage
         // Skip anomaly check as Aggregator's perceptron should be well-updated
-        if (rule.isCovering(instance) == true) {
+        if (rule.isCovering(instance)) {
           double error = rule.computeError(instance); // Use adaptive mode error
           boolean changeDetected = ((RuleActiveRegressionNode) rule.getLearningNode()).updateChangeDetection(error);
-          if (changeDetected == true) {
+          if (changeDetected) {
             ruleIterator.remove();
 
             this.sendRemoveRuleEvent(ruleID);
@@ -143,8 +140,7 @@ private boolean addRule(ActiveRule rule) {
 
   @Override
   public void onCreate(int id) {
-    this.processorId = id;
-    this.ruleSet = new LinkedList<ActiveRule>();
+    this.ruleSet = new LinkedList<>();
   }
 
   @Override

samoa-api/src/main/java/org/apache/samoa/learners/clusterers/LocalClustererProcessor.java

@@ -45,8 +45,7 @@ final public class LocalClustererProcessor implements Processor {
      *
      */
   private static final long serialVersionUID = -1577910988699148691L;
-  private static final Logger logger = LoggerFactory
-      .getLogger(LocalClustererProcessor.class);
+  private static final Logger logger = LoggerFactory.getLogger(LocalClustererProcessor.class);
   private LocalClustererAdapter model;
   private Stream outputStream;
   private int modelId;
@@ -84,7 +83,7 @@ public LocalClustererAdapter getLearner() {
    * Set the output streams.
    * 
    * @param outputStream
-   *          the new output stream {@link PredictionCombinerPE}.
+   *          the new output stream {@link Stream}.
    */
   public void setOutputStream(Stream outputStream) {
 

samoa-api/src/main/java/org/apache/samoa/learners/clusterers/SingleLearner.java

@@ -69,7 +69,7 @@ public void init(TopologyBuilder builder, Instances dataset, int parallelism) {
 
   protected void setLayout() {
     learnerP = new LocalClustererProcessor();
-    LocalClustererAdapter learner = (LocalClustererAdapter) this.learnerOption.getValue();
+    LocalClustererAdapter learner = this.learnerOption.getValue();
     learner.setDataset(this.dataset);
     learnerP.setLearner(learner);
 
@@ -96,7 +96,6 @@ public Processor getInputProcessor() {
    */
   @Override
   public Set<Stream> getResultStreams() {
-    Set<Stream> streams = ImmutableSet.of(this.resultStream);
-    return streams;
+    return ImmutableSet.of(this.resultStream);
   }
 }

samoa-local/src/test/java/org/apache/samoa/AlgosTest.java

@@ -20,7 +20,6 @@
  * #L%
  */
 
-import org.apache.samoa.LocalDoTask;
 import org.junit.Test;
 
 public class AlgosTest {