Test reference files missed in PR #261 and #258 + support to set DJL Engine random seed

moa/src/main/java/moa/classifiers/deeplearning/CAND.java

@@ -160,6 +160,12 @@ public class CAND extends AbstractClassifier implements MultiClassClassifier, Ca
             "Stats dump file name",
             "" );
 
+    public IntOption djlRandomSeed = new IntOption(
+            "djlRandomSeed",
+            'S',
+            "Random seed for DJL Engine",
+            10, 0, Integer.MAX_VALUE);
+
     @Override
     public void resetLearningImpl() {
         if (nn != null) {
@@ -428,6 +434,7 @@ class MLPConfigs{
             this.nn[i].numberOfLayers.setValue(numberOfLayersInEachMLP.getValue());
             this.nn[i].deltaForADWIN = nnConfigs[i].deltaForADWIN;
             this.nn[i].backPropLossThreshold.setValue(backPropLossThreshold.getValue());
+            this.nn[i].djlRandomSeed.setValue(djlRandomSeed.getValue());
             this.nn[i].initializeNetwork(instance);
         }
 

moa/src/main/java/moa/classifiers/deeplearning/MLP.java

@@ -31,6 +31,7 @@
 import com.github.javacliparser.FloatOption;
 import com.github.javacliparser.MultiChoiceOption;
 
+import ai.djl.engine.Engine;
 import ai.djl.Device;
 import ai.djl.Model;
 import ai.djl.basicmodelzoo.basic.Mlp;
@@ -48,9 +49,7 @@
 
 import java.text.DecimalFormat;
 import java.lang.Math;
-import java.util.ArrayList;
-import java.util.Arrays;
-import java.util.List;
+import java.util.*;
 
 import static ai.djl.ndarray.types.DataType.FLOAT32;
 
@@ -220,6 +219,12 @@ public static class NormalizeInfo{
 			new String[]{"GPU (use CPU if not available)", "CPU"},
 			deviceTypeOptionCPU);
 
+	public IntOption djlRandomSeed = new IntOption(
+			"djlRandomSeed",
+			'S',
+			"Random seed for DJL Engine",
+			10, 0, Integer.MAX_VALUE);
+
 
 	public double deltaForADWIN = 1.0E-5;
 
@@ -443,6 +448,12 @@ public void initializeNetwork(Instance inst) {
 			return;
 		}
 
+		Set<String> engines = Engine.getAllEngines();
+		Iterator<String> engineIterator = engines.iterator();
+		while (engineIterator.hasNext()){
+			Engine.getEngine(engineIterator.next()).setRandomSeed(djlRandomSeed.getValue());
+		}
+
 		votes = new double [inst.numClasses()];
 
 		if (useNormalization.isSet() || useOneHotEncode.isSet()) {

moa/src/main/java/moa/classifiers/meta/imbalanced/OnlineAdaBoost.java

@@ -227,7 +227,9 @@ protected void adjustEnsembleSize(int nClasses) {
     		for (int i = this.nEstimators; i < nClasses; i++) {
     			this.ensemble.add(this.baseLearner.copy()); 
     			this.nEstimators ++;
-    			this.adwinEnsemble.add(new ADWIN());
+                if (this.driftDetection) {
+                    this.adwinEnsemble.add(new ADWIN());
+                }
     			this.lambdaSc.add(0.0);
                 this.lambdaSw.add(0.0);
                 this.epsilon.add(0.0);               

moa/src/main/java/moa/classifiers/meta/imbalanced/OnlineAdaC2.java

@@ -256,7 +256,9 @@ protected void adjustEnsembleSize(int nClasses) {
     		for (int i = this.nEstimators; i < nClasses; i++) {
     			this.ensemble.add(this.baseLearner.copy()); 
     			this.nEstimators ++;
-    			this.adwinEnsemble.add(new ADWIN());
+                if (this.driftDetection) {
+                    this.adwinEnsemble.add(new ADWIN());
+                }
     			this.lambdaTP.add(0.0);
                 this.lambdaTN.add(0.0);
                 this.lambdaFP.add(0.0);

moa/src/main/java/moa/classifiers/meta/imbalanced/OnlineCSB2.java

@@ -249,7 +249,9 @@ protected void adjustEnsembleSize(int nClasses) {
     		for (int i = this.nEstimators; i < nClasses; i++) {
     			this.ensemble.add(this.baseLearner.copy()); 
     			this.nEstimators ++;
-    			this.adwinEnsemble.add(new ADWIN());    			
+                if (this.driftDetection) {
+                    this.adwinEnsemble.add(new ADWIN());
+                }
     			this.lambdaFP.add(0.0);
                 this.lambdaFN.add(0.0);
                 this.lambdaSum.add(0.0);

moa/src/main/java/moa/classifiers/meta/imbalanced/OnlineRUSBoost.java

@@ -291,7 +291,9 @@ protected void adjustEnsembleSize(int nClasses) {
     		for (int i = this.nEstimators; i < nClasses; i++) {
     			this.ensemble.add(this.baseLearner.copy()); 
     			this.nEstimators ++;
-    			this.adwinEnsemble.add(new ADWIN());
+				if (this.driftDetection) {
+					this.adwinEnsemble.add(new ADWIN());
+				}
     			this.lambdaSc.add(0.0);
                 this.lambdaPos.add(0.0);
                 this.lambdaNeg.add(0.0);

moa/src/main/java/moa/classifiers/meta/imbalanced/OnlineSMOTEBagging.java

@@ -231,7 +231,9 @@ protected void adjustEnsembleSize(int nClasses) {
     		for (int i = this.nEstimators; i < nClasses; i++) {
     			this.ensemble.add(this.baseLearner.copy()); 
     			this.nEstimators ++;
-    			this.adwinEnsemble.add(new ADWIN());    			
+				if (this.driftDetection) {
+					this.adwinEnsemble.add(new ADWIN());
+				}
 			}
     	}
     }

moa/src/main/java/moa/classifiers/meta/imbalanced/OnlineUnderOverBagging.java

@@ -209,7 +209,9 @@ protected void adjustEnsembleSize(int nClasses) {
     		for (int i = this.nEstimators; i < nClasses; i++) {
     			this.ensemble.add(this.baseLearner.copy()); 
     			this.nEstimators ++;
-    			this.adwinEnsemble.add(new ADWIN());    			
+                if (this.driftDetection) {
+                    this.adwinEnsemble.add(new ADWIN());
+                }
 			}
     	}
     }

moa/src/test/resources/moa/classifiers/deeplearning/CAND.ref

@@ -0,0 +1,145 @@
+--> classification-out0.arff
+moa.classifiers.deeplearning.CAND -o 2 -B 4 -h -n
+
+Index
+  10000
+Votes
+  0: 0.00238989
+  1: -0.00239092
+Measurements
+  classified instances: 9999
+  classifications correct (percent): 58.04580458
+  Kappa Statistic (percent): 0.4860987
+  Kappa Temporal Statistic (percent): 11.57251265
+  Kappa M Statistic (percent): -2.44200244
+Model measurements
+  model training instances: 9999
+
+Index
+  20000
+Votes
+  0: 0.00238989
+  1: -0.00239092
+Measurements
+  classified instances: 19999
+  classifications correct (percent): 57.20786039
+  Kappa Statistic (percent): 0.47593559
+  Kappa Temporal Statistic (percent): 10.65873264
+  Kappa M Statistic (percent): -2.58930712
+Model measurements
+  model training instances: 19999
+
+Index
+  30000
+Votes
+  0: -0.40148243
+  1: 0.40148208
+Measurements
+  classified instances: 29999
+  classifications correct (percent): 57.39857995
+  Kappa Statistic (percent): 0.49809449
+  Kappa Temporal Statistic (percent): 11.72204186
+  Kappa M Statistic (percent): -2.33824471
+Model measurements
+  model training instances: 29999
+
+Index
+  40000
+Votes
+  0: -0.37733972
+  1: -0.31610203
+Measurements
+  classified instances: 39999
+  classifications correct (percent): 57.30893272
+  Kappa Statistic (percent): 0.91250658
+  Kappa Temporal Statistic (percent): 11.75649837
+  Kappa M Statistic (percent): -2.1291866
+Model measurements
+  model training instances: 39999
+
+Index
+  50000
+Votes
+  0: -0.37733972
+  1: -0.31610203
+Measurements
+  classified instances: 49999
+  classifications correct (percent): 57.3751475
+  Kappa Statistic (percent): 1.11019024
+  Kappa Temporal Statistic (percent): 11.78077655
+  Kappa M Statistic (percent): -1.99081164
+Model measurements
+  model training instances: 49999
+
+Index
+  60000
+Votes
+  0: -0.37733972
+  1: -0.31610203
+Measurements
+  classified instances: 59999
+  classifications correct (percent): 57.22595377
+  Kappa Statistic (percent): 1.3553988
+  Kappa Temporal Statistic (percent): 11.54919869
+  Kappa M Statistic (percent): -1.83318784
+Model measurements
+  model training instances: 59999
+
+Index
+  70000
+Votes
+  0: -0.37733972
+  1: -0.31610203
+Measurements
+  classified instances: 69999
+  classifications correct (percent): 57.10795869
+  Kappa Statistic (percent): 1.46362385
+  Kappa Temporal Statistic (percent): 11.58750258
+  Kappa M Statistic (percent): -1.76247289
+Model measurements
+  model training instances: 69999
+
+Index
+  80000
+Votes
+  0: -0.37733972
+  1: -0.31610203
+Measurements
+  classified instances: 79999
+  classifications correct (percent): 57.18946487
+  Kappa Statistic (percent): 1.72083945
+  Kappa Temporal Statistic (percent): 11.84329069
+  Kappa M Statistic (percent): -1.57788587
+Model measurements
+  model training instances: 79999
+
+Index
+  90000
+Votes
+  0: -0.37733972
+  1: -0.31610203
+Measurements
+  classified instances: 89999
+  classifications correct (percent): 57.20063556
+  Kappa Statistic (percent): 1.74312949
+  Kappa Temporal Statistic (percent): 11.82556942
+  Kappa M Statistic (percent): -1.57964135
+Model measurements
+  model training instances: 89999
+
+Index
+  100000
+Votes
+  0: -0.37733972
+  1: -0.31610203
+Measurements
+  classified instances: 99999
+  classifications correct (percent): 57.17857179
+  Kappa Statistic (percent): 1.93990852
+  Kappa Temporal Statistic (percent): 11.78930455
+  Kappa M Statistic (percent): -1.51486416
+Model measurements
+  model training instances: 99999
+
+
+