Adds clustering & nearest neighbour protobuf serialization

Classification/Core/src/main/java/org/tribuo/classification/ensemble/FullyWeightedVotingCombiner.java

@@ -165,6 +165,11 @@ public ConfiguredObjectProvenance getProvenance() {
         return new ConfiguredObjectProvenanceImpl(this, "EnsembleCombiner");
     }
 
+    @Override
+    public Class<Label> getTypeWitness() {
+        return Label.class;
+    }
+
     /**
      * Exports this voting combiner to ONNX.
      * <p>

Classification/Core/src/main/java/org/tribuo/classification/ensemble/VotingCombiner.java

@@ -156,6 +156,11 @@ public ConfiguredObjectProvenance getProvenance() {
         return new ConfiguredObjectProvenanceImpl(this,"EnsembleCombiner");
     }
 
+    @Override
+    public Class<Label> getTypeWitness() {
+        return Label.class;
+    }
+
     /**
      * Exports this voting combiner to ONNX.
      * <p>

Clustering/Hdbscan/src/main/java/org/tribuo/clustering/hdbscan/HdbscanModel.java

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2021, 2022, Oracle and/or its affiliates. All rights reserved.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -16,6 +16,8 @@
 
 package org.tribuo.clustering.hdbscan;
 
+import com.google.protobuf.Any;
+import com.google.protobuf.InvalidProtocolBufferException;
 import com.oracle.labs.mlrg.olcut.util.Pair;
 import org.tribuo.Example;
 import org.tribuo.Excuse;
@@ -26,11 +28,16 @@
 import org.tribuo.Prediction;
 import org.tribuo.clustering.ClusterID;
 import org.tribuo.clustering.hdbscan.HdbscanTrainer.Distance;
+import org.tribuo.clustering.hdbscan.protos.ClusterExemplarProto;
+import org.tribuo.clustering.hdbscan.protos.HdbscanModelProto;
+import org.tribuo.impl.ModelDataCarrier;
 import org.tribuo.math.distance.DistanceType;
 import org.tribuo.math.la.DenseVector;
 import org.tribuo.math.la.SGDVector;
 import org.tribuo.math.la.SparseVector;
+import org.tribuo.math.la.Tensor;
 import org.tribuo.math.la.VectorTuple;
+import org.tribuo.protos.core.ModelProto;
 import org.tribuo.provenance.ModelProvenance;
 
 import java.io.IOException;
@@ -57,6 +64,11 @@
 public final class HdbscanModel extends Model<ClusterID> {
     private static final long serialVersionUID = 1L;
 
+    /**
+     * Protobuf serialization version.
+     */
+    public static final int CURRENT_VERSION = 0;
+
     private final List<Integer> clusterLabels;
 
     private final DenseVector outlierScoresVector;
@@ -76,13 +88,59 @@ public final class HdbscanModel extends Model<ClusterID> {
                  ImmutableOutputInfo<ClusterID> outputIDInfo, List<Integer> clusterLabels, DenseVector outlierScoresVector,
                  List<HdbscanTrainer.ClusterExemplar> clusterExemplars, DistanceType distType, double noisePointsOutlierScore) {
         super(name,description,featureIDMap,outputIDInfo,false);
-        this.clusterLabels = clusterLabels;
+        this.clusterLabels = Collections.unmodifiableList(clusterLabels);
         this.outlierScoresVector = outlierScoresVector;
-        this.clusterExemplars = clusterExemplars;
+        this.clusterExemplars = Collections.unmodifiableList(clusterExemplars);
         this.distType = distType;
         this.noisePointsOutlierScore = noisePointsOutlierScore;
     }
 
+    /**
+     * Deserialization factory.
+     * @param version The serialized object version.
+     * @param className The class name.
+     * @param message The serialized data.
+     */
+    public static HdbscanModel deserializeFromProto(int version, String className, Any message) throws InvalidProtocolBufferException {
+        if (version < 0 || version > CURRENT_VERSION) {
+            throw new IllegalArgumentException("Unknown version " + version + ", this class supports at most version " + CURRENT_VERSION);
+        }
+        HdbscanModelProto proto = message.unpack(HdbscanModelProto.class);
+
+        ModelDataCarrier<?> carrier = ModelDataCarrier.deserialize(proto.getMetadata());
+        if (!carrier.outputDomain().getOutput(0).getClass().equals(ClusterID.class)) {
+            throw new IllegalStateException("Invalid protobuf, output domain is not a clustering domain, found " + carrier.outputDomain().getClass());
+        }
+        @SuppressWarnings("unchecked") // guarded by getClass
+        ImmutableOutputInfo<ClusterID> outputDomain = (ImmutableOutputInfo<ClusterID>) carrier.outputDomain();
+
+        Tensor outlierScoresTensor = Tensor.deserialize(proto.getOutlierScoresVector());
+        if (!(outlierScoresTensor instanceof DenseVector)) {
+            throw new IllegalStateException("Invalid protobuf, outlier scores must be a dense vector, found " + outlierScoresTensor.getClass());
+        }
+        DenseVector outlierScoresVector = (DenseVector) outlierScoresTensor;
+
+        List<Integer> clusterLabels = new ArrayList<>(proto.getClusterLabelsList());
+        for (Integer i : clusterLabels) {
+            if (outputDomain.getOutput(i) == null && i != -1) {
+                throw new IllegalStateException("Invalid protobuf, found cluster id " + i + " which is not present in the domain " + outputDomain);
+            }
+        }
+        if (clusterLabels.size() != outlierScoresVector.size()) {
+            throw new IllegalStateException("Invalid protobuf, expected the same number of outlier scores as cluster labels, found " +outlierScoresVector.size() + " scores and " + clusterLabels.size() + " labels");
+        }
+
+        List<HdbscanTrainer.ClusterExemplar> exemplars = new ArrayList<>();
+        for (ClusterExemplarProto p : proto.getClusterExemplarsList()) {
+            exemplars.add(HdbscanTrainer.ClusterExemplar.deserialize(p));
+        }
+
+        DistanceType distType = DistanceType.valueOf(proto.getDistType());
+
+        return new HdbscanModel(carrier.name(), carrier.provenance(), carrier.featureDomain(),
+            outputDomain, clusterLabels, outlierScoresVector, exemplars, distType, proto.getNoisePointsOutlierScore());
+    }
+
     /**
      * Returns the cluster labels for the training data.
      * <p>
@@ -202,10 +260,32 @@ public Optional<Excuse<ClusterID>> getExcuse(Example<ClusterID> example) {
         return Optional.empty();
     }
 
+    @Override
+    public ModelProto serialize() {
+        ModelDataCarrier<ClusterID> carrier = createDataCarrier();
+
+        HdbscanModelProto.Builder modelBuilder = HdbscanModelProto.newBuilder();
+        modelBuilder.setMetadata(carrier.serialize());
+        modelBuilder.addAllClusterLabels(clusterLabels);
+        modelBuilder.setOutlierScoresVector(outlierScoresVector.serialize());
+        modelBuilder.setDistType(distType.name());
+        for (HdbscanTrainer.ClusterExemplar e : clusterExemplars) {
+            modelBuilder.addClusterExemplars(e.serialize());
+        }
+        modelBuilder.setNoisePointsOutlierScore(noisePointsOutlierScore);
+
+        ModelProto.Builder builder = ModelProto.newBuilder();
+        builder.setSerializedData(Any.pack(modelBuilder.build()));
+        builder.setClassName(HdbscanModel.class.getName());
+        builder.setVersion(CURRENT_VERSION);
+
+        return builder.build();
+    }
+
     @Override
     protected HdbscanModel copy(String newName, ModelProvenance newProvenance) {
         DenseVector copyOutlierScoresVector = outlierScoresVector.copy();
-        List<Integer> copyClusterLabels = Collections.unmodifiableList(clusterLabels);
+        List<Integer> copyClusterLabels = new ArrayList<>(clusterLabels);
         List<HdbscanTrainer.ClusterExemplar> copyExemplars = new ArrayList<>(clusterExemplars);
         return new HdbscanModel(newName, newProvenance, featureIDMap, outputIDInfo, copyClusterLabels,
             copyOutlierScoresVector, copyExemplars, distType, noisePointsOutlierScore);

Clustering/Hdbscan/src/main/java/org/tribuo/clustering/hdbscan/HdbscanTrainer.java

@@ -27,10 +27,12 @@
 import org.tribuo.Trainer;
 import org.tribuo.clustering.ClusterID;
 import org.tribuo.clustering.ImmutableClusteringInfo;
+import org.tribuo.clustering.hdbscan.protos.ClusterExemplarProto;
 import org.tribuo.math.distance.DistanceType;
 import org.tribuo.math.la.DenseVector;
 import org.tribuo.math.la.SGDVector;
 import org.tribuo.math.la.SparseVector;
+import org.tribuo.math.la.Tensor;
 import org.tribuo.math.neighbour.NeighboursQuery;
 import org.tribuo.math.neighbour.NeighboursQueryFactory;
 import org.tribuo.math.neighbour.NeighboursQueryFactoryType;
@@ -924,6 +926,26 @@ public boolean equals(Object o) {
         public int hashCode() {
             return Objects.hash(label, outlierScore, features, maxDistToEdge);
         }
+
+        ClusterExemplarProto serialize() {
+            ClusterExemplarProto.Builder builder = ClusterExemplarProto.newBuilder();
+
+            builder.setLabel(label);
+            builder.setOutlierScore(outlierScore);
+            builder.setFeatures(features.serialize());
+            builder.setMaxDistToEdge(maxDistToEdge);
+
+            return builder.build();
+        }
+
+        static ClusterExemplar deserialize(ClusterExemplarProto proto) {
+            Tensor tensor = Tensor.deserialize(proto.getFeatures());
+            if (!(tensor instanceof SGDVector)) {
+                throw new IllegalStateException("Invalid protobuf, features must be an SGDVector, found " + tensor.getClass());
+            }
+            SGDVector vector = (SGDVector) tensor;
+            return new ClusterExemplar(proto.getLabel(),proto.getOutlierScore(),vector,proto.getMaxDistToEdge());
+        }
     }
 
 }