fix for Jackson bindings in benchmarks (#281)

* fix for Jackson bindings in benchmarks * changes + shortening tests
aws · Oct 12, 2021 · c8b5039 · c8b5039
1 parent ed1bda6
commit c8b5039
Show file tree

Hide file tree

Showing 16 changed files with 37 additions and 60 deletions.
diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/config/ImputationMethod.java b/Java/core/src/main/java/com/amazon/randomcutforest/config/ImputationMethod.java
@@ -32,9 +32,13 @@ public enum ImputationMethod {
      * last known value in each input dimension
      */
     PREVIOUS,
+    /**
+     * linear interpolation
+     */
+    LINEAR,
     /**
      * use the RCF imputation; but would often require a minimum number of
-     * observations and would use PREVIOUS till that point
+     * observations and would uses a default (often LINEAR) till that point
      */
     RCF;
 }
diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/state/RandomCutForestMapper.java b/Java/core/src/main/java/com/amazon/randomcutforest/state/RandomCutForestMapper.java
@@ -271,7 +271,7 @@ public RandomCutForest toModel(RandomCutForestState state, ExecutionContext exec
                 .internalShinglingEnabled(state.isInternalShinglingEnabled()).randomSeed(seed);
 
         if (state.isCompact()) {
-            if (state.getPrecisionEnumValue() == Precision.FLOAT_32) {
+            if (Precision.valueOf(state.getPrecision()) == Precision.FLOAT_32) {
                 return singlePrecisionForest(builder, state, null, null, null);
             } else {
                 return doublePrecisionForest(builder, state, null, null, null);

diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/state/RandomCutForestState.java b/Java/core/src/main/java/com/amazon/randomcutforest/state/RandomCutForestState.java
@@ -21,7 +21,6 @@
 
 import lombok.Data;
 
-import com.amazon.randomcutforest.config.Precision;
 import com.amazon.randomcutforest.state.sampler.CompactSamplerState;
 import com.amazon.randomcutforest.state.store.PointStoreState;
 import com.amazon.randomcutforest.state.tree.CompactRandomCutTreeState;
@@ -81,7 +80,4 @@ public class RandomCutForestState {
 
     private boolean saveCoordinatorStateEnabled;
 
-    public Precision getPrecisionEnumValue() {
-        return Precision.valueOf(precision);
-    }
 }
diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/state/store/NodeStoreMapper.java b/Java/core/src/main/java/com/amazon/randomcutforest/state/store/NodeStoreMapper.java
@@ -53,7 +53,7 @@ public NodeStore toModel(NodeStoreState state, long seed) {
         int capacity = state.getCapacity();
         int[] cutDimension = ArrayPacking.unpackInts(state.getCutDimension(), state.isCompressed());
         double[] cutValue;
-        if (state.getPrecisionEnumValue() == Precision.FLOAT_32) {
+        if (Precision.valueOf(state.getPrecision()) == Precision.FLOAT_32) {
             cutValue = CommonUtils.toDoubleArray(ArrayPacking.unpackFloats(state.getCutValueData()));
         } else {
             cutValue = ArrayPacking.unpackDoubles(state.getCutValueData());
@@ -107,7 +107,7 @@ public NodeStoreState toState(NodeStore model) {
         }
 
         state.setCutDimension(ArrayPacking.pack(model.getCutDimension(), state.isCompressed()));
-        if (state.getPrecisionEnumValue() == Precision.FLOAT_32) {
+        if (Precision.valueOf(state.getPrecision()) == Precision.FLOAT_32) {
             state.setCutValueData(ArrayPacking.pack(CommonUtils.toFloatArray(model.getCutValue())));
         } else {
             state.setCutValueData(ArrayPacking.pack(model.getCutValue(), model.getCutValue().length));

diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/state/store/NodeStoreState.java b/Java/core/src/main/java/com/amazon/randomcutforest/state/store/NodeStoreState.java
@@ -19,8 +19,6 @@
 
 import lombok.Data;
 
-import com.amazon.randomcutforest.config.Precision;
-
 @Data
 public class NodeStoreState {
 
@@ -47,7 +45,4 @@ public class NodeStoreState {
     private int[] leafMass;
     private int[] leafPointIndex;
 
-    public Precision getPrecisionEnumValue() {
-        return Precision.valueOf(precision);
-    }
 }
diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/state/store/PointStoreDoubleMapper.java b/Java/core/src/main/java/com/amazon/randomcutforest/state/store/PointStoreDoubleMapper.java
@@ -41,7 +41,8 @@ public class PointStoreDoubleMapper implements IStateMapper<PointStoreDouble, Po
     public PointStoreDouble toModel(PointStoreState state, long seed) {
         checkNotNull(state.getRefCount(), "refCount must not be null");
         checkNotNull(state.getPointData(), "pointData must not be null");
-        checkArgument(state.getPrecisionEnumValue() == Precision.FLOAT_64, "precision must be " + Precision.FLOAT_64);
+        checkArgument(Precision.valueOf(state.getPrecision()) == Precision.FLOAT_64,
+                "precision must be " + Precision.FLOAT_64);
         int indexCapacity = state.getIndexCapacity();
         int dimensions = state.getDimensions();
         double[] store = ArrayPacking.unpackDoubles(state.getPointData(), state.getCurrentStoreCapacity() * dimensions);

diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/state/store/PointStoreFloatMapper.java b/Java/core/src/main/java/com/amazon/randomcutforest/state/store/PointStoreFloatMapper.java
@@ -42,7 +42,8 @@ public class PointStoreFloatMapper implements IStateMapper<PointStoreFloat, Poin
     public PointStoreFloat toModel(PointStoreState state, long seed) {
         checkNotNull(state.getRefCount(), "refCount must not be null");
         checkNotNull(state.getPointData(), "pointData must not be null");
-        checkArgument(state.getPrecisionEnumValue() == Precision.FLOAT_32, "precision must be " + Precision.FLOAT_32);
+        checkArgument(Precision.valueOf(state.getPrecision()) == Precision.FLOAT_32,
+                "precision must be " + Precision.FLOAT_32);
         int indexCapacity = state.getIndexCapacity();
         int dimensions = state.getDimensions();
         float[] store = ArrayPacking.unpackFloats(state.getPointData(), state.getCurrentStoreCapacity() * dimensions);

diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/state/store/PointStoreState.java b/Java/core/src/main/java/com/amazon/randomcutforest/state/store/PointStoreState.java
@@ -19,8 +19,6 @@
 
 import lombok.Data;
 
-import com.amazon.randomcutforest.config.Precision;
-
 /**
  * A class for storing the state of a
  * {@link com.amazon.randomcutforest.store.PointStoreDouble} or a
@@ -112,7 +110,4 @@ public class PointStoreState {
      */
     private int indexCapacity;
 
-    public Precision getPrecisionEnumValue() {
-        return Precision.valueOf(precision);
-    }
 }
diff --git a/Java/core/src/main/java/com/amazon/randomcutforest/state/store/SmallNodeStoreMapper.java b/Java/core/src/main/java/com/amazon/randomcutforest/state/store/SmallNodeStoreMapper.java
@@ -49,7 +49,8 @@ public class SmallNodeStoreMapper implements IStateMapper<SmallNodeStore, NodeSt
 
     @Override
     public SmallNodeStore toModel(NodeStoreState state, long seed) {
-        checkState(state.getPrecisionEnumValue() == Precision.FLOAT_32, " incorrect invocation of SmallNodeStore");
+        checkState(Precision.valueOf(state.getPrecision()) == Precision.FLOAT_32,
+                " incorrect invocation of SmallNodeStore");
         int capacity = state.getCapacity();
         short[] cutDimension = ArrayPacking.unpackShorts(state.getCutDimension(), state.isCompressed());
         float[] cutValue = ArrayPacking.unpackFloats(state.getCutValueData());

diff --git a/.../src/main/java/com/amazon/randomcutforest/state/tree/CompactRandomCutTreeFloatMapper.java b/.../src/main/java/com/amazon/randomcutforest/state/tree/CompactRandomCutTreeFloatMapper.java
@@ -43,8 +43,9 @@ public CompactRandomCutTreeFloat toModel(CompactRandomCutTreeState state, Compac
 
         INodeStore nodeStore;
 
-        if (AbstractCompactRandomCutTree.canUseSmallNodeStore(state.getNodeStoreState().getPrecisionEnumValue(),
-                state.getMaxSize(), state.getDimensions())) {
+        if (AbstractCompactRandomCutTree.canUseSmallNodeStore(
+                Precision.valueOf(state.getNodeStoreState().getPrecision()), state.getMaxSize(),
+                state.getDimensions())) {
             SmallNodeStoreMapper nodeStoreMapper = new SmallNodeStoreMapper();
             nodeStoreMapper.setPartialTreeStateEnabled(state.isPartialTreeState());
             nodeStore = nodeStoreMapper.toModel(state.getNodeStoreState());

diff --git a/.../core/src/test/java/com/amazon/randomcutforest/RandomCutForestShingledFunctionalTest.java b/.../core/src/test/java/com/amazon/randomcutforest/RandomCutForestShingledFunctionalTest.java
@@ -111,6 +111,7 @@ public void InternalShinglingTest(boolean rotation) {
         int shingleSize = 2;
         int dimensions = baseDimensions * shingleSize;
         long seed = new Random().nextLong();
+        System.out.println(seed);
 
         int numTrials = 1; // test is exact equality, reducing the number of trials
         int length = 4000 * sampleSize;
@@ -143,7 +144,7 @@ public void InternalShinglingTest(boolean rotation) {
             }
 
             for (int j = 0; j < shingledData.length; j++) {
-                // validate eaulity of points
+                // validate equality of points
                 for (int y = 0; y < baseDimensions; y++) {
                     int position = (rotation) ? (count % shingleSize) : shingleSize - 1;
                     assertEquals(dataWithKeys.data[count][y], shingledData[j][position * baseDimensions + y], 1e-10);

diff --git a/.../java/com/amazon/randomcutforest/parkservices/state/ThresholdedRandomCutForestMapper.java b/.../java/com/amazon/randomcutforest/parkservices/state/ThresholdedRandomCutForestMapper.java
@@ -19,6 +19,8 @@
 import lombok.Setter;
 
 import com.amazon.randomcutforest.RandomCutForest;
+import com.amazon.randomcutforest.config.ForestMode;
+import com.amazon.randomcutforest.config.TransformMethod;
 import com.amazon.randomcutforest.parkservices.ThresholdedRandomCutForest;
 import com.amazon.randomcutforest.parkservices.preprocessor.Preprocessor;
 import com.amazon.randomcutforest.parkservices.state.preprocessor.PreprocessorMapper;
@@ -58,9 +60,9 @@ public ThresholdedRandomCutForest toModel(ThresholdedRandomCutForestState state,
         tForest.setLastAnomalyAttribution(new DiVectorMapper().toModel(state.getLastAnomalyAttribution()));
         tForest.setLastAnomalyPoint(state.getLastAnomalyPoint());
         tForest.setLastExpectedPoint(state.getLastExpectedPoint());
-        tForest.setForestMode(state.getForestModeEnumValue());
+        tForest.setForestMode(ForestMode.valueOf(state.getForestMode()));
         tForest.setLastRelativeIndex(state.getLastRelativeIndex());
-        tForest.setTransformMethod(state.getTransformMethodEnumValue());
+        tForest.setTransformMethod(TransformMethod.valueOf(state.getTransformMethod()));
         return tForest;
     }
 

diff --git a/...n/java/com/amazon/randomcutforest/parkservices/state/ThresholdedRandomCutForestState.java b/...n/java/com/amazon/randomcutforest/parkservices/state/ThresholdedRandomCutForestState.java
@@ -19,8 +19,6 @@
 
 import lombok.Data;
 
-import com.amazon.randomcutforest.config.ForestMode;
-import com.amazon.randomcutforest.config.TransformMethod;
 import com.amazon.randomcutforest.parkservices.state.preprocessor.PreprocessorState;
 import com.amazon.randomcutforest.parkservices.state.threshold.BasicThresholderState;
 import com.amazon.randomcutforest.state.RandomCutForestState;
@@ -52,11 +50,4 @@ public class ThresholdedRandomCutForestState {
     private int lastRelativeIndex;
     private int lastReset;
 
-    public TransformMethod getTransformMethodEnumValue() {
-        return TransformMethod.valueOf(transformMethod);
-    }
-
-    public ForestMode getForestModeEnumValue() {
-        return ForestMode.valueOf(forestMode);
-    }
 }
diff --git a/...n/java/com/amazon/randomcutforest/parkservices/state/preprocessor/PreprocessorMapper.java b/...n/java/com/amazon/randomcutforest/parkservices/state/preprocessor/PreprocessorMapper.java
@@ -18,6 +18,9 @@
 import lombok.Getter;
 import lombok.Setter;
 
+import com.amazon.randomcutforest.config.ForestMode;
+import com.amazon.randomcutforest.config.ImputationMethod;
+import com.amazon.randomcutforest.config.TransformMethod;
 import com.amazon.randomcutforest.parkservices.preprocessor.Preprocessor;
 import com.amazon.randomcutforest.parkservices.state.statistics.DeviationMapper;
 import com.amazon.randomcutforest.parkservices.state.statistics.DeviationState;
@@ -41,12 +44,13 @@ public Preprocessor toModel(PreprocessorState state, long seed) {
             }
         }
         Preprocessor.Builder<?> preprocessorBuilder = new Preprocessor.Builder<>()
-                .forestMode(state.getForestModeEnumValue()).shingleSize(state.getShingleSize())
-                .dimensions(state.getDimensions()).imputationMethod(state.getImputationMethodEnumValue())
+                .forestMode(ForestMode.valueOf(state.getForestMode())).shingleSize(state.getShingleSize())
+                .dimensions(state.getDimensions())
+                .imputationMethod(ImputationMethod.valueOf(state.getImputationMethod()))
                 .fillValues(state.getDefaultFill()).inputLength(state.getInputLength()).weights(state.getWeights())
-                .transformMethod(state.getTransformMethodEnumValue()).startNormalization(state.getStartNormalization())
-                .useImputedFraction(state.getUseImputedFraction()).timeDeviation(timeStampDeviation)
-                .dataQuality(dataQuality);
+                .transformMethod(TransformMethod.valueOf(state.getTransformMethod()))
+                .startNormalization(state.getStartNormalization()).useImputedFraction(state.getUseImputedFraction())
+                .timeDeviation(timeStampDeviation).dataQuality(dataQuality);
 
         if (deviations != null) {
             preprocessorBuilder.deviations(deviations);

diff --git a/...in/java/com/amazon/randomcutforest/parkservices/state/preprocessor/PreprocessorState.java b/...in/java/com/amazon/randomcutforest/parkservices/state/preprocessor/PreprocessorState.java
@@ -19,9 +19,6 @@
 
 import lombok.Data;
 
-import com.amazon.randomcutforest.config.ForestMode;
-import com.amazon.randomcutforest.config.ImputationMethod;
-import com.amazon.randomcutforest.config.TransformMethod;
 import com.amazon.randomcutforest.parkservices.state.statistics.DeviationState;
 
 @Data
@@ -54,16 +51,4 @@ public class PreprocessorState {
     private DeviationState timeStampDeviationState;
     private DeviationState[] deviationStates;
 
-    public TransformMethod getTransformMethodEnumValue() {
-        return TransformMethod.valueOf(transformMethod);
-    }
-
-    public ForestMode getForestModeEnumValue() {
-        return ForestMode.valueOf(forestMode);
-    }
-
-    public ImputationMethod getImputationMethodEnumValue() {
-        return ImputationMethod.valueOf(imputationMethod);
-    }
-
 }
diff --git a/Java/parkservices/src/test/java/com/amazon/randomcutforest/parkservices/ConsistencyTest.java b/Java/parkservices/src/test/java/com/amazon/randomcutforest/parkservices/ConsistencyTest.java
@@ -68,8 +68,8 @@ public void InternalShinglingTest() {
     @Test
     public void ExternalShinglingTest() {
         int sampleSize = 256;
-        int baseDimensions = 2;
-        int shingleSize = 8;
+        int baseDimensions = 1;
+        int shingleSize = 4;
         int dimensions = baseDimensions * shingleSize;
         long seed = new Random().nextLong();
 
@@ -122,8 +122,8 @@ public void ExternalShinglingTest() {
     @Test
     public void MixedShinglingTest() {
         int sampleSize = 256;
-        int baseDimensions = 2;
-        int shingleSize = 8;
+        int baseDimensions = 1;
+        int shingleSize = 4;
         int dimensions = baseDimensions * shingleSize;
         long seed = new Random().nextLong();
 
@@ -158,7 +158,7 @@ public void MixedShinglingTest() {
             }
 
             for (int j = 0; j < length; j++) {
-                // validate eaulity of points
+                // validate equality of points
                 for (int y = 0; y < baseDimensions; y++) {
                     assertEquals(dataWithKeys.data[count][y], shingledData[j][(shingleSize - 1) * baseDimensions + y],
                             1e-10);