Checkstyle fixes

Author: davidkyle

x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/MachineLearning.java

@@ -195,15 +195,14 @@
 import org.elasticsearch.xpack.ml.dataframe.persistence.DataFrameAnalyticsConfigProvider;
 import org.elasticsearch.xpack.ml.dataframe.process.AnalyticsProcessFactory;
 import org.elasticsearch.xpack.ml.dataframe.process.AnalyticsProcessManager;
-import org.elasticsearch.xpack.ml.dataframe.process.results.AnalyticsResult;
 import org.elasticsearch.xpack.ml.dataframe.process.MemoryUsageEstimationProcessManager;
-import org.elasticsearch.xpack.ml.dataframe.process.results.MemoryUsageEstimationResult;
-import org.elasticsearch.xpack.ml.dataframe.process.NativeMemoryUsageEstimationProcessFactory;
 import org.elasticsearch.xpack.ml.dataframe.process.NativeAnalyticsProcessFactory;
+import org.elasticsearch.xpack.ml.dataframe.process.NativeMemoryUsageEstimationProcessFactory;
+import org.elasticsearch.xpack.ml.dataframe.process.results.AnalyticsResult;
+import org.elasticsearch.xpack.ml.dataframe.process.results.MemoryUsageEstimationResult;
 import org.elasticsearch.xpack.ml.inference.InferenceProcessor;
 import org.elasticsearch.xpack.ml.inference.ModelLoader;
 import org.elasticsearch.xpack.ml.inference.sillymodel.SillyModelLoader;
-import org.elasticsearch.xpack.ml.inference.tree.Tree;
 import org.elasticsearch.xpack.ml.inference.tree.TreeModelLoader;
 import org.elasticsearch.xpack.ml.job.JobManager;
 import org.elasticsearch.xpack.ml.job.JobManagerHolder;

x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/tree/Tree.java

@@ -1,3 +1,9 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
 package org.elasticsearch.xpack.ml.inference.tree;
 
 import java.util.ArrayList;
@@ -105,7 +111,7 @@ public static class Node {
             this(-1, -1, -1, false, value);
         }
 
-        boolean isLeaf() {
+        public boolean isLeaf() {
             return leftChild < 1;
         }
 
@@ -122,10 +128,22 @@ boolean isMissing(Double feature) {
             return feature == null;
         }
 
-        Double value() {
+        public Double value() {
             return thresholdValue;
         }
 
+        public int getFeatureIndex() {
+            return featureIndex;
+        }
+
+        public int getLeftChild() {
+            return leftChild;
+        }
+
+        public int getRightChild() {
+            return rightChild;
+        }
+
         @Override
         public String toString() {
             StringBuilder builder = new StringBuilder("{\n");

x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/tree/TreeEnsembleModel.java

@@ -1,3 +1,9 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
 package org.elasticsearch.xpack.ml.inference.tree;
 
 import org.apache.log4j.LogManager;

x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/tree/TreeModel.java

@@ -1,8 +1,16 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
 package org.elasticsearch.xpack.ml.inference.tree;
 
 import org.elasticsearch.ingest.IngestDocument;
 import org.elasticsearch.xpack.ml.inference.Model;
 
+import java.util.function.BiConsumer;
+
 public class TreeModel implements Model {
 
     private String targetFieldName;
@@ -15,9 +23,9 @@ public class TreeModel implements Model {
     }
 
     @Override
-    public IngestDocument infer(IngestDocument document) {
+    public void infer(IngestDocument document, BiConsumer<IngestDocument, Exception> handler) {
         Double prediction = ensemble.predictFromDoc(document.getSourceAndMetadata());
         document.setFieldValue(targetFieldName, prediction);
-        return document;
+        handler.accept(document, null);
     }
 }

x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/tree/TreeModelLoader.java

@@ -1,3 +1,9 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
 package org.elasticsearch.xpack.ml.inference.tree;
 
 import org.apache.log4j.LogManager;
@@ -46,7 +52,7 @@ public TreeModel load(String modelId, String processorTag, boolean ignoreMissing
         String index = ConfigurationUtils.readStringProperty(InferenceProcessor.TYPE, processorTag, config, INDEX);
         String targetField = ConfigurationUtils.readStringProperty(InferenceProcessor.TYPE, processorTag, config, TARGET_FIELD);
 
-        load(modelId, index, listener);
+        load(modelId, index, config, listener);
         latch.await();
         if (exception.get() != null) {
             throw exception.get();
@@ -59,11 +65,17 @@ public TreeModel load(String modelId, String processorTag, boolean ignoreMissing
         return model;
     }
 
-    public void load(String id, String index, ActionListener<TreeEnsembleModel> listener) {
+    @Override
+    public void consumeConfiguration(String processorTag, Map<String, Object> config) {
+        ConfigurationUtils.readStringProperty(InferenceProcessor.TYPE, processorTag, config, INDEX);
+        ConfigurationUtils.readStringProperty(InferenceProcessor.TYPE, processorTag, config, TARGET_FIELD);
+    }
+
+    public void load(String id, String index, Map<String, Object> config, ActionListener<TreeEnsembleModel> listener) {
         client.prepareGet(index, null, id).execute(ActionListener.wrap(
                 response -> {
                     if (response.isExists()) {
-                        listener.onResponse(createEnsemble(response.getSourceAsMap()));
+                        listener.onResponse(createEnsemble(response.getSourceAsMap(), featureMapFromConfig(config)));
                     } else {
                         listener.onFailure(new ResourceNotFoundException("missing model [{}], [{}]", id, index));
                     }
@@ -72,25 +84,23 @@ public void load(String id, String index, ActionListener<TreeEnsembleModel> list
         ));
     }
 
-    @SuppressWarnings("unchecked")
-    private TreeEnsembleModel createEnsemble(Map<String, Object> source) throws IOException {
+    private Map<String, Integer> featureMapFromConfig(Map<String, Object> config) {
+        // TODO, this was hard coded for a demo
         Map<String, Integer> featureMap = new HashMap<>();
         featureMap.put("f0", 0);
         featureMap.put("f1", 1);
         featureMap.put("f2", 2);
         featureMap.put("f3", 3);
+        return featureMap;
+    }
 
+    @SuppressWarnings("unchecked")
+    private TreeEnsembleModel createEnsemble(Map<String, Object> source, Map<String, Integer> featureMap) throws IOException {
         List<Map<String, Object>> trees = (List<Map<String, Object>>) source.get("ensemble");
         if (trees == null) {
             throw new IllegalStateException("missing trees");
         }
 
         return XgBoostJsonParser.parse(trees, featureMap);
     }
-
-    @Override
-    public void readConfiguration(String processorTag, Map<String, Object> config) {
-        ConfigurationUtils.readStringProperty(InferenceProcessor.TYPE, processorTag, config, INDEX);
-        ConfigurationUtils.readStringProperty(InferenceProcessor.TYPE, processorTag, config, TARGET_FIELD);
-    }
 }

x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/inference/tree/xgboost/XgBoostJsonParser.java

@@ -1,3 +1,10 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+
 package org.elasticsearch.xpack.ml.inference.tree.xgboost;
 
 import org.elasticsearch.xpack.ml.inference.tree.Tree;
@@ -14,8 +21,6 @@ public class XgBoostJsonParser {
     private static String SPLIT_CONDITION = "split_condition";
     private static String CHILDREN = "children";
 
-//    private static Logger logger = LogManager.getLogger(XgBoostJsonParser.class);
-
 
     public static TreeEnsembleModel parse(List<Map<String, Object>> json, Map<String, Integer> featureMap) throws IOException {
 
@@ -69,6 +74,12 @@ private static void parseTree(Map<String, Object> treeMap, Map<String, Integer>
                 throw new IOException("field [SPLIT_CONDITION] in node [" + nodeId + "] is not a double");
             }
 
+            // Fortunately XGBoost numbers the nodes the same way we do (breath first)
+            // Assuming a 1 based index a non-leaf node at index X will have its child nodes
+            // at indices 2X and 2X +1 (in practice we are using a zero based index so adjust
+            // accordingly). XGBoost uses the same numbering system so we don't have to point
+            // the child indices somewhere else on the new node, the child nodes will
+            // automatically slot in the in right place.
             treeBuilder.addJunction(nodeId, featureIndex, true, threshold);
 
             for (Map<String, Object> child : children) {

x-pack/plugin/ml/src/test/java/org/elasticsearch/xpack/ml/inference/InferenceProcessorTests.java

@@ -24,7 +24,6 @@ public class InferenceProcessorTests extends ESTestCase {
     public void testFactory() throws Exception {
         String processorTag = randomAlphaOfLength(10);
         String modelType = "test";
-        
         Model mockModel = mock(Model.class);
         ModelLoader mockLoader = mock(ModelLoader.class);
         when(mockLoader.load(eq("k2"), eq(processorTag), eq(false), any())).thenReturn(mockModel);