Default "prediction_field_name" to (dependent_variable + "_prediction") (#48232) (#48282)

przemekwitek · web-flow · commit e303683b323a · 2019-10-21T13:18:14.000+02:00
diff --git a/client/rest-high-level/src/test/java/org/elasticsearch/client/MachineLearningIT.java b/client/rest-high-level/src/test/java/org/elasticsearch/client/MachineLearningIT.java
@@ -126,8 +126,8 @@
 import org.elasticsearch.client.ml.dataframe.PhaseProgress;
 import org.elasticsearch.client.ml.dataframe.QueryConfig;
 import org.elasticsearch.client.ml.dataframe.evaluation.classification.Classification;
-import org.elasticsearch.client.ml.dataframe.evaluation.regression.MeanSquaredErrorMetric;
 import org.elasticsearch.client.ml.dataframe.evaluation.classification.MulticlassConfusionMatrixMetric;
+import org.elasticsearch.client.ml.dataframe.evaluation.regression.MeanSquaredErrorMetric;
 import org.elasticsearch.client.ml.dataframe.evaluation.regression.RSquaredMetric;
 import org.elasticsearch.client.ml.dataframe.evaluation.regression.Regression;
 import org.elasticsearch.client.ml.dataframe.evaluation.softclassification.AucRocMetric;
@@ -1297,6 +1297,7 @@ public void testPutDataFrameAnalyticsConfig_GivenRegression() throws Exception {
                 .setIndex("put-test-dest-index")
                 .build())
             .setAnalysis(org.elasticsearch.client.ml.dataframe.Regression.builder("my_dependent_variable")
+                .setPredictionFieldName("my_dependent_variable_prediction")
                 .setTrainingPercent(80.0)
                 .build())
             .setDescription("this is a regression")
@@ -1331,6 +1332,7 @@ public void testPutDataFrameAnalyticsConfig_GivenClassification() throws Excepti
                 .setIndex("put-test-dest-index")
                 .build())
             .setAnalysis(org.elasticsearch.client.ml.dataframe.Classification.builder("my_dependent_variable")
+                .setPredictionFieldName("my_dependent_variable_prediction")
                 .setTrainingPercent(80.0)
                 .setNumTopClasses(1)
                 .build())
diff --git a/x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/ml/dataframe/analyses/Classification.java b/x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/ml/dataframe/analyses/Classification.java
@@ -92,7 +92,7 @@ public Classification(String dependentVariable,
         }
         this.dependentVariable = ExceptionsHelper.requireNonNull(dependentVariable, DEPENDENT_VARIABLE);
         this.boostedTreeParams = ExceptionsHelper.requireNonNull(boostedTreeParams, BoostedTreeParams.NAME);
-        this.predictionFieldName = predictionFieldName;
+        this.predictionFieldName = predictionFieldName == null ? dependentVariable + "_prediction" : predictionFieldName;
         this.numTopClasses = numTopClasses == null ? DEFAULT_NUM_TOP_CLASSES : numTopClasses;
         this.trainingPercent = trainingPercent == null ? 100.0 : trainingPercent;
     }
@@ -113,6 +113,10 @@ public String getDependentVariable() {
         return dependentVariable;
     }
 
+    public String getPredictionFieldName() {
+        return predictionFieldName;
+    }
+
     public int getNumTopClasses() {
         return numTopClasses;
     }
diff --git a/x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/ml/dataframe/analyses/Regression.java b/x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/ml/dataframe/analyses/Regression.java
@@ -70,7 +70,7 @@ public Regression(String dependentVariable,
         }
         this.dependentVariable = ExceptionsHelper.requireNonNull(dependentVariable, DEPENDENT_VARIABLE);
         this.boostedTreeParams = ExceptionsHelper.requireNonNull(boostedTreeParams, BoostedTreeParams.NAME);
-        this.predictionFieldName = predictionFieldName;
+        this.predictionFieldName = predictionFieldName == null ? dependentVariable + "_prediction" : predictionFieldName;
         this.trainingPercent = trainingPercent == null ? 100.0 : trainingPercent;
     }
 
@@ -89,6 +89,10 @@ public String getDependentVariable() {
         return dependentVariable;
     }
 
+    public String getPredictionFieldName() {
+        return predictionFieldName;
+    }
+
     public double getTrainingPercent() {
         return trainingPercent;
     }
diff --git a/x-pack/plugin/core/src/test/java/org/elasticsearch/xpack/core/ml/dataframe/analyses/ClassificationTests.java b/x-pack/plugin/core/src/test/java/org/elasticsearch/xpack/core/ml/dataframe/analyses/ClassificationTests.java
@@ -73,6 +73,14 @@ public void testConstructor_GivenNumTopClassesIsGreaterThan1000() {
         assertThat(e.getMessage(), equalTo("[num_top_classes] must be an integer in [0, 1000]"));
     }
 
+    public void testGetPredictionFieldName() {
+        Classification classification = new Classification("foo", BOOSTED_TREE_PARAMS, "result", 3, 50.0);
+        assertThat(classification.getPredictionFieldName(), equalTo("result"));
+
+        classification = new Classification("foo", BOOSTED_TREE_PARAMS, null, 3, 50.0);
+        assertThat(classification.getPredictionFieldName(), equalTo("foo_prediction"));
+    }
+
     public void testGetNumTopClasses() {
         Classification classification = new Classification("foo", BOOSTED_TREE_PARAMS, "result", 7, 1.0);
         assertThat(classification.getNumTopClasses(), equalTo(7));
diff --git a/x-pack/plugin/core/src/test/java/org/elasticsearch/xpack/core/ml/dataframe/analyses/RegressionTests.java b/x-pack/plugin/core/src/test/java/org/elasticsearch/xpack/core/ml/dataframe/analyses/RegressionTests.java
@@ -19,6 +19,8 @@
 
 public class RegressionTests extends AbstractSerializingTestCase<Regression> {
 
+    private static final BoostedTreeParams BOOSTED_TREE_PARAMS = new BoostedTreeParams(0.0, 0.0, 0.5, 500, 1.0);
+
     @Override
     protected Regression doParseInstance(XContentParser parser) throws IOException {
         return Regression.fromXContent(parser, false);
@@ -42,32 +44,45 @@ protected Writeable.Reader<Regression> instanceReader() {
         return Regression::new;
     }
 
-    public void testConstructor_GivenTrainingPercentIsNull() {
-        Regression regression = new Regression("foo", new BoostedTreeParams(0.0, 0.0, 0.5, 500, 1.0), "result", null);
-        assertThat(regression.getTrainingPercent(), equalTo(100.0));
-    }
-
-    public void testConstructor_GivenTrainingPercentIsBoundary() {
-        Regression regression = new Regression("foo", new BoostedTreeParams(0.0, 0.0, 0.5, 500, 1.0), "result", 1.0);
-        assertThat(regression.getTrainingPercent(), equalTo(1.0));
-        regression = new Regression("foo", new BoostedTreeParams(0.0, 0.0, 0.5, 500, 1.0), "result", 100.0);
-        assertThat(regression.getTrainingPercent(), equalTo(100.0));
-    }
-
     public void testConstructor_GivenTrainingPercentIsLessThanOne() {
         ElasticsearchStatusException e = expectThrows(ElasticsearchStatusException.class,
-            () -> new Regression("foo", new BoostedTreeParams(0.0, 0.0, 0.5, 500, 1.0), "result", 0.999));
+            () -> new Regression("foo", BOOSTED_TREE_PARAMS, "result", 0.999));
 
         assertThat(e.getMessage(), equalTo("[training_percent] must be a double in [1, 100]"));
     }
 
     public void testConstructor_GivenTrainingPercentIsGreaterThan100() {
         ElasticsearchStatusException e = expectThrows(ElasticsearchStatusException.class,
-            () -> new Regression("foo", new BoostedTreeParams(0.0, 0.0, 0.5, 500, 1.0), "result", 100.0001));
+            () -> new Regression("foo", BOOSTED_TREE_PARAMS, "result", 100.0001));
 
         assertThat(e.getMessage(), equalTo("[training_percent] must be a double in [1, 100]"));
     }
 
+    public void testGetPredictionFieldName() {
+        Regression regression = new Regression("foo", BOOSTED_TREE_PARAMS, "result", 50.0);
+        assertThat(regression.getPredictionFieldName(), equalTo("result"));
+
+        regression = new Regression("foo", BOOSTED_TREE_PARAMS, null, 50.0);
+        assertThat(regression.getPredictionFieldName(), equalTo("foo_prediction"));
+    }
+
+    public void testGetTrainingPercent() {
+        Regression regression = new Regression("foo", BOOSTED_TREE_PARAMS, "result", 50.0);
+        assertThat(regression.getTrainingPercent(), equalTo(50.0));
+
+        // Boundary condition: training_percent == 1.0
+        regression = new Regression("foo", BOOSTED_TREE_PARAMS, "result", 1.0);
+        assertThat(regression.getTrainingPercent(), equalTo(1.0));
+
+        // Boundary condition: training_percent == 100.0
+        regression = new Regression("foo", BOOSTED_TREE_PARAMS, "result", 100.0);
+        assertThat(regression.getTrainingPercent(), equalTo(100.0));
+
+        // training_percent == null, default applied
+        regression = new Regression("foo", BOOSTED_TREE_PARAMS, "result", null);
+        assertThat(regression.getTrainingPercent(), equalTo(100.0));
+    }
+
     public void testFieldCardinalityLimitsIsNonNull() {
         assertThat(createTestInstance().getFieldCardinalityLimits(), is(not(nullValue())));
     }
diff --git a/x-pack/plugin/src/test/resources/rest-api-spec/test/ml/data_frame_analytics_crud.yml b/x-pack/plugin/src/test/resources/rest-api-spec/test/ml/data_frame_analytics_crud.yml
@@ -1470,6 +1470,7 @@ setup:
       "eta": 0.5,
       "maximum_number_trees": 400,
       "feature_bag_fraction": 0.3,
+      "prediction_field_name": "foo_prediction",
       "training_percent": 60.3
     }
   }}
@@ -1809,6 +1810,7 @@ setup:
       "eta": 0.5,
       "maximum_number_trees": 400,
       "feature_bag_fraction": 0.3,
+      "prediction_field_name": "foo_prediction",
       "training_percent": 60.3,
       "num_top_classes": 2
     }
@@ -1844,6 +1846,7 @@ setup:
   - match: { analysis: {
     "regression":{
       "dependent_variable": "foo",
+      "prediction_field_name": "foo_prediction",
       "training_percent": 100.0
     }
   }}

Original file line number	Diff line number	Diff line change
`@@ -92,7 +92,7 @@ public Classification(String dependentVariable,`
`92`	`92`	`}`
`93`	`93`	`this.dependentVariable = ExceptionsHelper.requireNonNull(dependentVariable, DEPENDENT_VARIABLE);`
`94`	`94`	`this.boostedTreeParams = ExceptionsHelper.requireNonNull(boostedTreeParams, BoostedTreeParams.NAME);`
`95`		`- this.predictionFieldName = predictionFieldName;`
	`95`	`+ this.predictionFieldName = predictionFieldName == null ? dependentVariable + "_prediction" : predictionFieldName;`
`96`	`96`	`this.numTopClasses = numTopClasses == null ? DEFAULT_NUM_TOP_CLASSES : numTopClasses;`
`97`	`97`	`this.trainingPercent = trainingPercent == null ? 100.0 : trainingPercent;`
`98`	`98`	`}`
`@@ -113,6 +113,10 @@ public String getDependentVariable() {`
`113`	`113`	`return dependentVariable;`
`114`	`114`	`}`
`115`	`115`
	`116`	`+ public String getPredictionFieldName() {`
	`117`	`+ return predictionFieldName;`
	`118`	`+ }`
	`119`	`+`
`116`	`120`	`public int getNumTopClasses() {`
`117`	`121`	`return numTopClasses;`
`118`	`122`	`}`
Original file line number	Diff line number	Diff line change
`@@ -70,7 +70,7 @@ public Regression(String dependentVariable,`
`70`	`70`	`}`
`71`	`71`	`this.dependentVariable = ExceptionsHelper.requireNonNull(dependentVariable, DEPENDENT_VARIABLE);`
`72`	`72`	`this.boostedTreeParams = ExceptionsHelper.requireNonNull(boostedTreeParams, BoostedTreeParams.NAME);`
`73`		`- this.predictionFieldName = predictionFieldName;`
	`73`	`+ this.predictionFieldName = predictionFieldName == null ? dependentVariable + "_prediction" : predictionFieldName;`
`74`	`74`	`this.trainingPercent = trainingPercent == null ? 100.0 : trainingPercent;`
`75`	`75`	`}`
`76`	`76`
`@@ -89,6 +89,10 @@ public String getDependentVariable() {`
`89`	`89`	`return dependentVariable;`
`90`	`90`	`}`
`91`	`91`
	`92`	`+ public String getPredictionFieldName() {`
	`93`	`+ return predictionFieldName;`
	`94`	`+ }`
	`95`	`+`
`92`	`96`	`public double getTrainingPercent() {`
`93`	`97`	`return trainingPercent;`
`94`	`98`	`}`