Add Tests

Author: muencseb

pom.xml

@@ -35,6 +35,25 @@
             <artifactId>spark-mllib_2.10</artifactId>
             <version>1.6.1</version>
         </dependency>
+        <dependency>
+            <groupId>junit</groupId>
+            <artifactId>junit</artifactId>
+            <version>4.12</version>
+            <scope>test</scope>
+        </dependency>
+        <dependency>
+            <groupId>junit</groupId>
+            <artifactId>junit</artifactId>
+            <version>4.12</version>
+            <scope>test</scope>
+        </dependency>
+        <dependency>
+            <groupId>org.hamcrest</groupId>
+            <artifactId>hamcrest-all</artifactId>
+            <version>1.3</version>
+            <scope>test</scope>
+        </dependency>
+
     </dependencies>
 
 

src/main/java/recommendation/model/RecommendationMlModel.java

@@ -30,13 +30,14 @@ public class RecommendationMlModel {
     private ALS als = new ALS();
     private MatrixFactorizationModel model;
     private ReentrantLock trainingLock = new ReentrantLock();
+    private JavaSparkContext javaSparkContext = new JavaSparkContext(SPARK_MASTER, SPARK_APP_NAME);
 
 
-    private RDDHelper rddHelper = new RDDHelper(new JavaSparkContext(SPARK_MASTER, SPARK_APP_NAME));
+    private RDDHelper rddHelper = new RDDHelper(javaSparkContext);
 
     private ReentrantReadWriteLock mutex = new ReentrantReadWriteLock();
 
-    private boolean modelIsReady = false;
+    private volatile boolean modelIsReady = false;
 
 
     public RecommendationMlModel(Callable<Collection<InputRating>> inputRatings) {
@@ -61,6 +62,10 @@ public Double getInterestPrediction(Integer userId, Integer eventId) throws Mode
         return prediction;
     }
 
+    public void close() {
+        javaSparkContext.close();
+    }
+
     private void asyncTrainModel(Callable<Collection<InputRating>> inputRatings) {
         try {
             asyncTrainModel(inputRatings.call());
@@ -89,6 +94,7 @@ private void trainModel(Collection<InputRating> ratings) {
         model = als.setRank(10).setIterations(10).run(ratingRDD);
         mutex.writeLock().unlock();
         modelIsReady = true;
+
     }
 
 

src/main/java/recommendation/service/RecommendationService.java

@@ -27,5 +27,9 @@ public Double getPrediction(Integer userId, Integer productId) throws ModelNotRe
         return recommendationMlModel.getInterestPrediction(userId, productId);
     }
 
+    public void close() {
+        recommendationMlModel.close();
+    }
+
 
 }

src/test/java/data/RDDHelperTest.java

@@ -0,0 +1,45 @@
+package data;
+
+import org.apache.commons.collections.CollectionUtils;
+import org.apache.spark.api.java.JavaRDD;
+import org.apache.spark.api.java.JavaSparkContext;
+import org.junit.AfterClass;
+import org.junit.Test;
+import recommendation.data.RDDHelper;
+
+import java.util.Arrays;
+import java.util.List;
+import java.util.stream.Collectors;
+import java.util.stream.IntStream;
+
+import static org.hamcrest.MatcherAssert.assertThat;
+import static org.hamcrest.core.IsInstanceOf.instanceOf;
+import static org.junit.Assert.assertTrue;
+
+public class RDDHelperTest {
+
+    private static JavaSparkContext jsc = new JavaSparkContext("local", "testjsc");
+    private RDDHelper rddHelper = new RDDHelper(jsc);
+
+    @AfterClass
+    public static void close() {
+        jsc.close();
+    }
+
+    @Test
+    public void shouldCreateRDDFromCollection() {
+        List<Integer> intList = IntStream.range(0, 100).boxed().collect(Collectors.toList());
+        JavaRDD intRDD = rddHelper.getRddFromCollection(intList);
+        assertTrue(CollectionUtils.isEqualCollection(intRDD.collect(), intList));
+        assertThat(intRDD, instanceOf(JavaRDD.class));
+    }
+
+    @Test
+    public void shouldCreateRDDFromCollectionWithConverter() {
+        List<Integer> intList = IntStream.range(0, 3).boxed().collect(Collectors.toList());
+        JavaRDD stringRDD = rddHelper.getRddFromCollection(intList, integer -> integer.toString());
+        assertThat(stringRDD, instanceOf(JavaRDD.class));
+        assertTrue(CollectionUtils.isEqualCollection(stringRDD.collect(), Arrays.asList("0", "1", "2")));
+        assertThat(stringRDD, instanceOf(JavaRDD.class));
+    }
+}

src/test/java/service/RecommendationServiceTest.java

@@ -0,0 +1,49 @@
+package service;
+
+import org.junit.BeforeClass;
+import org.junit.Test;
+import recommendation.data.InputRating;
+import recommendation.exceptions.ModelNotReadyException;
+import recommendation.service.RecommendationService;
+
+import java.util.ArrayList;
+
+import static org.hamcrest.MatcherAssert.assertThat;
+import static org.hamcrest.Matchers.greaterThan;
+import static org.hamcrest.Matchers.lessThan;
+import static org.junit.Assert.assertNotEquals;
+
+/**
+ * The testing was keepen simple, as sparkml is already tested. The testing focus is on the new features.
+ * This test just verifies, that sparkml was correctly called
+ */
+public class RecommendationServiceTest {
+
+
+    static RecommendationService recommendationService;
+
+    @BeforeClass
+    public static void initModel() {
+        ArrayList<InputRating> inputRatings = new ArrayList<>();
+        inputRatings.add(new InputRating(1, 1, 1));
+        inputRatings.add(new InputRating(1, 2, 0));
+        inputRatings.add(new InputRating(1, 3, 2));
+
+        inputRatings.add(new InputRating(2, 1, 1));
+        recommendationService = new RecommendationService(inputRatings);
+        while (!recommendationService.isModelReady()) {
+        }
+    }
+
+    @Test
+    public void shouldNotForgetKnownRating() throws ModelNotReadyException {
+        assertThat(recommendationService.getPrediction(1, 2), greaterThan(recommendationService.getPrediction(1, 1)));
+        assertNotEquals(0.0, recommendationService.getPrediction(1, 1));
+    }
+
+    @Test
+    public void shouldPredictFromOtherUser() throws ModelNotReadyException {
+        assertThat(recommendationService.getPrediction(2, 2), lessThan(recommendationService.getPrediction(1, 3)));
+    }
+
+}