migrate estimation cli kmeans to application layer

Author: lassewesth

algo/build.gradle

@@ -79,6 +79,7 @@ dependencies {
     testImplementation openGds.mockito.junit.jupiter
 
     testImplementation project(':centrality-algorithms')
+    testImplementation project(':community-algorithms')
     testImplementation project(':node-embedding-algorithms')
     testImplementation project(':path-finding-algorithms')
 }

algo/src/main/java/org/neo4j/gds/kmeans/Kmeans.java

@@ -39,9 +39,6 @@
 import java.util.concurrent.ExecutorService;
 
 public final class Kmeans extends Algorithm<KmeansResult> {
-
-    static final String KMEANS_DESCRIPTION =
-        "The Kmeans  algorithm clusters nodes into different communities based on Euclidean distance";
     private static final int UNASSIGNED = -1;
     private HugeIntArray bestCommunities;
     private final Graph graph;
@@ -166,7 +163,7 @@ private void kMeans(
         int restartIteration
     ) {
 
-        //note: currentDistanceFromCentroid is not reset to a [0,...,0] distance array but it does not have to
+        //note: currentDistanceFromCentroid is not reset to a [0,...,0] distance array, but it does not have to
         // it's used only in K-Means++ (where it is essentially reset; see func distanceFromLastSampledCentroid in KmeansTask)
         // or during final distance calculation where it is reset as well (see calculateFinalDistance in KmeansTask)
 
@@ -242,7 +239,7 @@ private void kMeans(
         }
         progressTracker.endSubTask(); // Main - end
 
-        double averageDistanceFromCentroid = calculatedistancePhase(tasks);
+        double averageDistanceFromCentroid = calculateDistancePhase(tasks);
         updateBestSolution(
             restartIteration,
             clusterManager,
@@ -262,7 +259,7 @@ private void initializeCentroids(ClusterManager clusterManager, KmeansSampler sa
         progressTracker.endSubTask(); // Initialization - end
     }
 
-    private void recomputeCentroids(ClusterManager clusterManager, List<KmeansTask> tasks) {
+    private void recomputeCentroids(ClusterManager clusterManager, Iterable<KmeansTask> tasks) {
         clusterManager.reset();
 
         for (KmeansTask task : tasks) {
@@ -367,7 +364,7 @@ private void calculateSilhouette() {
 
     }
 
-    private double calculatedistancePhase(List<KmeansTask> tasks) {
+    private double calculateDistancePhase(Iterable<KmeansTask> tasks) {
         for (KmeansTask task : tasks) {
             task.switchToPhase(TaskPhase.DISTANCE);
         }

algo/src/main/java/org/neo4j/gds/kmeans/KmeansAlgorithmFactory.java

@@ -24,7 +24,7 @@
 import org.junit.jupiter.params.ParameterizedTest;
 import org.junit.jupiter.params.provider.ValueSource;
 import org.neo4j.gds.api.Graph;
-import org.neo4j.gds.core.utils.progress.tasks.ProgressTracker;
+import org.neo4j.gds.applications.algorithms.community.CommunityAlgorithms;
 import org.neo4j.gds.extension.GdlExtension;
 import org.neo4j.gds.extension.GdlGraph;
 import org.neo4j.gds.extension.IdFunction;
@@ -265,6 +265,8 @@ void shouldComputeSilhouetteCorrectly() {
 
     @Test
     void shouldNotWorkForRestartsAndSeeds() {
+        var communityAlgorithms = new CommunityAlgorithms(null, null);
+
         var kmeansConfig = KmeansStreamConfigImpl.builder()
             .nodeProperty("kmeans")
             .concurrency(1)
@@ -273,31 +275,23 @@ void shouldNotWorkForRestartsAndSeeds() {
             .k(2)
             .numberOfRestarts(10)
             .build();
-
-        var kmeansAlgorithmFactory = new KmeansAlgorithmFactory<>();
-        assertThatThrownBy(() -> kmeansAlgorithmFactory.build(
-            lineGraph,
-            kmeansConfig,
-            ProgressTracker.NULL_TRACKER
-        )).hasMessageContaining("cannot be run");
+        assertThatThrownBy(() -> communityAlgorithms.kMeans(lineGraph, kmeansConfig))
+            .hasMessageContaining("cannot be run");
     }
 
     @Test
     void shouldNotWorkForDifferentSeedAndK() {
+        var communityAlgorithms = new CommunityAlgorithms(null, null);
+
         var kmeansConfig = KmeansStreamConfigImpl.builder()
             .nodeProperty("kmeans")
             .concurrency(1)
             .randomSeed(19L)
             .seedCentroids(List.of(List.of(1d)))
             .k(2)
             .build();
-
-        var kmeansAlgorithmFactory = new KmeansAlgorithmFactory<>();
-        assertThatThrownBy(() -> kmeansAlgorithmFactory.build(
-            lineGraph,
-            kmeansConfig,
-            ProgressTracker.NULL_TRACKER
-        )).hasMessageContaining("Incorrect");
+        assertThatThrownBy(() -> communityAlgorithms.kMeans(lineGraph, kmeansConfig))
+            .hasMessageContaining("Incorrect");
     }
 
     @Test