scalafmt: format 9 files to pass CI lint

Author: derrickburns

src/main/scala/com/massivedatascience/clusterer/ml/AgglomerativeBregman.scala

@@ -478,9 +478,9 @@ class AgglomerativeBregmanModel(
     with Logging
     with HasTrainingSummary {
 
-  private[ml] var modelDivergence: String = "squaredEuclidean"
-  private[ml] var modelSmoothing: Double  = 1e-10
-  private[ml] var modelLinkage: String    = "average"
+  private[ml] var modelDivergence: String  = "squaredEuclidean"
+  private[ml] var modelSmoothing: Double   = 1e-10
+  private[ml] var modelLinkage: String     = "average"
   private[ml] var kernel: ClusteringKernel = _
 
   /** Cluster centers as vectors for downstream consumers/tests. */

src/main/scala/com/massivedatascience/clusterer/ml/BalancedKMeans.scala

@@ -513,7 +513,6 @@ class BalancedKMeans(override val uid: String)
     withDistances.withColumn("_assignment", assignUdf(col("_row_id")))
   }
 
-
   override def transformSchema(schema: StructType): StructType = {
     require(
       schema.fieldNames.contains($(featuresCol)),

src/main/scala/com/massivedatascience/clusterer/ml/DPMeans.scala

@@ -365,7 +365,6 @@ class DPMeans(override val uid: String)
     copyValues(model)
   }
 
-
   override def transformSchema(schema: StructType): StructType = {
     validateAndTransformSchema(schema)
   }
@@ -437,7 +436,6 @@ class DPMeansModel private[ml] (
     result
   }
 
-
   override def transformSchema(schema: StructType): StructType = {
     validateAndTransformSchema(schema)
   }

src/main/scala/com/massivedatascience/clusterer/ml/MiniBatchKMeans.scala

@@ -336,7 +336,6 @@ class MiniBatchKMeans(override val uid: String)
     )
   }
 
-
   /** Initialize cluster centers. */
   private def initializeCenters(
       df: DataFrame,

src/main/scala/com/massivedatascience/clusterer/ml/RobustKMeans.scala

@@ -414,7 +414,8 @@ class RobustKMeansModel(
 
   override def transform(dataset: Dataset[_]): DataFrame = {
     val df     = dataset.toDF()
-    val kernel = ClusteringOps.createKernel(divergenceName, $(smoothing)).asInstanceOf[BregmanKernel]
+    val kernel =
+      ClusteringOps.createKernel(divergenceName, $(smoothing)).asInstanceOf[BregmanKernel]
     val mode   = OutlierMode.fromString(outlierModeName)
 
     val bcKernel  = df.sparkSession.sparkContext.broadcast(kernel)

src/main/scala/com/massivedatascience/clusterer/ml/SoftKMeans.scala

@@ -280,7 +280,6 @@ class SoftKMeans(override val uid: String)
     }
   }
 
-
   /** Initialize cluster centers. */
   private def initializeCenters(
       df: DataFrame,

src/main/scala/com/massivedatascience/clusterer/ml/StreamingKMeans.scala

@@ -231,7 +231,8 @@ class StreamingKMeansModel(
   // Mutable state for streaming updates (array contents are mutated, not references)
   @transient private val centerArrays: Array[Vector]   = initialCenters.map(Vectors.dense)
   @transient private val clusterWeights: Array[Double] = Array.fill(initialCenters.length)(1.0)
-  @transient private lazy val kernel: ClusteringKernel  = createKernel(divergenceName, smoothingValue)
+  @transient private lazy val kernel: ClusteringKernel =
+    createKernel(divergenceName, smoothingValue)
 
   /** Get current cluster centers as Vectors (defensive copy).
     */

src/main/scala/com/massivedatascience/clusterer/ml/df/CenterInitializer.scala

@@ -76,10 +76,10 @@ private[ml] object CenterInitializer extends Logging {
       kernel: ClusteringKernel
   ): Array[Array[Double]] = {
     initMode.toLowerCase match {
-      case "random"                        => initializeRandom(df, featuresCol, k, seed)
+      case "random"                       => initializeRandom(df, featuresCol, k, seed)
       case "k-means||" | "kmeansparallel" =>
         initializeKMeansPlusPlus(df, featuresCol, k, seed, kernel)
-      case other                           =>
+      case other                          =>
         throw new IllegalArgumentException(
           s"Unknown init mode: '$other'. Valid options: random, k-means||"
         )

src/test/scala/com/massivedatascience/clusterer/ml/df/BregmanKernelAccuracySuite.scala

@@ -911,7 +911,7 @@ class BregmanKernelAccuracySuite extends AnyFunSuite with Matchers {
   test("All kernels handle vectors with zeros (with smoothing)") {
     val testCases = Seq(
       (new SquaredEuclideanKernel(), Vectors.dense(0.0, 1.0, 0.0)),
-      (new KLDivergenceKernel(1e-6), Vectors.dense(0.0, 1.0)),           // Needs smoothing
+      (new KLDivergenceKernel(1e-6), Vectors.dense(0.0, 1.0)),          // Needs smoothing
       (new GeneralizedIDivergenceKernel(1e-6), Vectors.dense(0.0, 1.0)) // Needs smoothing
     )
 
@@ -1030,8 +1030,17 @@ class BregmanKernelAccuracySuite extends AnyFunSuite with Matchers {
   }
 
   test("ClusteringOps: createUpdateStrategy returns GradMeanUDAFUpdate for Bregman kernels") {
-    for (div <- Seq("squaredEuclidean", "kl", "itakuraSaito", "generalizedI", "logistic",
-      "spherical", "cosine")) {
+    for (
+      div <- Seq(
+               "squaredEuclidean",
+               "kl",
+               "itakuraSaito",
+               "generalizedI",
+               "logistic",
+               "spherical",
+               "cosine"
+             )
+    ) {
       withClue(s"divergence=$div:") {
         val updater = ClusteringOps.createUpdateStrategy(div)
         updater shouldBe a[strategies.GradMeanUDAFUpdate]
@@ -1046,8 +1055,9 @@ class BregmanKernelAccuracySuite extends AnyFunSuite with Matchers {
   }
 
   test("ClusteringOps: createKernel returns BregmanKernel for standard divergences") {
-    for (div <- Seq("squaredEuclidean", "kl", "itakuraSaito", "generalizedI", "logistic",
-      "spherical")) {
+    for (
+      div <- Seq("squaredEuclidean", "kl", "itakuraSaito", "generalizedI", "logistic", "spherical")
+    ) {
       withClue(s"divergence=$div:") {
         val kernel = ClusteringOps.createKernel(div)
         kernel shouldBe a[kernels.BregmanKernel]