Corrected :*: member of DecomposableCovariance

dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/DynaMLPipe.scala

@@ -581,11 +581,7 @@ object DynaMLPipe {
     * Creates an [[Encoder]] which replicates a
     * [[DenseVector]] instance n times.
     * */
-  val breezeDVReplicationEncoder = (n: Int) => Encoder((v: DenseVector[Double]) => {
-    Array.fill(n)(v)
-  }, (vs: Array[DenseVector[Double]]) => {
-   vs.head
-  })
+  val breezeDVReplicationEncoder = (n: Int) => genericReplicationEncoder[DenseVector[Double]](n)
 
   def trainParametricModel[
   G, T, Q, R, S, M <: ParameterizedLearner[G, T, Q, R, S]

dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/kernels/LocalScalarKernel.scala

@@ -5,6 +5,8 @@ import io.github.mandar2812.dynaml.DynaMLPipe
 import io.github.mandar2812.dynaml.algebra.{PartitionedPSDMatrix, PartitionedVector}
 import io.github.mandar2812.dynaml.pipes.{DataPipe, Encoder}
 
+import scala.reflect.ClassTag
+
 /**
   * Scalar Kernel defines algebraic behavior for kernels of the form
   * K: Index x Index -> Double, i.e. kernel functions whose output
@@ -127,12 +129,20 @@ class DecomposableCovariance[S](kernels: LocalScalarKernel[S]*)(
       coupleAndKern._2.evaluate(u,v)
     }))
   }
+
+  override def gradient(x: S, y: S): Map[String, Double] = {
+    val (xs, ys) = (encoding*encoding)((x,y))
+    xs.zip(ys).zip(kernels).map(coupleAndKern => {
+      val (u,v) = coupleAndKern._1
+      coupleAndKern._2.gradient(u,v)
+    }).reduceLeft(_++_)
+  }
 }
 
 object DecomposableCovariance {
 
   val :+: = DataPipe((l: Array[Double]) => l.sum)
 
-  val :*: = DataPipe((l: Array[Double]) => l.sum)
+  val :*: = DataPipe((l: Array[Double]) => l.product)
 
 }

dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/models/gp/AbstractGPRegressionModel.scala

@@ -31,6 +31,8 @@ import io.github.mandar2812.dynaml.optimization.GloballyOptWithGrad
 import io.github.mandar2812.dynaml.probability.MultGaussianPRV
 import org.apache.log4j.Logger
 
+import scala.reflect.ClassTag
+
 /**
   * Single-Output Gaussian Process Regression Model
   * Performs gp/spline smoothing/regression with
@@ -46,7 +48,7 @@ import org.apache.log4j.Logger
   */
 abstract class AbstractGPRegressionModel[T, I](
   cov: LocalScalarKernel[I], n: LocalScalarKernel[I],
-  data: T, num: Int)
+  data: T, num: Int)(implicit ev: ClassTag[I])
   extends ContinuousProcess[T, I, Double, MultGaussianPRV]
   with SecondOrderProcess[T, I, Double, Double, DenseMatrix[Double], MultGaussianPRV]
   with GloballyOptWithGrad {

dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/models/stp/AbstractSTPRegressionModel.scala

@@ -32,14 +32,16 @@ import io.github.mandar2812.dynaml.probability.MultStudentsTPRV
 import io.github.mandar2812.dynaml.probability.distributions.{BlockedMultivariateStudentsT, MultivariateStudentsT}
 import org.apache.log4j.Logger
 
+import scala.reflect.ClassTag
+
 /**
   * @author mandar2812 date 26/08/16.
   * Implementation of a Students' T Regression model.
   */
 abstract class AbstractSTPRegressionModel[T, I](
   mu: Double, cov: LocalScalarKernel[I],
   n: LocalScalarKernel[I],
-  data: T, num: Int)
+  data: T, num: Int)(implicit ev: ClassTag[I])
   extends ContinuousProcess[T, I, Double, MultStudentsTPRV]
   with SecondOrderProcess[T, I, Double, Double, DenseMatrix[Double], MultStudentsTPRV]
   with GloballyOptimizable {