Added MixturePipe Class

Author: mandar2812

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

@@ -0,0 +1,56 @@
+package io.github.mandar2812.dynaml.modelpipe
+
+import breeze.linalg.{DenseMatrix, DenseVector}
+import breeze.stats.distributions.{ContinuousDistr, Moments}
+import io.github.mandar2812.dynaml.algebra.{PartitionedPSDMatrix, PartitionedVector}
+import io.github.mandar2812.dynaml.models.gp.AbstractGPRegressionModel
+import io.github.mandar2812.dynaml.models.stp.AbstractSTPRegressionModel
+import io.github.mandar2812.dynaml.models.{ContinuousProcessModel, GenContinuousMixtureModel, SecondOrderProcessModel, StochasticProcessMixtureModel}
+import io.github.mandar2812.dynaml.optimization.GloballyOptimizable
+import io.github.mandar2812.dynaml.pipes.DataPipe2
+import io.github.mandar2812.dynaml.probability.{ContinuousRVWithDistr, MultGaussianPRV, MultStudentsTPRV}
+import io.github.mandar2812.dynaml.probability.distributions.{BlockedMultiVariateGaussian, BlockedMultivariateStudentsT, HasErrorBars}
+
+import scala.reflect.ClassTag
+
+/**
+  * Mixture Pipe takes a sequence of stochastic process models
+  * and associated probability weights and returns a mixture model.
+  * @author mandar2812 date 22/06/2017.
+  * */
+abstract class MixturePipe[
+T, I: ClassTag, Y, YDomain, YDomainVar,
+BaseDistr <: ContinuousDistr[YDomain]
+  with Moments[YDomain, YDomainVar]
+  with HasErrorBars[YDomain],
+W1 <: ContinuousRVWithDistr[YDomain, BaseDistr],
+BaseProcess <: ContinuousProcessModel[T, I, Y, W1]
+  with SecondOrderProcessModel[T, I, Y, Double, DenseMatrix[Double], W1]
+  with GloballyOptimizable] extends
+  DataPipe2[Seq[BaseProcess], DenseVector[Double],
+    GenContinuousMixtureModel[
+      T, I, Y, YDomain, YDomainVar,
+      BaseDistr, W1, BaseProcess]]
+
+
+class GPMixturePipe[T, I: ClassTag] extends
+  MixturePipe[T, I, Double, PartitionedVector, PartitionedPSDMatrix,
+    BlockedMultiVariateGaussian, MultGaussianPRV,
+    AbstractGPRegressionModel[T, I]] {
+
+  override def run(
+    models: Seq[AbstractGPRegressionModel[T, I]],
+    weights: DenseVector[Double]) =
+    StochasticProcessMixtureModel(models, weights)
+}
+
+class StudentTMixturePipe[T, I: ClassTag] extends
+  MixturePipe[T, I, Double, PartitionedVector, PartitionedPSDMatrix,
+    BlockedMultivariateStudentsT, MultStudentsTPRV,
+    AbstractSTPRegressionModel[T, I]] {
+
+  override def run(
+    models: Seq[AbstractSTPRegressionModel[T, I]],
+    weights: DenseVector[Double]) =
+    StochasticProcessMixtureModel(models, weights)
+}

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

@@ -10,8 +10,8 @@ import io.github.mandar2812.dynaml.probability.distributions.HasErrorBars
 import scala.reflect.ClassTag
 
 /**
-  * Created by mandar on 21/06/2017.
-  */
+  * @author mandar2812 date 21/06/2017.
+  * */
 abstract class MixtureMachine[
 T, I: ClassTag, Y, YDomain, YDomainVar,
 BaseDistr <: ContinuousDistr[YDomain]
@@ -71,7 +71,8 @@ BaseProcess <: ContinuousProcessModel[T, I, Y, W1]
     if(policy == "CSA") performCSA(initialConfig, options)
     else getEnergyLandscape(initialConfig, options, meanFieldPrior)
 
-  protected def modelProbabilities: DataPipe[Seq[(Double, Map[String, Double])], Seq[(Double, Map[String, Double])]] =
+  protected def modelProbabilities
+  : DataPipe[Seq[(Double, Map[String, Double])], Seq[(Double, Map[String, Double])]] =
     DataPipe(ProbGPCommMachine.calculateModelWeightsSigmoid(baselinePolicy))
 
   override def optimize(

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

@@ -2,8 +2,9 @@ package io.github.mandar2812.dynaml.optimization
 
 import breeze.linalg.DenseVector
 import io.github.mandar2812.dynaml.algebra.{PartitionedPSDMatrix, PartitionedVector}
+import io.github.mandar2812.dynaml.modelpipe._
 import io.github.mandar2812.dynaml.models.StochasticProcessMixtureModel
-import io.github.mandar2812.dynaml.models.gp.{AbstractGPRegressionModel, GaussianProcessMixture}
+import io.github.mandar2812.dynaml.models.gp.AbstractGPRegressionModel
 import io.github.mandar2812.dynaml.pipes.{DataPipe, DataPipe2}
 import io.github.mandar2812.dynaml.probability.MultGaussianPRV
 import io.github.mandar2812.dynaml.probability.distributions.BlockedMultiVariateGaussian
@@ -30,13 +31,12 @@ class ProbGPMixtureMachine[T, I: ClassTag](
 
   implicit val transform: DataPipe[T, Seq[(I, Double)]] = DataPipe(system.dataAsSeq)
 
-  override val confToModel = DataPipe((model_state: Map[String, Double]) =>
-    AbstractGPRegressionModel(
-    kernelPipe(model_state), noisePipe(model_state),
-    system.mean)(system.data, system.npoints))
+  override val confToModel = DataPipe(
+    (model_state: Map[String, Double]) =>
+      AbstractGPRegressionModel(
+        kernelPipe(model_state), noisePipe(model_state),
+        system.mean)(system.data, system.npoints))
 
-  override val mixturePipe = DataPipe2(
-    (models: Seq[AbstractGPRegressionModel[T, I]], weights: DenseVector[Double]) =>
-    StochasticProcessMixtureModel[T, I](models, weights))
+  override val mixturePipe = new GPMixturePipe[T, I]
 
 }

scripts/stochasticPriors.sc

@@ -1,5 +1,5 @@
 import breeze.linalg.eig
-import breeze.stats.distributions.{ContinuousDistr, Gamma, Gaussian}
+import breeze.stats.distributions.{ContinuousDistr, Gamma}
 import io.github.mandar2812.dynaml.kernels._
 import io.github.mandar2812.dynaml.models.bayes.{LinearTrendESGPrior, LinearTrendGaussianPrior}
 import io.github.mandar2812.dynaml.probability._