Version 1.4.3 begun!

 - Added feature in ParallelPipe to access individual pipes
 - Added scaling function for CoRegGPPrior

build.sbt

@@ -12,7 +12,7 @@ packageDescription := "DynaML is a scala library/repl for implementing and worki
   "which can be extended easily to implement advanced models for small and large scale applications.\n\n"+
   "But the library can also be used as an educational/research tool for data analysis."
 
-val mainVersion = "v1.4.2"
+val mainVersion = "v1.4.3-beta.1"
 
 val dataDirectory = settingKey[File]("The directory holding the data files for running example scripts")
 

copy-docs.sh

@@ -1,6 +1,6 @@
 #! /bin/bash
 
-cp -R dynaml-core/target/scala-2.11/api/* docs/api_docs/v1.4.2/dynaml-core/
-cp -R dynaml-pipes/target/scala-2.11/api/* docs/api_docs/v1.4.2/dynaml-pipes/
-cp -R dynaml-examples/target/scala-2.11/api/* docs/api_docs/v1.4.2/dynaml-examples/
-cp -R target/scala-2.11/api/* docs/api_docs/v1.4.2/dynaml-repl/
+cp -R dynaml-core/target/scala-2.11/api/* docs/api_docs/v1.4.3/dynaml-core/
+cp -R dynaml-pipes/target/scala-2.11/api/* docs/api_docs/v1.4.3/dynaml-pipes/
+cp -R dynaml-examples/target/scala-2.11/api/* docs/api_docs/v1.4.3/dynaml-examples/
+cp -R target/scala-2.11/api/* docs/api_docs/v1.4.3/dynaml-repl/

dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/models/bayes/GaussianProcessPrior.scala

@@ -28,7 +28,7 @@ import io.github.mandar2812.dynaml.modelpipe.GPRegressionPipe2
 import io.github.mandar2812.dynaml.models.gp.AbstractGPRegressionModel
 
 import scala.reflect.ClassTag
-import io.github.mandar2812.dynaml.pipes.{DataPipe, MetaPipe}
+import io.github.mandar2812.dynaml.pipes.{DataPipe, MetaPipe, ParallelPipe}
 import io.github.mandar2812.dynaml.probability.{MatrixNormalRV, MultGaussianPRV}
 
 /**
@@ -213,6 +213,7 @@ abstract class CoRegGPPrior[I: ClassTag, J: ClassTag, MeanFuncParams](
   noiseCovarianceI: LocalScalarKernel[I], noiseCovarianceJ: LocalScalarKernel[J]) extends
   GaussianProcessPrior[(I,J), MeanFuncParams](covarianceI:*covarianceJ, noiseCovarianceI:*noiseCovarianceJ) {
 
+  self =>
 
   def priorDistribution[U <: Seq[I], V <: Seq[J]](d1: U, d2: V): MatrixNormalRV = {
 
@@ -224,6 +225,26 @@ abstract class CoRegGPPrior[I: ClassTag, J: ClassTag, MeanFuncParams](
     MatrixNormalRV(m, u, v)
   }
 
+  /**
+    * Define a prior over the process which is a scaled version of the base GP.
+    *
+    * z ~ GP(m(.), K(.,.))
+    *
+    * y = g(x)&times;z
+    *
+    * y ~ GP(g(x)&times;m(x), g(x)K(x,x')g(x'))
+    **/
+  def *(scalingFunc: ParallelPipe[I, Double, J, Double]) =
+    CoRegGPPrior(
+      ScaledKernel[I](self.covarianceI, scalingFunc._1),
+      ScaledKernel[J](self.covarianceJ, scalingFunc._2),
+      ScaledKernel[I](self.noiseCovarianceI, scalingFunc._1),
+      ScaledKernel[J](self.noiseCovarianceJ, scalingFunc._2))(
+      MetaPipe((p: MeanFuncParams) => (x: (I, J)) => {
+        self.meanFunctionPipe(p)(x)*scalingFunc._1(x._1)*scalingFunc._2(x._2)
+      }),
+      self._meanFuncParams
+    )
 }
 
 object CoRegGPPrior {

dynaml-pipes/src/main/scala-2.11/io/github/mandar2812/dynaml/pipes/DataPipe.scala

@@ -98,13 +98,21 @@ object DataPipe {
 trait ParallelPipe[-Source1, +Result1, -Source2, +Result2]
   extends DataPipe[(Source1, Source2), (Result1, Result2)] {
 
+  val _1: DataPipe[Source1, Result1]
+  val _2: DataPipe[Source2, Result2]
+
 }
 
 object ParallelPipe {
   def apply[S1, D1, S2, D2](func1: (S1) => D1, func2: (S2) => D2):
   ParallelPipe[S1, D1, S2, D2] = {
     new ParallelPipe[S1, D1, S2, D2] {
+
       def run(data: (S1, S2)) = (func1(data._1), func2(data._2))
+
+      override val _1 = DataPipe(func1)
+
+      override val _2 = DataPipe(func2)
     }
   }
 }