org.gersteinlab.regulation.model
Class RateModelSqrDiff

java.lang.Object
  extended by org.gersteinlab.regulation.numeric.MultivariateFunction
      extended by org.gersteinlab.regulation.numeric.DifferentiableMultivariateFunction
          extended by org.gersteinlab.regulation.numeric.DoublyDifferentiableMultivariateFunction
              extended by org.gersteinlab.regulation.model.RateModelSqrDiff

public class RateModelSqrDiff
extends DoublyDifferentiableMultivariateFunction

        A function that equals the total squared difference between the
        predicted value of a differential equation rate model and some target
        values at a set of time points, treating the parameters of the model as
        the variables.
        

Version:
1.0 (September 1, 2008)
        Change History:
        1.0     - Initial version
        
Author:
Kevin Yuk-Lap Yip

Field Summary
 
Fields inherited from class org.gersteinlab.regulation.numeric.MultivariateFunction
xStored
 
Constructor Summary
RateModelSqrDiff(double[] tIn, double[][] yjIn, double[][][][] yksIn, double[][][][] dykdtsIn, java.lang.Class<? extends RateModel> modelClass)
          Create the function with the given set of points and target values and differentials.
 
Method Summary
 double f()
          Evaluate the total squared difference at the specified parameter values.
 double[] gf()
          Evaluate the gradient of the function at the specified parameter values.
 double[][] hf()
          Evaluate the Hessian of the function at the specified parameter values.
 void set(double[] a)
          Set the parameter values of the model.
 
Methods inherited from class org.gersteinlab.regulation.numeric.DoublyDifferentiableMultivariateFunction
hf
 
Methods inherited from class org.gersteinlab.regulation.numeric.DifferentiableMultivariateFunction
gf
 
Methods inherited from class org.gersteinlab.regulation.numeric.MultivariateFunction
f
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

RateModelSqrDiff

public RateModelSqrDiff(double[] tIn,
                        double[][] yjIn,
                        double[][][][] yksIn,
                        double[][][][] dykdtsIn,
                        java.lang.Class<? extends RateModel> modelClass)
                 throws java.lang.InstantiationException,
                        java.lang.IllegalAccessException,
                        java.lang.IllegalArgumentException,
                        java.lang.reflect.InvocationTargetException,
                        java.lang.NoSuchMethodException
Create the function with the given set of points and target values and differentials.

Parameters:
tIn - The sets of points
yjIn - The sets of target values First dimension: experiment Second dimension: time point
yksIn - The sets of predictors First dimention: predictor set Second dimension: predictor Third dimension: experiment Fourth dimension: time point
dykdtsIn - The sets of differentials of the predictors First dimension: predictor set Second dimension: predictor Third dimension: experiment Fourth dimension: time point
modelClass - The class of the rate model
Throws:
java.lang.InstantiationException
java.lang.IllegalAccessException
java.lang.IllegalArgumentException
java.lang.reflect.InvocationTargetException
java.lang.NoSuchMethodException
Method Detail

set

public void set(double[] a)
Set the parameter values of the model.

Overrides:
set in class MultivariateFunction
Parameters:
a - The parameter values

f

public double f()
Evaluate the total squared difference at the specified parameter values.

Overrides:
f in class MultivariateFunction
Returns:
The squred difference

gf

public double[] gf()
Evaluate the gradient of the function at the specified parameter values.

Overrides:
gf in class DifferentiableMultivariateFunction
Returns:
The gradient

hf

public double[][] hf()
Evaluate the Hessian of the function at the specified parameter values.

Overrides:
hf in class DoublyDifferentiableMultivariateFunction
Returns:
The Hessian