bijfit
Class DesignMatrixGamma

java.lang.Object
  |
  +--bijfit.DesignMatrix
        |
        +--bijfit.DesignMatrixGamma

public class DesignMatrixGamma
extends DesignMatrix

Design Matrix for gamma variate model, to be used in GeneralLeastSquares estimation. As described by Madsen, 1992 Gamma variate y(t) = A(t-t0)^alpha exp(-(t - t0) / beta) can be rewritten as y(t') = ymax t' ^ alpha exp(alpha(1 - t')) where t' = (t-t0) / (tmax - t0) and tmax is the t where y(t) = max. Linear form: ln(y(t')) = ln(ymax) + alpha(1 + ln(t') - t')) Copyright (c) 1999-2003, Michael Abramoff. All rights reserved.


Fields inherited from class bijfit.DesignMatrix
a
 
Constructor Summary
DesignMatrixGamma(float[] t, float tmax, float t0)
          Create a design matrix for the derived gamma variate function.
 
Method Summary
protected  double[] fgamma(float x, float xmax, float x0)
          Return a double[] vector containing the basis functions at x for a derived gamma variate.
 
Methods inherited from class bijfit.DesignMatrix
getMatrix, toString
 
Methods inherited from class java.lang.Object
, clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait
 

Constructor Detail

DesignMatrixGamma

public DesignMatrixGamma(float[] t,
                         float tmax,
                         float t0)
Create a design matrix for the derived gamma variate function.
Parameters:
t - the time stamps.
tmax - the time at which the curve reaches maximum height.
Method Detail

fgamma

protected double[] fgamma(float x,
                          float xmax,
                          float x0)
Return a double[] vector containing the basis functions at x for a derived gamma variate. Transforms x based on tmax
Parameters:
x - at which the derived gamma variate is to be evaluated.
Returns:
a double[] vector containing the coefficients at x.