drasys.or.stat.model
Class GeneralLinearModel

java.lang.Object
  |
  +--drasys.or.stat.model.GeneralLinearModel

public class GeneralLinearModel

extends java.lang.Object

implements GeneralLinearModelI

An implementation of the general linear model.

References:

Applied Linear Statistical Models (Irwin Series in Statistics)
    Michael H. Kutner, et al / Hardcover / Published 1996
Numerical Recipes in C : The Art of Scientific Computing
    William H. Press, et al / Hardcover / Published 1993

Constructor Summary

GeneralLinearModel(VectorI dependent, MatrixI independent)           Initialize the model with the variables.

 

Method Summary
VectorI	getCoefficients()           Returns the solution coefficients.
VectorI	getDependent()           Returns the dependent vector.
int	getDFE()           Returns the number of degrees of freedom for the errors.
int	getDFR()           Returns the number of degrees of freedom for the regression.
int	getDFT()           Returns the total number of degrees.
MatrixI	getIndependent()           Returns the independent matrix.
MatrixI	getInverseXX()           Returns the X'X inverse matrix.
double	getMSE()           Returns the mean of the squares of the error terms.
double	getMSR()           Returns the mean of the squares of the computed dependent values.
double	getMST()           Returns the mean of the squares of the sampled dependent values.
double	getRSquared()           Returns R^2.
double	getSSE()           Returns the sum of the squares of the error terms.
double	getSSR()           Returns the sum of the squares of the computed dependent values.
double	getSST()           Returns the sum of the squares of the sampled dependent values.
VectorI	getStandardErrors()           Returns the standard errors of the fitted independent variables.
VectorI	solve()           Solve for the fitting coefficients using all of the independent variables.
VectorI	solve(boolean forceZeroIntercept)           Solve for the fitting coefficients using all of the independent variables.
VectorI	solve(boolean forceZeroIntercept, int[] columnSelection)           Solve for the coefficients using only the dependent columns listed in 'columnSelection'.
java.lang.String	toString()

 

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

 

Constructor Detail

GeneralLinearModel

public GeneralLinearModel(VectorI dependent,
                          MatrixI independent)

Initialize the model with the variables.

getRSquared

public double getRSquared()

Returns R^2. R^2 = ModelSumOfSquares()/TotalSumOfSquares();

Specified by:

getRSquared in interface GeneralLinearModelI

getSST

public double getSST()

Returns the sum of the squares of the sampled dependent values. SST = SSR + SSE.

Specified by:

getSST in interface GeneralLinearModelI

getSSE

public double getSSE()

Returns the sum of the squares of the error terms. The error terms are the differences between the sampled and computed dependent values. SST = SSR + SSE.

Specified by:

getSSE in interface GeneralLinearModelI

getSSR

public double getSSR()

Returns the sum of the squares of the computed dependent values. SST = SSR + SSE.

Specified by:

getSSR in interface GeneralLinearModelI

getDFR

public int getDFR()

Returns the number of degrees of freedom for the regression.

Specified by:

getDFR in interface GeneralLinearModelI

getDFE

public int getDFE()

Returns the number of degrees of freedom for the errors.

Specified by:

getDFE in interface GeneralLinearModelI

getDFT

public int getDFT()

Returns the total number of degrees.

Specified by:

getDFT in interface GeneralLinearModelI

getMST

public double getMST()

Returns the mean of the squares of the sampled dependent values. MST = SST/DFT.

Specified by:

getMST in interface GeneralLinearModelI

getMSE

public double getMSE()

Returns the mean of the squares of the error terms. The error terms are the differences between the sampled and computed dependent values. MSE = SSE/DFE.

Specified by:

getMSE in interface GeneralLinearModelI

getMSR

public double getMSR()

Returns the mean of the squares of the computed dependent values. MSR = SSR/DFR.

Specified by:

getMSR in interface GeneralLinearModelI

solve

public VectorI solve()

Solve for the fitting coefficients using all of the independent variables. This is equivalent to 'columnSelection' = {0,1,2,...}. The fitting line is allowed to cross the dependent axis away from the origin. and the intercept term will be the last value in the solution vector.

Specified by:

solve in interface GeneralLinearModelI

solve

public VectorI solve(boolean forceZeroIntercept)

Solve for the fitting coefficients using all of the independent variables. This is equivalent to 'columnSelection' = {0,1,2,...}. If 'forceZeroIntercept' is true, the fitted line will be forced to cross the dependent axis at the origin and the intercept value will NOT appear at the end of the solution vector. Otherwise; The fitting line is allowed to cross the dependent axis away from the origin. and the intercept term will be the last value in the solution vector.

Specified by:

solve in interface GeneralLinearModelI

solve

public VectorI solve(boolean forceZeroIntercept,
                     int[] columnSelection)

Solve for the coefficients using only the dependent columns listed in 'columnSelection'. The resluting coefficients will be ordered to match 'columnSelection'. If 'forceZeroIntercept' is true, the fitted line will be forced to cross the dependent axis at the origin and the intercept value will NOT appear at the end of the solution vector. Otherwise; The fitting line is allowed to cross the dependent axis away from the origin. and the intercept term will be the last value in the solution vector.

Specified by:

solve in interface GeneralLinearModelI

getDependent

public VectorI getDependent()

Returns the dependent vector.

Specified by:

getDependent in interface GeneralLinearModelI

getIndependent

public MatrixI getIndependent()

Returns the independent matrix.

Specified by:

getIndependent in interface GeneralLinearModelI

getCoefficients

public VectorI getCoefficients()

Returns the solution coefficients. The coefficient at index 'n' corresponds to the column indexed by 'columnSelection[n]'. If a constant intercept term is included, its value will be in the last element.

Specified by:

getCoefficients in interface GeneralLinearModelI

getStandardErrors

public VectorI getStandardErrors()

Returns the standard errors of the fitted independent variables. If a constant intercept term is included, its standard error will be in the last element.

Specified by:

getStandardErrors in interface GeneralLinearModelI

getInverseXX

public MatrixI getInverseXX()

Returns the X'X inverse matrix. If a constant intercept term is included, it will occupy the last row and column.

Specified by:

getInverseXX in interface GeneralLinearModelI

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object