vec_math
Class ExpressionFit.AmoebaModel

java.lang.Object
  vec_math.ExpressionFit.AmoebaModel

All Implemented Interfaces:

DataModel

Enclosing class:

ExpressionFit

protected static class ExpressionFit.AmoebaModel
extends Object
implements DataModel

A multidimensional that evaluates a function f given as a string like

       a0*cos(a1*$t+a2)
       

on all input data, such that it returns

       Σ(y_i-y_av-f_i)&pow2;
       

The independet variable is always at index zero in the input data, and the dependent y_i is at the index provided at construct.

Field Summary

private VectorG[]	data Input data, on each VectorG only zero and yindex is used.
private ExpressionNode	fit The expression that is the model.
static String	TIME The name of the independant variable in the expression.
private double	yav The average of the dependant variable.
private int	yindex The index of the dependant variable.

Constructor Summary

protected	ExpressionFit.AmoebaModel(String model) Only constructs with an expression node.
	ExpressionFit.AmoebaModel(String model, VectorG[] mag, int y) Constructs a new Amoeba Model.

Method Summary

double	evaluateModel(VectorG a, VectorG time) Returns the model at a single point.
int	getMeasureCount() The number of dimensions is the variables count minus one.
double[]	getMeasurementErrors() Returns null.
double[]	getMeasures() Returns the measures.
double[]	getModel(VectorG a) The model
private double[]	getModelAtData(VectorG a) Returns the current model fit at all data points.
int	getParameterCount() The number of dimensions is the variables count minus one.
double[]	getResiduals(VectorG a) Returns the residuals of the measures to the data model.
VectorG[]	getTimes() Returns the times, this is an array of 1-d vectors initialized to data[].0.
double[]	getTotalErrors() Returns null.
private String	printModel(VectorG a) We convert the expression parsed for printing, and during evaluate.
protected boolean	setModelData(VectorG[] matrix) Sets the model data and the model index.
protected boolean	setModelIndex(int yi) Sets the model data index.

Methods inherited from class java.lang.Object

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

Field Detail

TIME

public static final String TIME

The name of the independant variable in the expression.

See Also:

Constant Field Values

data

private VectorG[] data

Input data, on each VectorG only zero and yindex is used.

yindex

private int yindex

The index of the dependant variable.

yav

private double yav

The average of the dependant variable.

fit

private ExpressionNode fit

The expression that is the model.

Constructor Detail

ExpressionFit.AmoebaModel

public ExpressionFit.AmoebaModel(String model,
                                 VectorG[] mag,
                                 int y)
                          throws ParseException

Constructs a new Amoeba Model.

Throws:

ParseException

ExpressionFit.AmoebaModel

protected ExpressionFit.AmoebaModel(String model)
                             throws ParseException

Only constructs with an expression node. Is protected to hinder unauthorized use. Subclasses must ensure that the model data and index is set, according to setModelData(vec_math.VectorG[]) and setModelIndex(int) before this multidimensional is used.

Throws:

ParseException

Method Detail

setModelData

protected boolean setModelData(VectorG[] matrix)

Sets the model data and the model index. Only after a call to this method and setModelIndex(int), the amoeba model may be used.

Returns:

True, if model can be used now.

setModelIndex

protected boolean setModelIndex(int yi)

Sets the model data index. . Only after a call to this method and setModelData(vec_math.VectorG[]), the amoeba model may be used.

Returns:

True, if model can be used now.

getParameterCount

public int getParameterCount()

The number of dimensions is the variables count minus one.

Specified by:

getParameterCount in interface DataModel

Returns:

The number of parameters in the model. This equals the dimension of the input vector in DataModel.getModel(vec_math.VectorG).

getMeasureCount

public int getMeasureCount()

The number of dimensions is the variables count minus one.

Specified by:

getMeasureCount in interface DataModel

Returns:

The number of measurements in the data model. This equals the dimension of the return vector in DataModel.getModel(vec_math.VectorG).

getMeasures

public double[] getMeasures()

Returns the measures.

Specified by:

getMeasures in interface DataModel

getTotalErrors

public double[] getTotalErrors()

Returns null.

Specified by:

getTotalErrors in interface DataModel

getMeasurementErrors

public double[] getMeasurementErrors()

Returns null.

Specified by:

getMeasurementErrors in interface DataModel

getTimes

public VectorG[] getTimes()

Returns the times, this is an array of 1-d vectors initialized to data[].0.

Specified by:

getTimes in interface DataModel

Returns:

An array of dimension DataModel.getMeasureCount().

getModel

public double[] getModel(VectorG a)

The model

Specified by:

getModel in interface DataModel

Parameters:

a - The model parameters

Returns:

A model for the data points depending on the model parameters.

getResiduals

public double[] getResiduals(VectorG a)

Returns the residuals of the measures to the data model. This equals calling getMeasures() and getModel(vec_math.VectorG) and returning the difference between both.

Specified by:

getResiduals in interface DataModel

Parameters:

a - The model parameters, not necessarily the best fit.

getModelAtData

private double[] getModelAtData(VectorG a)
                         throws ParseException

Returns the current model fit at all data points.

Throws:

ParseException

evaluateModel

public double evaluateModel(VectorG a,
                            VectorG time)

Returns the model at a single point.

Specified by:

evaluateModel in interface DataModel

Parameters:

a - The model parameters

time - The (one)dimensional vector of independant variables.

printModel

private String printModel(VectorG a)

We convert the expression parsed for printing, and during evaluate.