Package edu.cnu.mdi.splot.fit

Class FitResult

java.lang.Object

edu.cnu.mdi.splot.fit.FitResult

public final class FitResult
extends Object

Generic result for a (weighted) least-squares fit.

This class is intentionally model-agnostic: it stores parameter values and common diagnostics that apply across nonlinear and linear least-squares fits (Gaussian, erf/erfc, exponential, polynomial, etc.).

Field Summary

Fields

Modifier and Type	Field	Description
static final String	CHI
static final String	CHISQ
final double	chiSquare	Chi-square: sum r_i^2 = cost^2 (weighted if weights used).
final double	chiSquareReduced	Reduced chi-square: chiSquare / dof.
final double	cost	Least-squares cost: sqrt(sum r_i^2) where r_i are residuals.
final org.apache.commons.math3.linear.RealMatrix	covariance	Optional covariance matrix of parameters (p x p), or null if unavailable.
final int	dof	Degrees of freedom: max(1, nPoints - nParameters).
final int	evaluations	Evaluations used by an iterative optimizer; may be 0 for closed-form fits.
Evaluator	evaluator	"Use as a function" evaluator for the fit.
final int	iterations	Iterations used by an iterative optimizer; may be 0 for closed-form fits.
final double[]	params	Best-fit parameters.
static final String	PLUSMINUS
final double	rms	RMS of residuals (same residual space as optimizer).
IFitStringGetter	stringGetter	Gets descriptive strings from fitter

Constructor Summary

Constructors

Modifier	Constructor	Description
protected	FitResult(IFitStringGetter stringGetter, double[] params, org.apache.commons.math3.linear.RealMatrix covariance, double cost, double chiSquare, int dof, double chiSquareReduced, double rms, int iterations, int evaluations)	Create a fit result.

Method Summary

Modifier and Type	Method	Description
boolean	hasCovariance()
String	htmlSummary()	HTML summary of the fit result.
int	nParams()
double	param(int i)	Convenience access to parameter i.
double	paramStdError(int i)	Standard error for parameter i (NaN if unavailable).
double[]	paramStdErrors()	Standard errors for parameters computed from covariance: sqrt(cov[i,i]).
void	setEvaluator(Evaluator evaluator)	Set the evaluator.
String	singleLineSummary()	Single-line text summary of the fit result.
String	toString()

Methods inherited from class java.lang.Object

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

Field Details

CHI

public static final String CHI

See Also:

• Constant Field Values

CHISQ

public static final String CHISQ

See Also:

• Constant Field Values

PLUSMINUS

public static final String PLUSMINUS

See Also:

• Constant Field Values

params

public final double[] params

Best-fit parameters. Interpretation depends on the model.

stringGetter

public IFitStringGetter stringGetter

Gets descriptive strings from fitter

covariance

public final org.apache.commons.math3.linear.RealMatrix covariance

Optional covariance matrix of parameters (p x p), or null if unavailable.

cost

public final double cost

Least-squares cost: sqrt(sum r_i^2) where r_i are residuals. If weights are used, r_i are in the weighted residual space.

chiSquare

public final double chiSquare

Chi-square: sum r_i^2 = cost^2 (weighted if weights used).

dof

public final int dof

Degrees of freedom: max(1, nPoints - nParameters).

chiSquareReduced

public final double chiSquareReduced

Reduced chi-square: chiSquare / dof.

rms

public final double rms

RMS of residuals (same residual space as optimizer).

iterations

public final int iterations

Iterations used by an iterative optimizer; may be 0 for closed-form fits.

evaluations

public final int evaluations

Evaluations used by an iterative optimizer; may be 0 for closed-form fits.

evaluator

public Evaluator evaluator

"Use as a function" evaluator for the fit.

Constructor Details

FitResult

protected FitResult(IFitStringGetter stringGetter,
 double[] params,
 org.apache.commons.math3.linear.RealMatrix covariance,
 double cost,
 double chiSquare,
 int dof,
 double chiSquareReduced,
 double rms,
 int iterations,
 int evaluations)

Create a fit result.

Parameters:

stringGetter - access to descriptive strings about fitter

params - best-fit parameters

covariance - optional covariance matrix (may be null)

cost - least-squares cost

chiSquare - chi-square

dof - degrees of freedom

chiSquareReduced - reduced chi-square

rms - RMS of residuals

iterations - iterations used (0 if not applicable)

evaluations - evaluations used (0 if not applicable)

Method Details

setEvaluator

public void setEvaluator(Evaluator evaluator)

Set the evaluator.

Parameters:

evaluator -

nParams

public int nParams()

Returns:

number of parameters.

param

public double param(int i)

Convenience access to parameter i.

hasCovariance

public boolean hasCovariance()

Returns:

true if covariance is present.

paramStdErrors

public double[] paramStdErrors()

Standard errors for parameters computed from covariance: sqrt(cov[i,i]). Returns NaN for entries where covariance is missing or diagonal is non-positive.

paramStdError

public double paramStdError(int i)

Standard error for parameter i (NaN if unavailable).

htmlSummary

public String htmlSummary()

HTML summary of the fit result.

Returns:

HTML string

singleLineSummary

public String singleLineSummary()

Single-line text summary of the fit result.

Returns:

single-line string

toString

public String toString()

Overrides:

toString in class Object