stella.adapter

Class ModelDrift.PinholeDrift

java.lang.Object

astro.fits.AbstractFitsImageModel

stella.adapter.AbstractPinholeModel.AbstractPinholeFits

stella.adapter.ModelDrift.PinholeDrift

All Implemented Interfaces:

GradientMultidimensionalModel, LevenbergMarquardtCapable, MultidimensionalDataModel

Enclosing class:

ModelDrift

public static class ModelDrift.PinholeDrift
extends AbstractPinholeModel.AbstractPinholeFits
implements LevenbergMarquardtCapable, GradientMultidimensionalModel

A pinhole model is the model of a star with fitted intensity and FWHM close to the fixed pinhole position at the given coordinates. We fit four parameters, the Δx and Δy of the relative star shift, the total flux of the star and the FWHM of the star.

Field Summary

Fields

Modifier and Type	Field and Description
private double	beta The beta of the moffat function.
private VectorG	m0 A first estimate of the model parameters.
private double	pindip For modeling the pinhole.
private Point2D	pinhole The fixed pinhole position.
private double	pinsize The pinhole size in pixel.
private double	pinsteep For modeling the pinhole, the steepness.

Constructor Summary

Constructors

Modifier	Constructor and Description
private	PinholeDrift(MImageHDU measured, Rectangle box, double back, double backerr, Point2D pos, double size, double k, double deep, double moffat) Constructs a new data model for the pinhole.

Method Summary

Modifier and Type	Method and Description
VectorG	evaluateModel(VectorG dxdy, VectorG xy) Per definition, the independant variable, time, is here a two-dimensional vector specifying the location of the pixel.
VectorG[]	evaluateModelGradient(VectorG dxdy, VectorG at)
Multidimensional	getChiSquareModel()
VectorG	getFirstModel() Returns a first estimate.
VectorG	getFirstStep() A step estimate.
private MyMatrix	getGradientMatrix(VectorG dxdy) Returns the gradient matrix.
private VectorG[]	getGradientsAtMeasure(int row, VectorG dxdy)
VectorG	getNegativeChi2HalfGradient(VectorG better)
int	getParameterCount() The model parameters are Shift of star to pinhole in x-direction, pixel Shift of star to pinhole in y-direction, pixel Total ADU flux of star FWHM of star in pixel. Shift of true pinhole to pinhole-0 in x-direction, pixel Shift of true pinhole to pinhole-0 in y-direction, pixel Thus we
QuadMatrix	getPseudoHessian(VectorG better)
double[]	getStarModel(VectorG dxdy) Per definition, the model data is a Vector of all data points, not an area.

Methods inherited from class stella.adapter.AbstractPinholeModel.AbstractPinholeFits

applyBackground, applyBackground

Methods inherited from class astro.fits.AbstractFitsImageModel

aboveBackground, dimension, dimensionOfIndependent, equalData, getArea, getFloatImageData, getFloatImageDataErrors, getFloatImageModel, getFloatImageResiduals, getImageData, getImageDataErrors, getImageModel, getImageResiduals, getMeasureCount, getMeasurementErrors, getMeasures, getMetric, getModel, getResiduals, getRms, getTimes, getTotalErrors, getUpLeft, getYX, photonAndReadoutNoise, setAduErrors, subtract

Methods inherited from class java.lang.Object

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

Methods inherited from interface math.MultidimensionalDataModel

dimension, dimensionOfIndependent, getMeasureCount, getMeasurementErrors, getMeasures, getMetric, getModel, getResiduals, getTimes, getTotalErrors

Field Detail

pinhole

private final Point2D pinhole

The fixed pinhole position.

pinsize

private double pinsize

The pinhole size in pixel.

pinsteep

private final double pinsteep

For modeling the pinhole, the steepness.

pindip

private final double pindip

For modeling the pinhole.

beta

private final double beta

The beta of the moffat function.

m0

private final VectorG m0

A first estimate of the model parameters.

Constructor Detail

PinholeDrift

private PinholeDrift(MImageHDU measured,
                     Rectangle box,
                     double back,
                     double backerr,
                     Point2D pos,
                     double size,
                     double k,
                     double deep,
                     double moffat)
              throws IOException,
                     BasicFitsException

Constructs a new data model for the pinhole. The model is constructed with the measured data (plus the applicable box) and the measured backgound noise. The modeling parameters for the pinhole and the beta of the stellar Moffat function are also stated.

Throws:

IOException

BasicFitsException

Method Detail

getParameterCount

public int getParameterCount()

The model parameters are

• Shift of star to pinhole in x-direction, pixel
• Shift of star to pinhole in y-direction, pixel
• Total ADU flux of star
• FWHM of star in pixel.
• Shift of true pinhole to pinhole-0 in x-direction, pixel
• Shift of true pinhole to pinhole-0 in y-direction, pixel

Thus we

Specified by:

getParameterCount in interface MultidimensionalDataModel

Returns:

6.

getFirstModel

public VectorG getFirstModel()

Returns a first estimate.

Specified by:

getFirstModel in class AbstractPinholeModel.AbstractPinholeFits

getFirstStep

public VectorG getFirstStep()

A step estimate.

Specified by:

getFirstStep in class AbstractPinholeModel.AbstractPinholeFits

getStarModel

public double[] getStarModel(VectorG dxdy)

Per definition, the model data is a Vector of all data points, not an area. The model parameters are

• Shift of star to pinhole-0 in x-direction, pixel
• Shift of star to pinhole-0 in y-direction, pixel
• Total ADU flux of star
• FWHM of star in pixel.
• Shift of true pinhole to pinhole-0 in x-direction, pixel
• Shift of true pinhole to pinhole-0 in y-direction, pixel

Specified by:

getStarModel in class AbstractPinholeModel.AbstractPinholeFits

evaluateModel

public VectorG evaluateModel(VectorG dxdy,
                             VectorG xy)

Per definition, the independant variable, time, is here a two-dimensional vector specifying the location of the pixel. It can be non-integer. The model parameters are

• Shift of star to pinhole in x-direction, pixel
• Shift of star to pinhole in y-direction, pixel
• Total ADU flux of star
• FWHM of star in pixel.

Specified by:

evaluateModel in interface MultidimensionalDataModel

Parameters:

dxdy - The model parameters as defined above

The - pixel location as a two-dim vector.

evaluateModelGradient

public VectorG[] evaluateModelGradient(VectorG dxdy,
                                       VectorG at)

Specified by:

evaluateModelGradient in interface GradientMultidimensionalModel

getGradientsAtMeasure

private VectorG[] getGradientsAtMeasure(int row,
                                        VectorG dxdy)

getGradientMatrix

private MyMatrix getGradientMatrix(VectorG dxdy)

Returns the gradient matrix. Only the stellar model via the Moffat function depends on the model parameters.

getNegativeChi2HalfGradient

public VectorG getNegativeChi2HalfGradient(VectorG better)

Specified by:

getNegativeChi2HalfGradient in interface LevenbergMarquardtCapable

getPseudoHessian

public QuadMatrix getPseudoHessian(VectorG better)

Specified by:

getPseudoHessian in interface LevenbergMarquardtCapable

getChiSquareModel

public Multidimensional getChiSquareModel()

Specified by:

getChiSquareModel in interface LevenbergMarquardtCapable