Target function - The target function for OLA Averaging Kernels.

Functions
bool	checkTargetDim (MATRIX *f, int Nz)
void	newTarget (MATRIX f, int N, int Ny, int Nz, BaseDirectory *baseDir, bool debug, bool useFourierTransform)
MAP_DEPTH_MATRIX *	constructTarget (void(*NormalizeFz)(), void(*Buildup)(int, int, MAP_DEPTH_MATRIX *, FloatComplex *, FloatComplex **))
void	writeTarget (Float sigmah, int sigmaStep)
void	resetTarget (Float sigmah, int sigmaStep)
MATRIX	readFullTarget (char *inputFilename, MultiDimFits targetInfo, int N, int Ny, int Nz, Float sigmaCurrentValue, Float sigmaCurrentStep, BaseDirectory *baseDir, bool debug, bool useFourierTransform)
void	readTarget (TargetInfo targetInfo, int N, int Ny, int Nz, Float sigmaCurrentValue, Float sigmaCurrentStep, BaseDirectory *baseDir, bool debug, bool useFourierTransform)
MAP_DEPTH_MATRIX *	getTarget ()
void	freeTarget ()

---

Detailed Description

The target function is implemented as a 2-D Gaussian in x and y, multiplied by a a 1-D function of z. The 1-D function is an input parameter. The Gaussian has an input parameter, sigma_h, which specifies the horizontal full-width at half-maximum (FWHM) of the function.

The target function is precalculated for x, y and z values, then Fourier transformed.

---

Function Documentation

bool checkTargetDim	(	MATRIX *	f,
		int	Nz
	)

Verify the dimensions of the target function are compatible with the dimensions of the kernels. Oh, but if the inversion is NOT OLA, then only allow (1xNz) ... so don't use this function.

Parameters:

[in,out]	f	1D function f(z) for target function when z > 1. This parameter is in/out because it is NFz0 x Nz, and if the FITS file is Nz x NFz0, then we can easily take the transpose and still run the inversions.
[in]	Nz	Size of the z-dimension of the kernels.

Returns:

bool Returns true if dimensions are compatible, false otherwise.

MAP_DEPTH_MATRIX* constructTarget	(	void(*)()	NormalizeFz,
		void(*)(int, int, MAP_DEPTH_MATRIX *, FloatComplex *, FloatComplex **)	Buildup
	)

Construct target for multiple depths

Construct Target function from preset parameters (Gaussian keeps sigma_h, Target_f should already be set to f(z;z0) using newTarget() function).

Returns:

Target function stored in a 4-D MAP_DEPTH_MATRIX in wavespace.

void freeTarget

(

)

Free space associated with target function.

MAP_DEPTH_MATRIX* getTarget

(

)

Returns a pointer to target function represented by a Map-Depth matrix.

Returns:

Pointer to MAP_DEPTH_MATRIX representing the generated target function.

void newTarget	(	MATRIX	f,
		int	N,
		int	Ny,
		int	Nz,
		BaseDirectory *	baseDir,
		bool	debug,
		bool	useFourierTransform
	)

Set new parameters for Target function.

Parameters:

[in]	f	A MATRIX that contains a 1D function for z-dimension; but it can contain multiple 1D functions, one for each target depth.
[in]	N	X-dimension of target function in frequency space.
[in]	Ny	Y-dimension in frequency space if useFourierTransform is true; different than Nx because of Fourier Transform; otherwise, Y-dimension in real space.
[in]	Nz	Z-dimension.
[in]	baseDir	Output base directory.
[in]	debug	True if running with debug flag; False otherwise.
[in]	useFourierTransform	True if transforming target function to wavespace; False otherwise. Multiply 2D Gaussian by this function to get 3D Target function.

MATRIX readFullTarget	(	char *	inputFilename,
		MultiDimFits	targetInfo,
		int	N,
		int	Ny,
		int	Nz,
		Float	sigmaCurrentValue,
		Float	sigmaCurrentStep,
		BaseDirectory *	baseDir,
		bool	debug,
		bool	useFourierTransform
	)

readFullTarget Read the full target function from a fits file, not just the depth function, f(z,z0).

Parameters:

[in]	inputFilename	Full path the to fits file containing the target function.
[in]	targetInfo	Read the two types of input data for a target function. First type is a matrix (Nfz0 x Nz) containing the z-dimension for possibly multiple target depth functions, f(z;z_0). The second type is a complete target function of dimension (Nx x Ny x Nz x Nbeta). Multiple target functions are only for OLA inversions.
[in]	inputFilename	Character string for name of file containing target function.
[in]	targetInfo	Internal structure (MultiDimFits) that contains information about an open fits file.
[in]	N	Size of Nx dimension of kernels
[in]	Ny	Size of Ny dimension of kernels
[in]	Nz	Size of Nz dimension of kernels
[in]	sigmaCurrentValue	Current value of sigma_h parameter (FWHM).
[in]	sigmaCurrentStep	sigmaStep The nth sigma step in the parameter range.
[in]	baseDir	pointer to BaseDirectory containing the base directory for all output from the inversion.
[in]	debug	When set to true, intermediate values are written to output directory.
[in]	useFourierTransform	true if Fourier transforming the target function on input.

void readTarget	(	TargetInfo	targetInfo,
		int	N,
		int	Ny,
		int	Nz,
		Float	sigmaCurrentValue,
		Float	sigmaCurrentStep,
		BaseDirectory *	baseDir,
		bool	debug,
		bool	useFourierTransform
	)

Read the two types of input data for a target function. First type is a matrix (Nfz0 x Nz) containing the z-dimension for possibly multiple target depth functions, f(z;z_0). The second type is a complete target function of dimension (Nx x Ny x Nz x Nbeta). Multiple target functions are only for OLA inversions.

Parameters:

[in]	targetInfo	Internal structure containing info about the target function.
[in]	N	Size of Nx dimension of kernels
[in]	Ny	Size of Ny dimension of kernels
[in]	Nz	Size of Nz dimension of kernels
[in]	sigmaCurrentValue	Current value of sigma_h parameter (FWHM).
[in]	sigmaCurrentStep	sigmaStep The nth sigma step in the parameter range.
[in]	baseDir	pointer to BaseDirectory containing the base directory for all output from the inversion.
[in]	debug	When set to true, intermediate values are written to output directory.
[in]	useFourierTransform	true if Fourier transforming the target function on input.

void resetTarget	(	Float	sigmah,
		int	sigmaStep
	)

Change the target function by recalculating with sigmah.

Parameters:

[in]	sigmah	Full-width at half maximum of function.
[in]	sigmaStep	The nth sigma step in the parameter range.

void writeTarget	(	Float	sigmah,
		int	sigmaStep
	)

Write target function to output FITS file. Note: the output filename is assumed to be set using newTarget.

Parameters:

[in]	sigmah	Current value for sigmah parameter, where sigmah is the full-width at half-maximum (FWHM) of the Gaussian.
[in]	sigmaStep	The nth parameter value for sigmah.