Tools for processing signals

Classes
struct	mx::sigproc::autocorrelationFromPSD< T >
	Functor for calculating the autocorrelation given a PSD. More...

struct	mx::sigproc::linearPredictor< _realT >
	A class to support linear prediction. More...

Modules
	Power Spectra

	The Fourier Basis

	The Zernike Basis

	Window Functions

	Image Filters

	Signal Processing Files

	Circular Buffer
	A circular buffer class.

Functions
template<typename T >
void	mx::sigproc::autocorrelation (T ac, size_t Nac, T sig, size_t Nsig)
	Calculate the autocorrelation of a time-series.

template<typename T >
void	mx::sigproc::autocorrelation (std::vector< T > &ac, std::vector< T > &sig)
	Calculate the autocorrelation of a time-series.

template<typename realT >
int	mx::sigproc::basisMask (improc::eigenCube< realT > &modes, improc::eigenImage< realT > &mask)
	Mask a basis set.

template<typename realT >
int	mx::sigproc::basisMeanSub (improc::eigenCube< realT > &modes, improc::eigenImage< realT > &mask, bool postMult=true)
	Mean-subtract a basis set.

template<typename realT >
int	mx::sigproc::basisNormalize (improc::eigenCube< realT > &modes, improc::eigenImage< realT > &mask)
	Normalize a basis set.

template<typename realT >
int	mx::sigproc::basisAmplitudes (std::vector< realT > &amps, improc::eigenImage< realT > &im, improc::eigenCube< realT > &modes, improc::eigenImage< realT > &mask, bool subtract=false, int meanIgnore=0, int N=-1)
	Measure the amplitudes of a set of basis modes fit to an image. Optionally subtract them.

template<int progress = 0, typename eigenTout , typename eigenTin >
void	mx::sigproc::gramSchmidt (eigenTout &out, const eigenTin &in, bool normalize=true)
	Perform Gram-Schmidt ortogonalization of a basis set, and normalize the result.

template<int progress = 0, typename eigenTout , typename eigenTin , typename eigenTWin >
void	mx::sigproc::gramSchmidt (eigenTout &out, const eigenTin &in, const eigenTWin &window)
	Perform Gram-Schmidt ortogonalization of a basis set on a window, and normalize the result.

template<int progress = 0, typename eigenTout , typename eigenTout2 , typename eigenTin >
void	mx::sigproc::gramSchmidtSpectrum (eigenTout &out, eigenTout2 &spect, const eigenTin &in, typename eigenTin::Scalar normPix=0.0)
	Perform Gram-Schmidt ortogonalization of a basis set, and normalize the result, while recording the spectrum.

Function Documentation

◆ autocorrelation() [1/2]

template<typename T >

void mx::sigproc::autocorrelation	(	std::vector< T > &	ac,
		std::vector< T > &	sig
	)

Calculate the autocorrelation of a time-series.

Template Parameters

T	is the real type of the data and autocorrelation.

Todo:: this should probably re-allocate if ac.size() != sig.size(), or at least <

Parameters

[out]	ac	will contain the autocorrelation on return. If ac.size()==0 then it is resized to sig.size().
[in]	sig	is the input time-series (signal)

Definition at line 81 of file autocorrelation.hpp.

References mx::sigproc::autocorrelation().

◆ autocorrelation() [2/2]

template<typename T >

void mx::sigproc::autocorrelation	(	T *	ac,
		size_t	Nac,
		T *	sig,
		size_t	Nsig
	)

Calculate the autocorrelation of a time-series.

Template Parameters

T	is the real type of the data and autocorrelation.

Parameters

[out]	ac	is the pre-allocated array of length Nac which will contain the autocorrelation on return
[in]	Nac	is the length of ac
[in]	sig	is the input time-series (signal)
[in]	Nsig	is the length of the input time-series

Definition at line 47 of file autocorrelation.hpp.

Referenced by mx::sigproc::autocorrelation().

◆ basisAmplitudes()

template<typename realT >

int mx::sigproc::basisAmplitudes	(	std::vector< realT > &	amps,
		improc::eigenImage< realT > &	im,
		improc::eigenCube< realT > &	modes,
		improc::eigenImage< realT > &	mask,
		bool	subtract = `false`,
		int	meanIgnore = `0`,
		int	N = `-1`
	)

Measure the amplitudes of a set of basis modes fit to an image. Optionally subtract them.

Mode subtraction occurs one by one, so subtraction will work with non-orthogonal basis sets.

Returns: 0 on success; -1 on error

Template Parameters

realT the floating point type.

Parameters

[out]	amps	the amplitudes of each mode fit to the image (will be resized).
	im	[in.out] the image to fit. Is subtracted in place if desired.
[in]	modes	the modes to fit.
[in]	mask	the 1/0 mask which defines the domain of the fit.
[in]	subtract	[optional] if true then the modes are subtracted as they are fit to the image
[in]	meanIgnore	[optional] if 1 then the mean, or if 2 the median, value is subtracted before fitting. If subtract is false, this value is added back after the subtraction.
[in]	N	[optional] the number of modes to actually fit. If N < 0 then all modes are fit.

Definition at line 150 of file basisUtils2D.hpp.

References mx::improc::imageMedian().

◆ basisMask()

template<typename realT >

int mx::sigproc::basisMask	(	improc::eigenCube< realT > &	modes,
		improc::eigenImage< realT > &	mask
	)

Mask a basis set.

Multiplies each mode in a basis set by a mask.

Returns: 0 on success; -1 on error

Template Parameters

realT a real floating point type.

Parameters

	modes	[in.out] the basis to normalize.
[in]	mask	1/0 mask defining the domain of the basis

Definition at line 51 of file basisUtils2D.hpp.

◆ basisMeanSub()

template<typename realT >

int mx::sigproc::basisMeanSub	(	improc::eigenCube< realT > &	modes,
		improc::eigenImage< realT > &	mask,
		bool	postMult = `true`
	)

Mean-subtract a basis set.

Subtracts the mean of each mode calculated over a domain defined by a mask.

Returns: 0 on success; -1 on error

Template Parameters

realT a real floating point type.

Parameters

	modes	[in.out] the basis to normalize.
[in]	mask	1/0 mask defining the domain of the basis
[in]	postMult	[optional] if true, then each image is multiplied by the mask after subtraction.

Definition at line 74 of file basisUtils2D.hpp.

◆ basisNormalize()

template<typename realT >

int mx::sigproc::basisNormalize	(	improc::eigenCube< realT > &	modes,
		improc::eigenImage< realT > &	mask
	)

Normalize a basis set.

RMS normalize each mode over a domain defined by a mask.

Returns: 0 on success; -1 on error

Template Parameters

realT a real floating point type.

Parameters

	modes	[in.out] the basis to normalize.
[in]	mask	1/0 mask defining the domain of the normalization

Definition at line 105 of file basisUtils2D.hpp.

◆ gramSchmidt() [1/2]

template<int progress = 0, typename eigenTout , typename eigenTin >

void mx::sigproc::gramSchmidt	(	eigenTout &	out,
		const eigenTin &	in,
		bool	normalize = `true`
	)

Perform Gram-Schmidt ortogonalization of a basis set, and normalize the result.

Performs the stabilized Gram-Schmidt procedure on the input basis set, which is the columns of the array. Optionally the output is normalized.

Template Parameters

progress	if true, then the loop index is printed for progress reporting
eigenTout	is the Eigen array type of the desired output
eigenTin	is the Eigen array type of the input

Parameters

[out]	out	the orthonormal basis set constructed from the input
[in]	in	a basis set, where each column represents one vector.
[in]	normalize	[optional] whether or not to normalize the output

Definition at line 48 of file gramSchmidt.hpp.

◆ gramSchmidt() [2/2]

template<int progress = 0, typename eigenTout , typename eigenTin , typename eigenTWin >

void mx::sigproc::gramSchmidt	(	eigenTout &	out,
		const eigenTin &	in,
		const eigenTWin &	window
	)

Perform Gram-Schmidt ortogonalization of a basis set on a window, and normalize the result.

Performs the stabilized Gram-Schmidt procedure on the input basis set over a window (or weight function), followed by normalization of the result.

Parameters

out	[out] is the orthonormal basis set constructed from the input
in	[in] is a basis set, where each column represents one vector.
window	[in] is the window, or weighting function

Template Parameters

progress	if true, then the loop index is printed for progress reporting
eigenTout	is the Eigen array type of the desired output
eigenTin	is the Eigen array type of the input
eigenTWin	is the Eigen array type of the window

Definition at line 103 of file gramSchmidt.hpp.

◆ gramSchmidtSpectrum()

template<int progress = 0, typename eigenTout , typename eigenTout2 , typename eigenTin >

void mx::sigproc::gramSchmidtSpectrum	(	eigenTout &	out,
		eigenTout2 &	spect,
		const eigenTin &	in,
		typename eigenTin::Scalar	normPix = `0.0`
	)

Perform Gram-Schmidt ortogonalization of a basis set, and normalize the result, while recording the spectrum.

Performs the stabilized Gram-Schmidt procedure on the input basis set, followed by normalization of the result. Also records the spectrum, that is the coefficients of the linear expansion in the orginal basis set for the resultant basis set.

Template Parameters

progress	if true, then the loop index is printed for progress reporting
eigenTout	is the Eigen array type of the output orthogonalized array
eigenTout2	is the Eigen array type of the spectrum
eigenTin	is the Eigen array type of the input

Parameters

[out]	out	the orthonormal basis set constructed from the input
[out]	spect	the spectrum
[in]	in	a basis set, where each column represents one vector
[in]	normPix	[optional] area of (usually number of pixels in) the orthogonal region for normalization. If 0 the basis is not renormalized

Definition at line 170 of file gramSchmidt.hpp.

Classes

Modules

Functions

Function Documentation

◆ autocorrelation() [1/2]

◆ autocorrelation() [2/2]

◆ basisAmplitudes()

◆ basisMask()

◆ basisMeanSub()

◆ basisNormalize()

◆ gramSchmidt() [1/2]

◆ gramSchmidt() [2/2]

◆ gramSchmidtSpectrum()