Collaboration diagram for Basic building blocks:

Modules
	Generic fill/copy functionality for STIR object from/to iterators

	Implementation details for buildblock

	Items relating to vectors and (multi-dimensional) arrays

	Items relating to coordinates

	Items related to simple geometric calculations

	Items related to ancillary information such as radionuclide, patient info etc

	Items related to projection data

	Items related to image data

	Data processors

	Items related to date/time processing

	Items relating to threading of certain STIR functions.

	Items relating to CUDA and STIR buildblock objects.

Files
file	ArrayFilter2DUsingConvolution.cxx
	Implementations for class ArrayFilter2DUsingConvolution.

file	ArrayFilter3DUsingConvolution.cxx
	Implementations for class ArrayFilter3DUsingConvolution.

file	buildblock_registries.cxx
	File that registers all stir::RegisterObject children in buildblock.

file	ByteOrder.cxx
	This file initialises ByteOrder::native_order.

file	DetectorCoordinateMap.cxx
	Implementation of class stir::DetectorCoordinateMap.

file	ExamData.cxx
	implementation of stir::ExamData

file	ExamInfo.cxx
	implementation of stir::ExamInfo

file	GatedDiscretisedDensity.cxx
	Implementation of class stir::GatedDiscretisedDensity.

file	GeneralisedPoissonNoiseGenerator.cxx
	Implements stir::GeneralisedPoissonNoiseGenerator.

file	interfile_keyword_functions.cxx
	Implementations for stir::standardise_interfile_keyword.

file	KeyParser.cxx
	Implementations for class stir::KeyParser.

file	line.cxx
	Implementations for class stir::Line.

file	ML_norm.cxx
	utilities for finding normalisation factors using an ML approach

file	num_threads.cxx
	Implementation of functions related to setting/getting the number of threads.

file	overlap_interpolate.cxx
	Implementation of stir::overlap_interpolate.

file	ParsingObject.cxx
	Non-linline implementations for class stir::ParsingObject.

file	PatientPosition.cxx
	Implementations of class stir::PatientPosition.

file	recon_array_functions.cxx
	implementations for functions declared in recon_array_functions.h

file	Scanner.cxx
	Implementations for class stir::Scanner.

file	TimeFrameDefinitions.cxx
	Implementation of class stir::TimeFrameDefinitions.

file	TimeGateDefinitions.cxx
	Implementation of class stir::TimeGateDefinitions.

file	Verbosity.cxx
	Declaration of class stir::Verbosity.

file	zoom.cxx
	Implementations of the stir::zoom functions.

file	AbsTimeIntervalFromDynamicData.cxx
	Declaration of class stir::AbsTimeIntervalFromDynamicData.

file	AbsTimeIntervalWithParsing.cxx
	Declaration of class stir::AbsTimeIntervalWithParsing.

file	DAVArrayFilter3D.cxx

file	ModifiedInverseAveragingImageFilterAll.cxx
	Implementations for class ModifiedInverseAveragingImageFilterAll.

file	ModifiedInverseAverigingImageFilter.cxx
	Implementations for class ModifiedInverseAverigingImageFilter.

file	NonseparableSpatiallyVaryingFilters.cxx
	Implementations for class NonseparableSpatiallyVaryingFilters.

file	NonseparableSpatiallyVaryingFilters3D.cxx
	Implementations for class NonseparableSpatiallyVaryingFilters3D.

file	Quaternion.cxx
	Implementation of class Quaternion.

file	DataSymmetriesForDensels_PET_CartesianGrid.cxx
	non-inline implementations for class stir::DataSymmetriesForDensels_PET_CartesianGrid

file	ArrayFilter2DUsingConvolution.h
	Declaration of class ArrayFilter2DUsingConvolution.

file	ArrayFilter3DUsingConvolution.h
	Declaration of class ArrayFilter3DUsingConvolution.

file	assign.h
	defines the stir::assign function to assign values to different data types

file	ByteOrder.h
	This file declares the stir::ByteOrder class.

file	ByteOrderDefine.h
	Definition of STIRIsNativeByteOrderBigEndian and STIRIsNativeByteOrderLittleEndian preprocessor symbols.

file	common.h
	basic configuration include file

file	gcc.h
	configuration for gcc

file	visualc.h
	configuration for Visual C++

file	copy_fill.h
	Declaration of stir::copy_to and stir::fill_from templates.

file	CPUTimer.h
	declares the stir::CPUTimer class.

file	CPUTimer.inl
	inline implementations for stir::CPUTimer

file	cross_product.h
	defines the cross-product of 2 CartesianCoordinate3D numbers

file	decay_correction_factor.h
	Simple functions to compute the decay correction factor.

file	deprecated.h
	This file declares a deprecation macro.

file	DetectorCoordinateMap.h
	Declaration of class stir::DetectorCoordinateMap.

file	error.h
	Declaration of stir::error()

file	ExamData.h
	declaration of stir::ExamData

file	ExamInfo.h
	This file declares the class stir::ExamInfo.

file	ExamInfo.inl
	This file declares the class stir::ExamInfo.

file	extract_line.h
	Declaration of stir::extract_line.

file	extract_line.inl
	Implementation of stir::extract_line.

file	FactoryRegistry.h
	Declaration of class stir::FactoryRegistry.

file	FactoryRegistry.inl
	Inline implementations for stir::FactoryRegistry.

file	FilePath.h
	Declaration of class stir::FilePath This is a class implementing basic filesytem functionality. Parts of this class were copied from Functions for filename manipulations.

file	FilePath.inl
	Implementations of inline functions for class stir::FilePath.

file	GeneralisedPoissonNoiseGenerator.h
	Declares stir::GeneralisedPoissonNoiseGenerator.

file	HigherPrecision.h
	class stir::HigherPrecision

file	ImagingModality.h
	Declaration of class stir::ImagingModality.

file	info.h
	Declaration of stir::info()

file	interfile_keyword_functions.h
	Functions useful for manipulating Interfile keywords.

file	interpolate.h
	declares functions for interpolation

file	is_null_ptr.h
	Definition of stir::is_null_ptr functions.

file	KeyParser.h
	Declaration of class stir::KeyParser.

file	line.h
	declaration of class stir::Line

file	linear_regression.h
	Declaration of stir::linear_regression()

file	linear_regression.inl
	Implementation of inline functions for stir::linear_regression()

file	min_positive_element.h
	Declares the stir::min_positive_element() function.

file	ML_norm.h
	Preliminary things for ML normalisation factor estimation.

file	modulo.h
	defines stir::modulo() and related functions

file	more_algorithms.h
	Declaration of some functions missing from std::algorithm.

file	more_algorithms.inl
	Implementation of some functions missing from std::algorithm.

file	NestedIterator.h
	This file declares the stir::NestedIterator class and supporting function objects.

file	NestedIterator.inl
	inline implementations for stir::NestedIterator.

file	NestedIteratorHelpers.h
	This file defines supporting function objects for stir::NestedIterator.

file	NumericInfo.h
	This file declares the class stir::NumericInfo.

file	NumericType.h
	This file declares the stir::NumericType class.

file	NumericType.inl
	Implementation of inline methods of class stir::NumericType.

file	ParsingObject.h
	Declaration of class stir::ParsingObject.

file	PatientPosition.h
	Declaration of class stir::PatientPosition.

file	recon_array_functions.h
	a variety of useful functions

file	DataSymmetriesForBins.inl
	inline implementations for class stir::DataSymmetriesForBins

file	DataSymmetriesForDensels.inl
	inline implementations for class stir::DataSymmetriesForDensels

file	RegisteredObject.h
	Declaration of class stiir::RegisteredObject.

file	RegisteredObject.inl
	Inline implementations for class stir::RegisteredObject.

file	RegisteredParsingObject.h
	Declaration of class stir::RegisteredParsingObject.

file	RegisteredParsingObject.inl
	Inline implementations for class stir::RegisteredParsingObject.

file	round.h
	Declaration of the stir::round functions.

file	round.inl
	Implementation of the stir::round functions.

file	RunTests.h
	defines the stir::RunTests class

file	Scanner.h
	Declaration of class stir::Scanner.

file	Scanner.inl
	implementation of inline functions of class Scanner

file	CreateTailMaskFromACFs.h
	Compute a mask for the "tails" in the sinogram.

file	SeparableGaussianImageFilter.h
	Declaration of class stir::SeparableGaussianImageFilter.

file	shared_ptr.h
	Import of std::shared_ptr, std::dynamic_pointer_cast and std::static_pointer_cast (or corresponding boost versions if STIR_USE_BOOST_SHARED_PTR is set, i.e. normally when std::shared_ptr doesn't exist) into the stir namespace.

file	stream.h
	Input/output of basic vector-like types to/from streams.

file	stream.inl
	Input/output of basic vector-like types to/from streams.

file	Succeeded.h
	Declaration of class stir::Succeeded.

file	thresholding.h
	Declaration of functions that threshold sequences (specified by iterators).

file	TimedBlock.h
	Class stir::TimedBlock.

file	TimedObject.h
	declares the stir::TimedObject class

file	TimedObject.inl
	inline implementations for stir::TimedObject

file	TimeFrameDefinitions.h
	Declaration of class stir::TimeFrameDefinitions.

file	TimeGateDefinitions.h
	Declaration of class stir::TimeGateDefinitions.

file	Timer.h
	This declares the stir::Timer class.

file	Timer.inl
	inline implementations for stir::Timer

file	TOF_conversions.h
	Implementations of inline functions for TOF time to mm.

file	unique_ptr.h
	Import of std::unique_ptr into the stir namespace, together with work-arounds for other compilers.

file	utilities.h
	This file declares various utility functions.

file	utilities.inl
	inline implementations for utility.h

file	Verbosity.h
	Declaration of class stir::Verbosity.

file	warning.h
	Declaration of stir::warning()

file	zoom.h
	This file declares various zooming functions.

file	ZoomOptions.h
	Declaration of class stir::ZoomOptions.

file	AbsTimeInterval.h
	Declaration of class stir::AbsTimeInterval.

file	AbsTimeIntervalFromDynamicData.h
	Declaration of class stir::AbsTimeIntervalFromDynamicData.

file	AbsTimeIntervalFromECAT7ACF.h
	Declaration of class stir::AbsTimeIntervalFromECAT7ACF.

file	AbsTimeIntervalWithParsing.h
	Declaration of class stir::AbsTimeIntervalWithParsing.

file	ModifiedInverseAveragingImageFilterAll.h
	This is a messy, first, attempt to design spatially varying filter Given the kernel which in this case is a lospass filter with a DC gain 1 the filter is design such that the output kernel varies depending on the k0 and k1 ( for more details on these factors look at Fessler)

file	ModifiedInverseAverigingArrayFilter.h
	This is a messy first attempt to design spatially varying filter Given the kernel which in this case is a lospass filter with a DC gain 1 the filter is design such that the output kernel varies depending on the k0 and k1 ( for more details on these factors look at Fessler)

file	ModifiedInverseAverigingImageFilter.h
	This is a messy, first, attempt to design spatially varying filter Given the kernel which in this case is a lospass filter with a DC gain 1 the filter is design such that the output kernel varies depending on the k0 and k1 ( for more details on these factors look at Fessler)

file	NonseparableSpatiallyVaryingFilters.h
	This is a messy, first, attempt to design spatially varying filter Given the kernel which in this case is a lospass filter with a DC gain 1 the filter is design such that the output kernel varies depending on the k0 and k1 ( for more details on these factors look at Fessler)

file	NonseparableSpatiallyVaryingFilters3D.h
	This is a messy, first, attempt to design spatially varying filter Given the kernel which in this case is a lospass filter with a DC gain 1 the filter is design such that the output kernel varies depending on the k0 and k1 ( for more details on these factors look at Fessler)

file	Quaternion.h
	Declaration of class stir::Quaternion.

file	Quaternion.inl
	Implementation of class stir::Quaternion.

file	SeparableGaussianArrayFilter.h

file	SeparableGaussianImageFilter.h
	Declaration of class SeparableGaussianImageFilter.

file	SeparableLowPassArrayFilter.h

file	SeparableLowPassImageFilter.h
	Declaration of class SeparableLowPassImageFilter.

file	InvertAxis.cxx
	Implementation of function stir::invert_axis.

file	warp_image.cxx
	Implementation of function stir::warp_image.

file	test_GeneralisedPoissonNoiseGenerator.cxx
	tests for the stir::GeneralisedPoissonNoiseGenerator class

Classes
class	stir::CPUTimer
	A class for measuring elapsed CPU time. More...

class	stir::DynamicDiscretisedDensity
	Class of multiple image frames, one for each time frame Each time frame is a DiscretisedDensity<3,float> More...

class	stir::DynamicProjData
	Dynamic projection data. More...

class	stir::ExamData
	base class for data objects such as ProjData etcProvides an ExamInfo member. More...

class	stir::ExamInfo
	a class for storing information about 1 exam (or scan) More...

class	stir::FactoryRegistry< Key, Factory, Compare >
	This class can be used to store 'factories' and their corresponding keys. It is essentially a map, but with some extra embelishments. More...

class	stir::FilePath
	The FilePath class. More...

class	stir::GeneralisedPoissonNoiseGenerator
	Generates noise realisations according to Poisson statistics but allowing for scaling. More...

struct	stir::HigherPrecision< T >
	Helper class to get a type with higher precision. More...

class	stir::ImagingModality
	Class for encoding the modality. More...

class	stir::KeyArgument
	A class that enumerates the possible types that can be used to store parsed values. Used (only) by KeyParser. More...

class	stir::map_element
	Class to store the Interfile keywords and their actions. More...

class	stir::KeyParser
	A class to parse Interfile headers. More...

class	stir::NestedIterator< topleveliterT, GetRestRangeFunctionT >
	Class NestedIterator implements a (forward) iterator using a pair of 'nested' iterators. More...

class	stir::NumericInfo< NUMBER >
	class NumericInfo<NUMBER> defines properties for the type NUMBER. More...

class	stir::NumericType
	provides names for some numeric types and methods for finding their properties. More...

struct	stir::TypeForNumericType< numeric_type_enum >
	A helper class that specifies the C++ type for a particular NumericType. More...

class	stir::ParseAndCreateFrom< OutputT, InputT, ParserT >
	template for adding keywords to a parser and creating an object More...

class	stir::ParsingObject
	A base class for objects that want to be able to parse parameter files. More...

class	stir::PatientPosition

class	stir::SymmetryOperation
	Encodes symmetry operation on image coordinates and projection data coordinates. More...

class	stir::RegisteredObject< Root >
	Helper class to provide registry mechanisms to a `Base` classSuppose you have a hierarchy of classes with (nearly) all public functionality provided by virtual functions of the `Base` (here called `Root`) class. The aim is then to be able to select at run-time which of the nodes will be used. More...

class	stir::RegisteredObjectBase
	Base class for all classes that can parse .par files (and more?)The only reason that this class exists is such that KeyParser can store different types of objects, and get some basic info from it. More...

class	stir::RegisteredParsingObject< Derived, Base, Parent >
	Parent class for all leaves in a RegisteredObject hierarchy that do parsing of parameter files. More...

class	stir::RunTests
	A base class for making test classesWith a derived class, an application could look like. More...

class	stir::Scanner
	A class for storing some info on the scanner. More...

class	stir::TimedBlock< TimerT=Timer >
	Helper class for measuring execution time of a block of code.It starts the timer in ctor, stops in dtor. Do not create unnamed instances of this class, as they are quite useless: you cannot predict destruction time. More...

class	stir::TimedObject
	base class for all objects which need timers. At the moment, there's only a CPU timer. More...

class	stir::TimeFrameDefinitions
	Class used for storing time frame durations. More...

class	stir::TimeGateDefinitions
	Class used for storing time gate durations. More...

class	stir::Timer
	A general base class for timers. Interface is like a stop watch. More...

class	stir::Verbosity
	This class enables the user to control the on-screen output. More...

class	stir::ZoomOptions
	This class enables the user to choose between different zooming optionsThe 3 possible values determine a global scale factor used for the end result: (i) preserve sum (locally) (ii) preserve values (like interpolation) (iii) preserve projections: using a STIR forward projector on the zoomed image will give (approximately) the same projections. More...

class	stir::AbsTimeInterval
	Base class for specifying a time interval (in absolute time) More...

class	stir::AbsTimeIntervalFromDynamicData
	class for specifying a time interval via a dynamic scan More...

class	stir::AbsTimeIntervalFromECAT7ACF
	class for specifying a time interval via an ECAT7 .a file More...

class	stir::AbsTimeIntervalWithParsing
	class for specifying a time interval via parsing of explicit times More...

class	stir::SeparableArrayFunctionObject2< num_dimensions, elemT >
	This class implements an `n` -dimensional ArrayFunctionObject whose operation is separable. More...

class	stir::ImagingModalityTests
	Test class for ImagingModality. More...

class	stir::ByteOrder
	This class provides member functions to find out what byte-order your machine is and to swap numbers. More...

Macros
#define	STIRIsNativeByteOrderBigEndian 1
	A macro that is defined to 1 when the compilation is on a big endian machine, otherwise it is set to 0. More...

#define	STIRIsNativeByteOrderLittleEndian 0
	A macro that is defined to 1 when the compilation is on a little endian machine, and to 0 otherwise. More...

#define	_PI boost::math::constants::pi<double>()
	The constant pi to high precision. More...

Functions
template<class NUMBER >
NUMBER	stir::square (const NUMBER &x)
	returns the square of a number, templated. More...

template<class coordT >
CartesianCoordinate3D< coordT >	stir::cross_product (const CartesianCoordinate3D< coordT > &a, const CartesianCoordinate3D< coordT > &b)
	the cross-product for 3-dimensional coordinates. More...

double	stir::decay_correction_factor (const double isotope_halflife, const double start_time, const double end_time)
	Compute decay-correction factor for a time frame. More...

double	stir::decay_correction_factor (const double isotope_halflife, const double rel_time)
	Computes the decay-correction factor for activity at a given time point. More...

void	stir::error (const char *const s,...)
	Print error with format string a la `printf` and throw exception. More...

template<class STRING >
void	stir::error (const STRING &string)
	Use this function for writing error messages and throwing an exception. More...

template<class elemT >
Array< 1, elemT >	stir::extract_line (const Array< 3, elemT > &, const BasicCoordinate< 3, int > &index, const int dimension)
	extracts a line from an array in the direction of the specified dimension. More...

template<class STRING >
void	stir::info (const STRING &string, const int verbosity_level=1)
	Use this function for writing informational messages. More...

template<class Value , class DataType , class CoordinatesType >
void	stir::linear_regression (Value &constant, Value &scale, Value &chi_square, Value &variance_of_constant, Value &variance_of_scale, Value &covariance_of_constant_with_scale, const VectorWithOffset< DataType > &measured_data, const VectorWithOffset< CoordinatesType > &coordinates, const VectorWithOffset< float > &weights, const bool use_estimated_variance=true)
	Implements standard linear regression on VectorWithOffset data. More...

template<class Value , class DataIter , class CoordinatesIter , class WeightsIter >
void	stir::linear_regression (Value &constant, Value &scale, Value &chi_square, Value &variance_of_constant, Value &variance_of_scale, Value &covariance_of_constant_with_scale, DataIter measured_data_begin, DataIter measured_data_end, CoordinatesIter coords_begin, WeightsIter weights_begin, const bool use_estimated_variance=true)
	Implements standard linear regression. More...

template<typename ForwardIter_t >
ForwardIter_t	stir::min_positive_element (ForwardIter_t start, ForwardIter_t end)
	Finds where the smallest strictly positive element occurs. More...

template<class iterT >
iterT	stir::abs_max_element (iterT start, iterT end)
	Like std::max_element, but comparing after taking absolute value. More...

template<class IterT , class elemT >
elemT	stir::sum (IterT start, IterT end, elemT init)
	Compute the sum of a sequence using operator+=(), using an initial value. More...

template<class IterT >
std::iterator_traits< IterT >::value_type	stir::sum (IterT start, IterT end)
	Compute the sum of a sequence using operator+=(). More...

template<class IterT >
std::iterator_traits< IterT >::value_type	stir::average (IterT start, IterT end)
	Compute the average of a sequence using sum(start,end). More...

char *	stir::strupr (char *const str)
	make C-string uppercase

void	stir::warning (const char *const s,...)
	Print warning with format string a la `printf`. More...

template<class STRING >
void	stir::warning (const STRING &string, const int verbosity_level=1)
	Use this function for writing warning messages. More...

Variables
const int	stir::max_filename_length = 1000
	some large value to say how long filenames can be in the (deprecated) function ask_filename_with_extension(char *,const std::string&, const std::string&)

Detailed Description

Library with things that are not specific to reconstructions. This includes multi-dimensional arrays, images, image processors, projection data,...

Macro Definition Documentation

◆ STIRIsNativeByteOrderBigEndian

#define STIRIsNativeByteOrderBigEndian 1

A macro that is defined to 1 when the compilation is on a big endian machine, otherwise it is set to 0.

Usage: #if STIRIsNativeByteOrderBigEndian
// code specific for big endian machines
#else
// code for little endian machines
#endif

Relation to the ByteOrder class.

This should be used only in the case that you really need to know this at compilation time. Try to use class ByteOrder instead.

The definition of class ByteOrder is independent of this macro.

Warning: It is recommended to check these settings by comparing with class ByteOrder. This is done by the ByteOrderTests class.; This macro depends on preprocessor defines set by your compiler. The list of defines is bound to be incomplete though. The current list sets LittleEndian for alpha, x86 or PowerPC processors or Windows using some macros which are known to work with gcc and Visual Studio (at some point also with CodeWarrior). Otherwise, the architecture is assumed to be big endian. ByteOrder does not rely on this type of conditionals.; The value of the macro in the doxygen documentation depends on what type of machine doxygen was run.

◆ STIRIsNativeByteOrderLittleEndian

#define STIRIsNativeByteOrderLittleEndian 0

A macro that is defined to 1 when the compilation is on a little endian machine, and to 0 otherwise.

See also: STIRIsNativeByteOrderBigEndian for details.

◆ _PI

#define _PI boost::math::constants::pi<double>()

The constant pi to high precision.

Referenced by stir::Scanner::check_consistency(), stir::CListModeDataROOT::CListModeDataROOT(), stir::ScatterSimulation::detection_efficiency_no_scatter(), stir::ScatterSimulation::downsample_scanner(), stir::extend_segment(), stir::ProjDataInfoCylindricalNoArcCorr::find_scanner_coordinates_given_cartesian_coordinates(), stir::from_min_pi_plus_pi_to_0_2pi(), stir::ProjDataInfoCylindricalNoArcCorr::get_bin(), stir::ProjMatrixByDenselUsingRayTracing::get_element(), stir::Ellipsoid::get_geometric_volume(), stir::EllipsoidalCylinder::get_geometric_volume(), stir::ProjDataInfoCylindrical::get_LOR(), stir::inverse_fourier_1d_for_real_data_corrupting_input(), stir::EllipsoidalCylinder::is_inside_shape(), stir::LORInCylinderCoordinates< coordT >::is_swapped(), stir::LORAs2Points< float >::is_swapped(), stir::LORInAxialAndSinogramCoordinates< coordT >::is_swapped(), stir::Scanner::list_names(), stir::LORInAxialAndNoArcCorrSinogramCoordinates< coordT >::LORInAxialAndNoArcCorrSinogramCoordinates(), stir::LORInAxialAndSinogramCoordinates< coordT >::LORInAxialAndSinogramCoordinates(), stir::ProjDataInfoCylindrical::parameter_info(), stir::Scanner::parameter_info(), stir::InterfilePDFSHeaderSPECT::post_processing(), stir::Shape3DWithOrientation::post_processing(), stir::ProjDataInfoCylindrical::ProjDataInfoCylindrical(), stir::FourierRebinning::rebin(), stir::ProjDataInfoTests::run_coordinate_test(), stir::ProjDataInfoTests::run_coordinate_test_for_realistic_scanner(), stir::ProjDataInfoTests::run_lor_get_s_test(), stir::ScannerTests::run_tests(), stir::ProjDataInfoCylindricalArcCorrTests::run_tests(), stir::Scanner::Scanner(), stir::SeparableGaussianArrayFilter< num_dimensions, elemT >::SeparableGaussianArrayFilter(), stir::Shape3DWithOrientation::set_key_values(), stir::ColsherFilter::set_up(), stir::ProjMatrixByBinSPECTUB::set_up(), stir::erfTests::test_FastErf(), stir::to_0_2pi(), and stir::BackProjectorByBinUsingInterpolation::use_piecewise_linear_interpolation().

Function Documentation

◆ square()

template<class NUMBER >

NUMBER stir::square ( const NUMBER & x )

inline

returns the square of a number, templated.

Referenced by stir::absolute_max_eigenvector_using_power_method(), stir::LogcoshPrior< elemT >::accumulate_Hessian_times_input(), stir::ProjMatrixByDenselOnCartesianGridUsingElement::calculate_proj_matrix_elems_for_one_densel(), stir::LORAs2Points< float >::change_representation_for_block(), stir::coordinate_between_2_lines(), stir::distance_between_line_and_point(), stir::ScatterSimulation::energy_lower_limit(), stir::find_LOR_intersections_with_cylinder(), stir::ProjDataInfoCylindricalNoArcCorr::find_scanner_coordinates_given_cartesian_coordinates(), stir::ProjDataInfo::get_costheta(), stir::ProjMatrixByDenselUsingRayTracing::get_element(), stir::ProjDataInfoGeneric::get_LOR(), stir::ProjDataInfoCylindrical::get_LOR(), stir::RigidObject3DTransformation::get_relative_transformation(), stir::ProjDataInfoCylindrical::get_tantheta(), stir::EllipsoidalCylinder::is_inside_shape(), stir::make_det_pair_data(), stir::ScatterSimulation::max_cos_angle(), stir::randoms_from_singles(), stir::RigidObject3DTransformation::RigidObject3DTransformation(), stir::MatchTrackerAndScanner::run(), stir::QuaternionTests::run_tests(), stir::Shape3DWithOrientation::scale(), stir::SeparableGaussianArrayFilter< num_dimensions, elemT >::SeparableGaussianArrayFilter(), stir::PLSPrior< elemT >::set_anatomical_filename(), stir::RelativeDifferencePrior< elemT >::set_kappa_sptr(), stir::LogcoshPrior< elemT >::set_scalar(), stir::KOSMAPOSLReconstruction< TargetT >::set_sigma_m(), stir::ProjMatrixElemsForOneDensel::square_sum(), stir::ProjMatrixElemsForOneBin::square_sum(), stir::truncate_rim(), and stir::RelativeDifferencePrior< elemT >::value().

◆ cross_product()

template<class coordT >

CartesianCoordinate3D<coordT> stir::cross_product	(	const CartesianCoordinate3D< coordT > &	a,
		const CartesianCoordinate3D< coordT > &	b
	)

the cross-product for 3-dimensional coordinates.

Warning: This implements minus the 'usual' definition of the cross-product. This is done because STIR uses a left-handed coordinate system. The definition of cross_product is such that ${a, b, a\times b}$ forms a left-handed coordinate system.

◆ decay_correction_factor() [1/2]

double stir::decay_correction_factor	(	const double	isotope_halflife,
		const double	start_time,
		const double	end_time
	)

inline

Compute decay-correction factor for a time frame.

This function computes the factor needed to convert average number of counts per second to activity at time 0, i.e. it returns

$\frac{(t_2-t_1)}{ \int_{t_1}^{t_2} \! 2^{-t/\mathrm{halflife}} \, dt}$

Referenced by stir::DynamicDiscretisedDensity::decay_correct_frames(), stir::PlasmaData::decay_correct_PlasmaData(), stir::randoms_from_singles(), and stir::PatlakPlot::set_model_matrix().

◆ decay_correction_factor() [2/2]

double stir::decay_correction_factor	(	const double	isotope_halflife,
		const double	rel_time
	)

inline

Computes the decay-correction factor for activity at a given time point.

This function computes the correction factor to convert activity at t0 + rel_time to activity at t0, i.e.

$2^{(\mathrm{rel\_time} / \mathrm{halflife})}$

◆ error() [1/2]

void stir::error	(	const char *const	s,
			...
	)

Print error with format string a la printf and throw exception.

The arguments are the same as if you would call printf(). The error message is written to stderr, preceeded by "ERROR:", a std::string is constructed with the error message, and throw is called with the string as argument.

Note that because we throw an exception, the caller can catch it. Prior to STIR 2.1, this was not possible as stir::error directly called std::exit.

Todo:: As opposed to using printf-style calling sequence, use a syntax like stir::info

Example: error("Error opening file %s", filename);

Deprecated:: (use 1 argument version instead)

◆ error() [2/2]

template<class STRING >

void stir::error ( const STRING & string )

inline

Use this function for writing error messages and throwing an exception.

The argument is expected to be a string, but could be anything for which std::ostream::operator<< would work.

This function currently first writes a newline, then ERROR:, then string and then another newline to std::cerr. Then it throws an exception.

Todo:: At a later stage, it will also write to a log-file.

boost::format is useful in this context.

Example: error(boost::format("Incorrect number of subsets: %d") % num_subsets);
error("This does not work");

◆ extract_line()

template<class elemT >

Array< 1, elemT > stir::extract_line	(	const Array< 3, elemT > &	input_array,
		const BasicCoordinate< 3, int > &	index,
		const int	dimension
	)

inline

extracts a line from an array in the direction of the specified dimension.

Todo:: make n-dimensional version

References stir::VectorWithOffset< T >::get_max_index(), and stir::VectorWithOffset< T >::get_min_index().

Referenced by stir::interpolate_line().

◆ info()

template<class STRING >

void stir::info	(	const STRING &	string,
		const int	verbosity_level = `1`
	)

Use this function for writing informational messages.

The argument is expected to be a string, but could be anything for which std::ostream::operator<< would work.

This function currently first writes a newline, then INFO, then string and then another newline to std::cerr.

Todo:: At a later stage, it will also write to a log-file.

boost::format is useful in this context.

Example: info(boost::format("Running sub-iteration %1% of total %2%") % subiter_num % total);
info("Running a really complicated algorithm");

References stir::get().

◆ linear_regression() [1/2]

template<class Value , class DataType , class CoordinatesType >

void stir::linear_regression	(	Value &	constant,
		Value &	scale,
		Value &	chi_square,
		Value &	variance_of_constant,
		Value &	variance_of_scale,
		Value &	covariance_of_constant_with_scale,
		const VectorWithOffset< DataType > &	measured_data,
		const VectorWithOffset< CoordinatesType > &	coordinates,
		const VectorWithOffset< float > &	weights,
		const bool	use_estimated_variance = `true`
	)

inline

Implements standard linear regression on VectorWithOffset data.

The linear_regression function does straightforward (1 dimensional) weighted least squares fitting.

input

3 VectorWithOffsets of measured_data, coordinates, weights
1 optional boolean value use_estimated_variance (default=true)

output

fitted parameters : constant, scale
goodness of fit measures : chi_square and the (co)variances

This solves the minimisation problem:

     Find constant, scale such that
       chi_square =
          sum_i weights[i]*
                (constant + scale*coordinates[i] - measured_data[i])^2
     is minimal.

When use_estimated_variance == false, the (co)variances are computed assuming that 1/weights[i] is the standard deviation on measured_data[i]. In particular, this means that the (co)variances depend only on the weights and the coordinates.

Alternatively, when use_estimated_variance == true, the weights are considered to be really only proportional to the 1/variance. Then the estimated variance is used to get sensible estimates of the errors:

     estimated_variance = chi_square/(measured_data.get_length() - 2)
     estimated_covariance_matrix = original_covariance_matrix * estimated_variance

References stir::VectorWithOffset< T >::begin(), stir::VectorWithOffset< T >::end(), stir::VectorWithOffset< T >::get_max_index(), and stir::VectorWithOffset< T >::get_min_index().

Referenced by stir::PatlakPlot::apply_linear_regression(), and stir::linear_regressionTests::run_tests().

◆ linear_regression() [2/2]

template<class Value , class DataIter , class CoordinatesIter , class WeightsIter >

void stir::linear_regression	(	Value &	constant,
		Value &	scale,
		Value &	chi_square,
		Value &	variance_of_constant,
		Value &	variance_of_scale,
		Value &	covariance_of_constant_with_scale,
		DataIter	measured_data_begin,
		DataIter	measured_data_end,
		CoordinatesIter	coords_begin,
		WeightsIter	weights_begin,
		const bool	use_estimated_variance = `true`
	)

inline

Implements standard linear regression.

This function takes the data as iterators for maximum flexibility. Note that it is assumed (but not checked) that the measured_data, coordinates and weights iterators run over the same range.

See also: linear_regression(Value& , Value&, Value& , Value& , Value& , Value& , const VectorWithOffset<DataType>& , const VectorWithOffset<CoordinatesType>& , const VectorWithOffset<float>& , const bool )

◆ min_positive_element()

template<typename ForwardIter_t >

ForwardIter_t stir::min_positive_element	(	ForwardIter_t	start,
		ForwardIter_t	end
	)

Finds where the smallest strictly positive element occurs.

Parameters

start	start of the sequence. Usually object.begin().
end	end of the sequence in iterator sense (so actually one beyond the last element). Usually object.end().

Returns: an iterator that points to the element in the sequence which has the smallest (strictly) positive value. If no (strictly) positive element is found, or if the sequence is empty, the end argument is returned.

The iterator type has to satisfy the requirements of a forward iterator, and its value_type has to be comparable using < and <=.

Referenced by stir::threshold_min_to_small_positive_value().

◆ abs_max_element()

template<class iterT >

iterT stir::abs_max_element	(	iterT	start,
		iterT	end
	)

inline

Like std::max_element, but comparing after taking absolute value.

This function using stir::norm_squared(), so works for complex numbers as well.

References stir::norm_squared().

Referenced by stir::absolute_max_eigenvector_using_power_method().

◆ sum() [1/2]

template<class IterT , class elemT >

elemT stir::sum	(	IterT	start,
		IterT	end,
		elemT	init
	)

inline

Compute the sum of a sequence using operator+=(), using an initial value.

Sadly std::accumulate uses operator+(). For non-trivial objects, this might call an inefficient version for the addition. As an alternative, std::accumulate could be called with a function object that calls operator+=.

Referenced by stir::average(), stir::ParametricQuadraticPrior< TargetT >::compute_value(), stir::ProjData::copy_to(), stir::find_centre_of_gravity(), stir::make_det_pair_data(), stir::Array< 1, elemT >::release_const_full_data_ptr(), stir::SeparableGaussianArrayFilter< num_dimensions, elemT >::SeparableGaussianArrayFilter(), stir::SeparableMetzArrayFilter< num_dimensions, elemT >::SeparableMetzArrayFilter(), stir::sum(), and stir::ModelMatrix< 2 >::write_to_file().

◆ sum() [2/2]

template<class IterT >

std::iterator_traits< IterT >::value_type stir::sum	(	IterT	start,
		IterT	end
	)

inline

Compute the sum of a sequence using operator+=().

Sadly std::accumulate uses operator+(). For non-trivial objects, this might call an inefficient version for the addition. Alternatively, std::accumulate could be called with a function object that calls operator+=. Still, in that case you need to specify an initial value.

If the range is empty, this function calls operator*=(0) to initialise the object to 0.

Warning: Currently, no work-around is present for old compilers that do not support std::iterator_traits.

References stir::sum().

◆ average()

template<class IterT >

std::iterator_traits< IterT >::value_type stir::average	(	IterT	start,
		IterT	end
	)

inline

Compute the average of a sequence using sum(start,end).

Warning: Currently, no work-around is present for old compilers that do not support std::iterator_traits.

References stir::sum().

Referenced by stir::RigidObject3DTransformation::find_closest_transformation().

◆ warning() [1/2]

void stir::warning	(	const char *const	s,
			...
	)

Print warning with format string a la printf.

The arguments are the same as if you would call printf(). The warning message is written to stderr, preceeded by "WARNING:".

Example: warning("Cannot open file %s, but I will work around it", filename);

Deprecated:: (use 1 argument version instead)

◆ warning() [2/2]

template<class STRING >

void stir::warning	(	const STRING &	string,
		const int	verbosity_level = `1`
	)

inline

Use this function for writing warning messages.

The argument is expected to be a string, but could be anything for which std::ostream::operator<< would work.

This function currently first writes a newline, then WARNING:, then string and then another newline to std::cerr.

Todo:: At a later stage, it will also write to a log-file.

boost::format is useful in this context.

Example: warning(boost::format("Type is like this: %1%. Not sure if that will work.") % projdata_info.parameter_info());
warning("This might not work");

References stir::get().

Modules

Files

Classes

Macros

Functions

Variables

Detailed Description

Macro Definition Documentation

◆ STIRIsNativeByteOrderBigEndian

◆ STIRIsNativeByteOrderLittleEndian

◆ _PI

Function Documentation

◆ square()

◆ cross_product()

◆ decay_correction_factor() [1/2]

◆ decay_correction_factor() [2/2]

◆ error() [1/2]

◆ error() [2/2]

◆ extract_line()

◆ info()

◆ linear_regression() [1/2]

◆ linear_regression() [2/2]

◆ min_positive_element()

◆ abs_max_element()

◆ sum() [1/2]

◆ sum() [2/2]

◆ average()

◆ warning() [1/2]

◆ warning() [2/2]