Spectral table model class. More...

#include <GModelSpectralTable.hpp>

Inheritance diagram for GModelSpectralTable:

Public Member Functions
	GModelSpectralTable (void)
	Void constructor. More...

	GModelSpectralTable (const GFilename &filename, const double &norm)
	File constructor. More...

	GModelSpectralTable (const GEbounds &ebounds, const GModelSpectralTablePars &pars, const GNdarray &spectra)
	Table model constructor. More...

	GModelSpectralTable (const GXmlElement &xml)
	XML constructor. More...

	GModelSpectralTable (const GModelSpectralTable &model)
	Copy constructor. More...

virtual	~GModelSpectralTable (void)
	Destructor. More...

virtual GModelSpectralTable &	operator= (const GModelSpectralTable &model)
	Assignment operator. More...

virtual void	clear (void)
	Clear table model. More...

virtual GModelSpectralTable *	clone (void) const
	Clone table model. More...

virtual std::string	classname (void) const
	Return class name. More...

virtual std::string	type (void) const
	Return model type. More...

virtual double	eval (const GEnergy &srcEng, const GTime &srcTime=GTime(), const bool &gradients=false) const
	Evaluate spectral table model. More...

virtual double	flux (const GEnergy &emin, const GEnergy &emax) const
	Returns model photon flux between emin, emax More...

virtual double	eflux (const GEnergy &emin, const GEnergy &emax) const
	Returns model energy flux between emin, emax More...

virtual GEnergy	mc (const GEnergy &emin, const GEnergy &emax, const GTime &time, GRan &ran) const
	Returns MC energy between [emin, emax]. More...

virtual void	read (const GXmlElement &xml)
	Read model from XML element. More...

virtual void	write (GXmlElement &xml) const
	Write model into XML element. More...

virtual std::string	print (const GChatter &chatter=NORMAL) const
	Print table model information. More...

int	nspectra (void) const
	Return number of spectra in table model. More...

GModelSpectralTablePar &	table_par (const int &index)
	Return reference to table parameter. More...

const GModelSpectralTablePar &	table_par (const int &index) const
	Return const reference to table parameter. More...

GModelSpectralTablePar &	table_par (const std::string &name)
	Return reference to table parameter. More...

const GModelSpectralTablePar &	table_par (const std::string &name) const
	Return const reference to table parameter. More...

double	norm (void) const
	Return normalization factor. More...

void	norm (const double &norm)
	Set normalization factor. More...

const GEbounds &	ebounds (void) const
	Return reference to energy boundaries. More...

void	load (const GFilename &filename)
	Load table from file. More...

void	save (const GFilename &filename, const bool &clobber=false) const
	Save table into file. More...

const GFilename &	filename (void) const
	Return file name. More...

void	energy_scale (const std::string &name)
	Set energy scale. More...

Public Member Functions inherited from GModelSpectral
	GModelSpectral (void)
	Void constructor. More...

	GModelSpectral (const GModelSpectral &model)
	Copy constructor. More...

virtual	~GModelSpectral (void)
	Destructor. More...

virtual GModelSpectral &	operator= (const GModelSpectral &model)
	Assignment operator. More...

virtual GModelPar &	operator[] (const int &index)
	Returns model parameter. More...

virtual const GModelPar &	operator[] (const int &index) const
	Returns model parameter (const version) More...

virtual GModelPar &	operator[] (const std::string &name)
	Returns reference to model parameter. More...

virtual const GModelPar &	operator[] (const std::string &name) const
	Returns reference to model parameter (const version) More...

GModelPar &	at (const int &index)
	Returns model parameter. More...

const GModelPar &	at (const int &index) const
	Returns model parameter (const version) More...

bool	has_par (const std::string &name) const
	Checks if parameter name exists. More...

int	size (void) const
	Return number of parameters. More...

void	autoscale (void)
	Autoscale parameters. More...

Public Member Functions inherited from GBase
virtual	~GBase (void)
	Destructor. More...

Protected Member Functions
void	init_members (void)
	Initialise class members. More...

void	copy_members (const GModelSpectralTable &model)
	Copy class members. More...

void	free_members (void)
	Delete class members. More...

void	set_par_pointers (void)
	Set parameter pointers. More...

void	set_energy_nodes (void)
	Set energy nodes from energy boundaries. More...

void	scale_energy (void)
	Scale energy. More...

bool	has_energy_scale (void) const
	Signal that energy scale was set. More...

GFitsBinTable	create_par_table (void) const
	Create PARAMETERS FITS table. More...

GFitsBinTable	create_eng_table (void) const
	Create ENERGIES FITS table. More...

GFitsBinTable	create_spec_table (void) const
	Create SPECTRA FITS table. More...

void	load_par (const GFits &fits)
	Load data from PARAMETERS extension. More...

void	load_eng (const GFits &fits)
	Load data from ENERGIES extension. More...

void	load_spec (const GFits &fits)
	Load data from SPECTRA extension. More...

int	par_index (const std::string &name) const
	Return index for parameter name. More...

void	update (void) const
	Update cache for spectral table model computation. More...

void	update_flux (void) const
	Update flux cache. More...

void	update_mc (const GEnergy &emin, const GEnergy &emax) const
	Update MC cache. More...

Protected Member Functions inherited from GModelSpectral
void	init_members (void)
	Initialise class members. More...

void	copy_members (const GModelSpectral &model)
	Copy class members. More...

void	free_members (void)
	Delete class members. More...

Protected Attributes
GModelPar	m_norm
	Normalization factor. More...

GModelSpectralTablePars	m_table_pars
	Table model parameters. More...

GNdarray	m_spectra
	Spectra. More...

GEbounds	m_ebounds
	Energy boundaries. More...

GFilename	m_filename
	Filename of table. More...

std::string	m_escale_par
	Energy scaling parameter. More...

double	m_escale
	Energy scale. More...

double	m_log10escale
	log10 of energy scale More...

int	m_npars
	Number of parameters. More...

int	m_nebins
	Number of energy bins. More...

std::vector< double >	m_last_values
	Last parameter values. More...

GNodeArray	m_lin_nodes
	Energy nodes of function. More...

GNodeArray	m_log_nodes
	log10(Energy) nodes of function More...

GNdarray	m_lin_values
	Function values and grad's. More...

GNdarray	m_log_values
	log10(Function) values and grad's More...

std::vector< double >	m_prefactor
	Power-law normalisations. More...

std::vector< double >	m_gamma
	Power-law indices. More...

std::vector< double >	m_epivot
	Power-law pivot energies. More...

std::vector< double >	m_flux
	Photon fluxes. More...

std::vector< double >	m_eflux
	Energy fluxes. More...

GEnergy	m_mc_emin
	Minimum energy. More...

GEnergy	m_mc_emax
	Maximum energy. More...

std::vector< double >	m_mc_cum
	Cumulative distribution. More...

std::vector< double >	m_mc_min
	Lower boundary for MC. More...

std::vector< double >	m_mc_max
	Upper boundary for MC. More...

std::vector< double >	m_mc_exp
	Exponent for MC. More...

Protected Attributes inherited from GModelSpectral
std::vector< GModelPar * >	m_pars
	Parameter pointers. More...

Detailed Description

Spectral table model class.

Definition at line 55 of file GModelSpectralTable.hpp.

Constructor & Destructor Documentation

GModelSpectralTable::GModelSpectralTable ( void )

Void constructor.

Definition at line 88 of file GModelSpectralTable.cpp.

References init_members().

Referenced by clone().

GModelSpectralTable::GModelSpectralTable	(	const GFilename &	filename,
		const double &	norm
	)

File constructor.

Parameters

[in]	filename	File name of table model.
[in]	norm	Normalization factor.

Constructs a spectral table model from a FITS file. See the load() method for more information about the expected structure of the FITS file.

Definition at line 107 of file GModelSpectralTable.cpp.

References init_members(), load(), m_norm, GOptimizerPar::range(), GOptimizerPar::remove_range(), GOptimizerPar::scale(), and GOptimizerPar::value().

GModelSpectralTable::GModelSpectralTable	(	const GEbounds &	ebounds,
		const GModelSpectralTablePars &	pars,
		const GNdarray &	spectra
	)

Table model constructor.

Parameters

[in]	ebounds	Energy boundaries.
[in]	pars	Table model parameters.
[in]	spectra	Spectra.

Constructs a spectral table model from energy boundaries, table model parameters, and spectra.

Definition at line 139 of file GModelSpectralTable.cpp.

References GNdarray::dim(), ebounds(), G_CONST, init_members(), m_ebounds, m_spectra, m_table_pars, set_energy_nodes(), set_par_pointers(), GNdarray::shape(), GModelSpectralTablePars::size(), GEbounds::size(), and gammalib::str().

GModelSpectralTable::GModelSpectralTable ( const GXmlElement & xml )

explicit

XML constructor.

Parameters

[in] xml XML element.

Constructs a spectral table model by extracting information from an XML element. See the read() method for more information about the expected structure of the XML element.

Definition at line 207 of file GModelSpectralTable.cpp.

References init_members(), and read().

GModelSpectralTable::GModelSpectralTable ( const GModelSpectralTable & model )

Copy constructor.

Parameters

[in] model Table model.

Definition at line 226 of file GModelSpectralTable.cpp.

References copy_members(), and init_members().

GModelSpectralTable::~GModelSpectralTable ( void )

virtual

Destructor.

Definition at line 243 of file GModelSpectralTable.cpp.

References free_members().

Member Function Documentation

std::string GModelSpectralTable::classname ( void ) const

inlinevirtual

Return class name.

Returns: String containing the class name ("GModelSpectralTable").

Implements GModelSpectral.

Definition at line 164 of file GModelSpectralTable.hpp.

void GModelSpectralTable::clear ( void )

virtual

Clear table model.

Implements GModelSpectral.

Definition at line 298 of file GModelSpectralTable.cpp.

References free_members(), GModelSpectral::free_members(), init_members(), and GModelSpectral::init_members().

Referenced by load().

GModelSpectralTable * GModelSpectralTable::clone ( void ) const

virtual

Clone table model.

Implements GModelSpectral.

Definition at line 316 of file GModelSpectralTable.cpp.

References GModelSpectralTable().

void GModelSpectralTable::copy_members ( const GModelSpectralTable & model )

protected

Copy class members.

Parameters

[in] model Table model.

Definition at line 1222 of file GModelSpectralTable.cpp.

References m_ebounds, m_eflux, m_epivot, m_escale, m_escale_par, m_filename, m_flux, m_gamma, m_last_values, m_lin_nodes, m_lin_values, m_log10escale, m_log_nodes, m_log_values, m_mc_cum, m_mc_emax, m_mc_emin, m_mc_exp, m_mc_max, m_mc_min, m_nebins, m_norm, m_npars, m_prefactor, m_spectra, m_table_pars, set_energy_nodes(), and set_par_pointers().

Referenced by GModelSpectralTable(), and operator=().

GFitsBinTable GModelSpectralTable::create_eng_table ( void ) const

protected

Create ENERGIES FITS table.

Returns: Binary FITS table containing table model energies

The method creates a binary FITS table that is compliant with https://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/general/ogip_92_009/node7.html

Definition at line 1636 of file GModelSpectralTable.cpp.

References GFitsTable::append(), GFitsHDU::card(), GEbounds::emax(), GEbounds::emin(), GFitsHDU::extname(), GEnergy::keV(), m_ebounds, and GEbounds::size().

Referenced by save().

GFitsBinTable GModelSpectralTable::create_par_table ( void ) const

protected

Create PARAMETERS FITS table.

Returns: Binary FITS table containing table model parameters

The method creates a binary FITS table that is compliant with https://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/general/ogip_92_009/node6.html

Definition at line 1497 of file GModelSpectralTable.cpp.

References GFitsTable::append(), GFitsHDU::card(), GFitsHDU::extname(), GOptimizerPar::has_max(), GOptimizerPar::has_min(), GOptimizerPar::is_fixed(), m_table_pars, GOptimizerPar::max(), GOptimizerPar::min(), GOptimizerPar::name(), GModelSpectralTablePars::size(), GModelSpectral::size(), and GOptimizerPar::value().

Referenced by save().

GFitsBinTable GModelSpectralTable::create_spec_table ( void ) const

protected

Create SPECTRA FITS table.

Returns: Binary FITS table containing table model spectra

The method creates a binary FITS table that is compliant with https://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/general/ogip_92_009/node8.html

Definition at line 1681 of file GModelSpectralTable.cpp.

References GFitsTable::append(), GFitsHDU::card(), GNdarray::dim(), GFitsHDU::extname(), m_ebounds, m_spectra, m_table_pars, GNdarray::shape(), GModelSpectralTablePars::size(), GEbounds::size(), and GFitsTableCol::unit().

Referenced by save().

const GEbounds & GModelSpectralTable::ebounds ( void ) const

inline

Return reference to energy boundaries.

Returns: Reference to energy boundaries.

Returns a reference to energy boundaries.

Definition at line 221 of file GModelSpectralTable.hpp.

References m_ebounds.

Referenced by GModelSpectralTable(), and update().

double GModelSpectralTable::eflux	(	const GEnergy &	emin,
		const GEnergy &	emax
	)		const

virtual

Returns model energy flux between emin, emax

Parameters

[in]	emin	Minimum photon energy.
[in]	emax	Maximum photon energy.

Returns: Energy flux (erg/cm2/s).

Computes

\[ \int_{\tt emin}^{\tt emax} S_{\rm E}(E | t) E \, dE \]

where

[emin, emax] is an energy interval, and
\(S_{\rm E}(E | t)\) is the spectral model (ph/cm2/s/MeV).

Implements GModelSpectral.

Definition at line 584 of file GModelSpectralTable.cpp.

References GNodeArray::inx_left(), m_eflux, m_epivot, m_gamma, m_lin_nodes, m_prefactor, GEnergy::MeV(), gammalib::MeV2erg, norm(), gammalib::plaw_energy_flux(), GNodeArray::set_value(), update(), and update_flux().

Referenced by update_flux().

void GModelSpectralTable::energy_scale ( const std::string & name )

Set energy scale.

Parameters

[in] name Table parameter name for energy scaling.

Exceptions

GException::invalid_argument Table model parameter name not found.s

Sets a given table parameter name as energy scale for table model interpolation.

Definition at line 1029 of file GModelSpectralTable.cpp.

References G_READ, m_escale_par, m_log10escale, m_table_pars, GOptimizerPar::name(), scale_energy(), and GModelSpectralTablePars::size().

Referenced by read().

double GModelSpectralTable::eval	(	const GEnergy &	srcEng,
		const GTime &	srcTime = `GTime()`,
		const bool &	gradients = `false`
	)		const

virtual

Evaluate spectral table model.

Parameters

[in]	srcEng	True photon energy.
[in]	srcTime	True photon arrival time.
[in]	gradients	Compute gradients?

Returns: Model value (ph/cm2/s/MeV).

Evaluates

\[ S_{\rm E}(E | t) = {\tt m\_norm} \times \left( w_l F_l(p) + w_r F_r(p) \right) \]

where

\({\tt m\_norm}\) is the normalization factor of the spectral table model,
\(w_l\) is the weight of the spectral vector \(F_l(p)\) with the largest energy \(E_l\) that satisfies \(E<E_l\), and
\(w_r\) is the weight of the spectral vector \(F_r(p)\) with the smallest energy \(E_r\) that satisfies \(E>E_r\).

The weights are computed using

\[ w_r = \frac{\log_{10} E - \log_{10} E_l}{\log_{10} E_r - \log_{10} E_l} \]

and \(w_l = 1 - w_r\).

If gradient is true, the method also computes the parameter gradients using

\[ \frac{\delta S_{\rm E}(E | t)}{\delta {\tt m\_norm}} = \frac{S_{\rm E}(E | t)}{{\tt m\_norm}} \]

and

\[ \frac{\delta S_{\rm E}(E | t)}{\delta p} = {\tt m\_norm} \times \left( w_l \frac{\delta F_l(p)}{\delta p} + w_r \frac{\delta F_r(p)}{\delta p} \right) \]

for all other parameters.

For the computation of \(F_l(p)\), \(F_r(p)\), \(\frac{\delta F_l(p)}{\delta p}\), and \(\frac{\delta F_r(p)}{\delta p}\) see the update() method.

Implements GModelSpectral.

Definition at line 377 of file GModelSpectralTable.cpp.

References GOptimizerPar::factor_gradient(), GOptimizerPar::factor_value(), GNodeArray::inx_left(), GNodeArray::inx_right(), GOptimizerPar::is_free(), gammalib::is_infinite(), gammalib::is_notanumber(), gammalib::ln10, log10(), GEnergy::log10MeV(), m_escale_par, m_lin_values, m_log10escale, m_log_nodes, m_norm, m_table_pars, GOptimizerPar::name(), norm(), GOptimizerPar::scale(), GNodeArray::set_value(), GModelSpectralTablePars::size(), table_par(), update(), GOptimizerPar::value(), GNodeArray::wgt_grad_left(), GNodeArray::wgt_grad_right(), GNodeArray::wgt_left(), and GNodeArray::wgt_right().

const GFilename & GModelSpectralTable::filename ( void ) const

inline

Return file name.

Returns: Name of table file.

Returns the name of the table file.

Definition at line 235 of file GModelSpectralTable.hpp.

References m_filename.

Referenced by load(), and save().

double GModelSpectralTable::flux	(	const GEnergy &	emin,
		const GEnergy &	emax
	)		const

virtual

Returns model photon flux between emin, emax

Parameters

[in]	emin	Minimum photon energy.
[in]	emax	Maximum photon energy.

Returns: Photon flux (ph/cm2/s).

Computes

\[ \int_{\tt emin}^{\tt emax} S_{\rm E}(E | t) dE \]

where

[emin, emax] is an energy interval, and
\(S_{\rm E}(E | t)\) is the spectral model (ph/cm2/s/MeV).

Implements GModelSpectral.

Definition at line 491 of file GModelSpectralTable.cpp.

References GNodeArray::inx_left(), m_epivot, m_flux, m_gamma, m_lin_nodes, m_prefactor, GEnergy::MeV(), norm(), gammalib::plaw_photon_flux(), GNodeArray::set_value(), update(), and update_flux().

Referenced by update_flux(), and update_mc().

void GModelSpectralTable::free_members ( void )

protected

Delete class members.

Definition at line 1272 of file GModelSpectralTable.cpp.

Referenced by clear(), operator=(), and ~GModelSpectralTable().

bool GModelSpectralTable::has_energy_scale ( void ) const

protected

Signal that energy scale was set.

Returns: True if energy scale was set

Definition at line 1479 of file GModelSpectralTable.cpp.

References m_escale.

Referenced by set_energy_nodes().

void GModelSpectralTable::init_members ( void )

protected

Initialise class members.

Definition at line 1164 of file GModelSpectralTable.cpp.

References GModelSpectralTablePars::clear(), GNodeArray::clear(), GEnergy::clear(), GFilename::clear(), GEbounds::clear(), GNdarray::clear(), GOptimizerPar::clear(), GOptimizerPar::free(), GOptimizerPar::gradient(), GOptimizerPar::has_grad(), m_ebounds, m_eflux, m_epivot, m_escale, m_escale_par, m_filename, m_flux, m_gamma, m_last_values, m_lin_nodes, m_lin_values, m_log10escale, m_log_nodes, m_log_values, m_mc_cum, m_mc_emax, m_mc_emin, m_mc_exp, m_mc_max, m_mc_min, m_nebins, m_norm, m_npars, m_prefactor, m_spectra, m_table_pars, GOptimizerPar::name(), GOptimizerPar::range(), GOptimizerPar::scale(), set_par_pointers(), and GOptimizerPar::value().

Referenced by clear(), GModelSpectralTable(), and operator=().

void GModelSpectralTable::load ( const GFilename & filename )

Load table from file.

Parameters

[in] filename File name.

Loads table model from FITS file. The format of the FITS file complies with https://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/general/ogip_92_009/ogip_92_009.html

Definition at line 827 of file GModelSpectralTable.cpp.

References clear(), GFits::close(), filename(), GFilename::is_empty(), load_eng(), load_par(), load_spec(), m_filename, set_energy_nodes(), and set_par_pointers().

Referenced by GModelSpectralTable(), and read().

void GModelSpectralTable::load_eng ( const GFits & fits )

protected

Load data from ENERGIES extension.

Parameters

[in] fits FITS file.

The method loads data from the ENERGIES binary table. The format of the table needs to be compliant with https://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/general/ogip_92_009/node7.html

If energy units are provided in the ENERGIES extension the method decodes the units and interprets the energy values correspondingly.

Definition at line 1867 of file GModelSpectralTable.cpp.

References GEbounds::append(), GFitsHDU::integer(), m_ebounds, and GFits::table().

Referenced by load().

void GModelSpectralTable::load_par ( const GFits & fits )

protected

Load data from PARAMETERS extension.

Parameters

[in] fits FITS file.

Exceptions

GException::invalid_value Non-positive parameter value encountered for logarithmic parameters.

The method loads data from the PARAMETERS binary table. The format of the table needs to be compliant with https://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/general/ogip_92_009/node6.html

Definition at line 1771 of file GModelSpectralTable.cpp.

References GModelSpectralTablePars::append(), GOptimizerPar::factor_range(), GOptimizerPar::fix(), GOptimizerPar::free(), GOptimizerPar::has_grad(), GFitsHDU::integer(), m_table_pars, max(), GModelSpectralTablePar::method(), min(), GFitsHDU::real(), GOptimizerPar::scale(), and GFits::table().

Referenced by load().

void GModelSpectralTable::load_spec ( const GFits & fits )

protected

Load data from SPECTRA extension.

Parameters

[in] fits FITS file.

The method loads data from the SPECTRA binary table. The format of the table needs to be compliant with https://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/general/ogip_92_009/node8.html

Definition at line 1911 of file GModelSpectralTable.cpp.

References GFitsHDU::integer(), GNodeArray::inx_left(), GNodeArray::inx_right(), m_spectra, m_table_pars, GFitsHDU::real(), GNodeArray::set_value(), GFits::table(), GNodeArray::wgt_left(), and GNodeArray::wgt_right().

Referenced by load().

GEnergy GModelSpectralTable::mc	(	const GEnergy &	emin,
		const GEnergy &	emax,
		const GTime &	time,
		GRan &	ran
	)		const

virtual

Returns MC energy between [emin, emax].

Parameters

[in]	emin	Minimum photon energy.
[in]	emax	Maximum photon energy.
[in]	time	True photon arrival time.
[in,out]	ran	Random number generator.

Returns: Energy.

Returns Monte Carlo energy by randomly drawing from a broken power law defined by the file function.

Implements GModelSpectral.

Definition at line 675 of file GModelSpectralTable.cpp.

References gammalib::check_energy_interval(), exp(), G_MC, log(), m_mc_cum, m_mc_exp, m_mc_max, m_mc_min, GEnergy::MeV(), GRan::uniform(), update(), update_flux(), and update_mc().

double GModelSpectralTable::norm ( void ) const

inline

Return normalization factor.

Returns: Normalization factor.

Returns the normalization factor.

Definition at line 192 of file GModelSpectralTable.hpp.

References m_norm, and GOptimizerPar::value().

Referenced by eflux(), eval(), flux(), read(), update_mc(), and write().

void GModelSpectralTable::norm ( const double & norm )

inline

Set normalization factor.

Parameters

[in] norm Normalization factor.

Sets the normalization factor.

Definition at line 206 of file GModelSpectralTable.hpp.

References m_norm, and GOptimizerPar::value().

int GModelSpectralTable::nspectra ( void ) const

Return number of spectra in table model.

Returns: Number of spectra.

If the table model is empty the method returns zero.

Definition at line 925 of file GModelSpectralTable.cpp.

References m_table_pars, and GModelSpectralTablePars::size().

Referenced by print(), and scale_energy().

GModelSpectralTable & GModelSpectralTable::operator= ( const GModelSpectralTable & model )

virtual

Assignment operator.

Parameters

[in] model Table model.

Returns: Table model.

Definition at line 265 of file GModelSpectralTable.cpp.

References copy_members(), free_members(), init_members(), and GModelSpectral::operator=().

int GModelSpectralTable::par_index ( const std::string & name ) const

protected

Return index for parameter name.

Parameters

[in] name Parameter name.

Returns: Parameter index.

Exceptions

GException::invalid_argument Parameter name not found in spectral table.

Definition at line 1991 of file GModelSpectralTable.cpp.

References G_PAR_INDEX, m_table_pars, and GModelSpectral::size().

Referenced by scale_energy(), and table_par().

std::string GModelSpectralTable::print ( const GChatter & chatter = NORMAL ) const

virtual

Print table model information.

Parameters

[in] chatter Chattiness.

Returns: String containing table model information.

Implements GModelSpectral.

Definition at line 1074 of file GModelSpectralTable.cpp.

References GModelSpectralTablePars::append(), GNdarray::dim(), GEbounds::emax(), GEbounds::emin(), m_ebounds, m_filename, GModelSpectral::m_pars, m_spectra, m_table_pars, max(), min(), nspectra(), gammalib::parformat(), GEnergy::print(), GNdarray::shape(), SILENT, GModelSpectralTablePars::size(), GEbounds::size(), GModelSpectral::size(), gammalib::str(), and GFilename::url().

void GModelSpectralTable::read ( const GXmlElement & xml )

virtual

Read model from XML element.

Parameters

[in] xml XML element containing power law model information.

Reads the spectral information from an XML element. The format of the XML elements is

<spectrum type="TableModel" file="..">
  <parameter name="Normalization" scale=".." value=".." min=".." max=".." free=".."/>
</spectrum>

Optionally, values for the model table parameters can also be provided.

Implements GModelSpectral.

Definition at line 739 of file GModelSpectralTable.cpp.

References GXmlElement::attribute(), energy_scale(), G_READ, GXmlElement::has_attribute(), load(), m_norm, m_table_pars, GOptimizerPar::name(), norm(), GModelPar::read(), GModelSpectralTablePars::size(), gammalib::xml_file_expand(), gammalib::xml_get_par(), and gammalib::xml_has_par().

Referenced by GModelSpectralTable().

void GModelSpectralTable::save	(	const GFilename &	filename,
		const bool &	clobber = `false`
	)		const

Save table into file.

Parameters

[in]	filename	File name.
[in]	clobber	Overwrite existing file?

Save the table model into a FITS file. The FITS file will contain three binary table extensions:

  * PARAMETERS - Table model parameters
  * ENERGIES - Table model energies
  * SPECTRA - Table model spectra

The format of the FITS file complies with https://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/general/ogip_92_009/ogip_92_009.html

Definition at line 878 of file GModelSpectralTable.cpp.

References GFits::append(), GFitsHDU::card(), create_eng_table(), create_par_table(), create_spec_table(), filename(), m_filename, GFits::saveto(), and GFilename::url().

void GModelSpectralTable::scale_energy ( void )

protected

Scale energy.

Scales the energy axis so that all spectra are relative to the energy scaling parameter. This scaling is useful for better interpolation of spectra that have a feature that depends of an energy parameter.

Definition at line 1359 of file GModelSpectralTable.cpp.

References GNdarray::index(), GNodeArray::inx_left(), GNodeArray::inx_right(), log10(), m_ebounds, m_escale, m_escale_par, m_last_values, m_lin_nodes, m_log10escale, m_log_nodes, m_spectra, m_table_pars, nspectra(), par_index(), GNodeArray::set_value(), GModelSpectralTablePars::size(), GNodeArray::size(), GEbounds::size(), table_par(), GModelSpectralTablePar::values(), GNodeArray::wgt_left(), and GNodeArray::wgt_right().

Referenced by energy_scale().

void GModelSpectralTable::set_energy_nodes ( void )

protected

Set energy nodes from energy boundaries.

Definition at line 1308 of file GModelSpectralTable.cpp.

References GNodeArray::append(), GEbounds::elogmean(), has_energy_scale(), log10(), m_ebounds, m_escale, m_lin_nodes, m_lin_values, m_log10escale, m_log_nodes, m_log_values, GEnergy::MeV(), and GEbounds::size().

Referenced by copy_members(), GModelSpectralTable(), and load().

void GModelSpectralTable::set_par_pointers ( void )

protected

Set parameter pointers.

Definition at line 1282 of file GModelSpectralTable.cpp.

References m_norm, GModelSpectral::m_pars, m_table_pars, and GModelSpectralTablePars::size().

Referenced by copy_members(), GModelSpectralTable(), init_members(), and load().

GModelSpectralTablePar & GModelSpectralTable::table_par ( const int & index )

Return reference to table parameter.

Parameters

[in] index Table parameter index [0,...,size()[.

Returns: Reference to table parameter.

Definition at line 954 of file GModelSpectralTable.cpp.

References G_TABLE_PAR, m_table_pars, and GModelSpectral::size().

Referenced by eval(), and scale_energy().

const GModelSpectralTablePar & GModelSpectralTable::table_par ( const int & index ) const

Return const reference to table parameter.

Parameters

[in] index Table parameter index [0,...,size()[.

Returns: Const reference to table parameter.

Definition at line 973 of file GModelSpectralTable.cpp.

References G_TABLE_PAR, m_table_pars, and GModelSpectral::size().

GModelSpectralTablePar & GModelSpectralTable::table_par ( const std::string & name )

Return reference to table parameter.

Parameters

[in] name Table parameter name.

Returns: Reference to table parameter.

Definition at line 992 of file GModelSpectralTable.cpp.

References m_table_pars, and par_index().

const GModelSpectralTablePar & GModelSpectralTable::table_par ( const std::string & name ) const

Return const reference to table parameter.

Parameters

[in] name Table parameter name.

Returns: Const reference to table parameter.

Definition at line 1008 of file GModelSpectralTable.cpp.

References m_table_pars, and par_index().

std::string GModelSpectralTable::type ( void ) const

inlinevirtual

Return model type.

Returns: "TableModel".

Returns the type of the spectral model.

Implements GModelSpectral.

Definition at line 178 of file GModelSpectralTable.hpp.

Referenced by write().

void GModelSpectralTable::update ( void ) const

protected

Update cache for spectral table model computation.

Update interval vectors for function values and gradients. An update is performed in case that some of the parameter values have changed. The method sets the following cache members:

 m_npars (number of model parameters)
 m_nebins (number of energy bins in spectra)
 m_last_values (last model parameter values)
 m_lin_values (function values)

The array \({\tt m\_lin\_values}\) holds the vector \(F(E|p)\) as function of energy \(E\), computed for the current set of parameters \(p\). The computation is done using the N-dimensional linear interpolation

\[ F(E|p) = \sum_{k=1}^M \left( \prod_{i=1}^N w_x(p) \right) F_p(E) \]

where

\(M=2^N\) is the number of parameter combinations,
\(N\) is the number of table model parameters,
\(w_x(p) = \{w_0^l, w_0^r, w_1^l, w_1^r, ...\}\) is an array of subsequently arranged left and right weighting factors for the linear interpolation of parameters, where \(w_0^l\) and \(w_0^r\) are the left and right weighting factors for the first parameter, \(w_1^l\) and \(w_1^r\) are the left and right weighting factors for the second parameter, and so on,
\(x=2 k + i/2^k \mod 2\) is the index in the array of weighting factors for parameter combination \(k\) and parameter \(i\), and
\(F_p(E)\) are the table model spectra, computed for a grid of possible parameter values.

For each parameter \(i\), the weighting factors \(w_i^l(p)\) and \(w_i^r(p)\) are computed using

\[ w_i^r(p) = \frac{p - p_l}{p_r - p_l} \]

and \(w_i^l(p) = 1 - w_i^r(p)\), where \(p_l\) is the largest parameter value that satisfies \(p<p_l\) and \(p_r\) is the smallest parameter value that satisfies \(p>p_r\).

The method also computes the gradients

\[ \frac{\delta F(E|p)}{\delta p} = \sum_{k=1}^M \left( \frac{\delta w_{x_p}(p)}{\delta p} \prod_{i=1 \land i \neq i_p}^N w_x(p) \right) F_p(E) \]

where \(x_p = 2 k + i_p/2^k \mod 2\).

Definition at line 2078 of file GModelSpectralTable.cpp.

References ebounds(), GNodeArray::inx_left(), GNodeArray::inx_right(), m_last_values, m_lin_values, m_log_values, m_nebins, m_npars, m_spectra, m_table_pars, GNodeArray::set_value(), GModelSpectralTablePars::size(), GEbounds::size(), GOptimizerPar::value(), GNodeArray::wgt_grad_left(), GNodeArray::wgt_grad_right(), GNodeArray::wgt_left(), and GNodeArray::wgt_right().

Referenced by eflux(), eval(), flux(), and mc().

void GModelSpectralTable::update_flux ( void ) const

protected

Update flux cache.

Definition at line 2293 of file GModelSpectralTable.cpp.

References eflux(), flux(), log(), m_eflux, m_epivot, m_flux, m_gamma, m_lin_nodes, m_lin_values, m_prefactor, gammalib::MeV2erg, gammalib::plaw_energy_flux(), gammalib::plaw_photon_flux(), pow(), GNodeArray::size(), and sqrt().

Referenced by eflux(), flux(), and mc().

void GModelSpectralTable::update_mc	(	const GEnergy &	emin,
		const GEnergy &	emax
	)		const

protected

Update MC cache.

Parameters

[in]	emin	Minimum energy.
[in]	emax	Maximum energy.

This method sets up an array of indices and the cumulative distribution function needed for MC simulations.

Definition at line 2356 of file GModelSpectralTable.cpp.

References abs(), flux(), GNodeArray::inx_left(), log(), m_epivot, m_flux, m_gamma, m_lin_nodes, m_mc_cum, m_mc_emax, m_mc_emin, m_mc_exp, m_mc_max, m_mc_min, m_prefactor, GEnergy::MeV(), norm(), gammalib::plaw_photon_flux(), pow(), and GNodeArray::set_value().

Referenced by mc().

void GModelSpectralTable::write ( GXmlElement & xml ) const

virtual

Write model into XML element.

Parameters

[in] xml XML element into which model information is written.

Writes the spectral information into an XML element. The format of the XML element is

<spectrum type="FileFunction" file="..">
  <parameter name="Normalization" scale=".." value=".." min=".." max=".." free=".."/>
</spectrum>

In addition, the method writes the model table parameters into the XML file.

Note that the function nodes will not be written since they will not be altered by any method.

Implements GModelSpectral.

Definition at line 788 of file GModelSpectralTable.cpp.

References GXmlElement::attribute(), G_WRITE, m_escale_par, m_filename, m_norm, m_table_pars, GOptimizerPar::name(), norm(), GModelSpectralTablePars::size(), type(), GModelPar::write(), gammalib::xml_check_type(), gammalib::xml_file_reduce(), and gammalib::xml_need_par().

Member Data Documentation

GEbounds GModelSpectralTable::m_ebounds

protected

Energy boundaries.

Definition at line 130 of file GModelSpectralTable.hpp.

Referenced by copy_members(), create_eng_table(), create_spec_table(), ebounds(), GModelSpectralTable(), init_members(), load_eng(), print(), scale_energy(), and set_energy_nodes().

std::vector<double> GModelSpectralTable::m_eflux

mutableprotected

Energy fluxes.

Definition at line 148 of file GModelSpectralTable.hpp.

Referenced by copy_members(), eflux(), init_members(), and update_flux().

std::vector<double> GModelSpectralTable::m_epivot

mutableprotected

Power-law pivot energies.

Definition at line 146 of file GModelSpectralTable.hpp.

Referenced by copy_members(), eflux(), flux(), init_members(), update_flux(), and update_mc().

double GModelSpectralTable::m_escale

protected

Energy scale.

Definition at line 133 of file GModelSpectralTable.hpp.

Referenced by copy_members(), has_energy_scale(), init_members(), scale_energy(), and set_energy_nodes().

std::string GModelSpectralTable::m_escale_par

protected

Energy scaling parameter.

Definition at line 132 of file GModelSpectralTable.hpp.

Referenced by copy_members(), energy_scale(), eval(), init_members(), scale_energy(), and write().

GFilename GModelSpectralTable::m_filename

mutableprotected

Filename of table.

Definition at line 131 of file GModelSpectralTable.hpp.

Referenced by copy_members(), filename(), init_members(), load(), print(), save(), and write().

std::vector<double> GModelSpectralTable::m_flux

mutableprotected

Photon fluxes.

Definition at line 147 of file GModelSpectralTable.hpp.

Referenced by copy_members(), flux(), init_members(), update_flux(), and update_mc().

std::vector<double> GModelSpectralTable::m_gamma

mutableprotected

Power-law indices.

Definition at line 145 of file GModelSpectralTable.hpp.

Referenced by copy_members(), eflux(), flux(), init_members(), update_flux(), and update_mc().

std::vector<double> GModelSpectralTable::m_last_values

mutableprotected

Last parameter values.

Definition at line 139 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), scale_energy(), and update().

GNodeArray GModelSpectralTable::m_lin_nodes

mutableprotected

Energy nodes of function.

Definition at line 140 of file GModelSpectralTable.hpp.

Referenced by copy_members(), eflux(), flux(), init_members(), scale_energy(), set_energy_nodes(), update_flux(), and update_mc().

GNdarray GModelSpectralTable::m_lin_values

mutableprotected

Function values and grad's.

Definition at line 142 of file GModelSpectralTable.hpp.

Referenced by copy_members(), eval(), init_members(), set_energy_nodes(), update(), and update_flux().

double GModelSpectralTable::m_log10escale

protected

log10 of energy scale

Definition at line 134 of file GModelSpectralTable.hpp.

Referenced by copy_members(), energy_scale(), eval(), init_members(), scale_energy(), and set_energy_nodes().

GNodeArray GModelSpectralTable::m_log_nodes

mutableprotected

log10(Energy) nodes of function

Definition at line 141 of file GModelSpectralTable.hpp.

Referenced by copy_members(), eval(), init_members(), scale_energy(), and set_energy_nodes().

GNdarray GModelSpectralTable::m_log_values

mutableprotected

log10(Function) values and grad's

Definition at line 143 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), set_energy_nodes(), and update().

std::vector<double> GModelSpectralTable::m_mc_cum

mutableprotected

Cumulative distribution.

Definition at line 151 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), mc(), and update_mc().

GEnergy GModelSpectralTable::m_mc_emax

mutableprotected

Maximum energy.

Definition at line 150 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), and update_mc().

GEnergy GModelSpectralTable::m_mc_emin

mutableprotected

Minimum energy.

Definition at line 149 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), and update_mc().

std::vector<double> GModelSpectralTable::m_mc_exp

mutableprotected

Exponent for MC.

Definition at line 154 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), mc(), and update_mc().

std::vector<double> GModelSpectralTable::m_mc_max

mutableprotected

Upper boundary for MC.

Definition at line 153 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), mc(), and update_mc().

std::vector<double> GModelSpectralTable::m_mc_min

mutableprotected

Lower boundary for MC.

Definition at line 152 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), mc(), and update_mc().

int GModelSpectralTable::m_nebins

mutableprotected

Number of energy bins.

Definition at line 138 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), and update().

GModelPar GModelSpectralTable::m_norm

protected

Normalization factor.

Definition at line 127 of file GModelSpectralTable.hpp.

Referenced by copy_members(), eval(), GModelSpectralTable(), init_members(), norm(), read(), set_par_pointers(), and write().

int GModelSpectralTable::m_npars

mutableprotected

Number of parameters.

Definition at line 137 of file GModelSpectralTable.hpp.

Referenced by copy_members(), init_members(), and update().

std::vector<double> GModelSpectralTable::m_prefactor

mutableprotected

Power-law normalisations.

Definition at line 144 of file GModelSpectralTable.hpp.

Referenced by copy_members(), eflux(), flux(), init_members(), update_flux(), and update_mc().

GNdarray GModelSpectralTable::m_spectra

protected

Spectra.

Definition at line 129 of file GModelSpectralTable.hpp.

Referenced by copy_members(), create_spec_table(), GModelSpectralTable(), init_members(), load_spec(), print(), scale_energy(), and update().

GModelSpectralTablePars GModelSpectralTable::m_table_pars

protected

Table model parameters.

Definition at line 128 of file GModelSpectralTable.hpp.

Referenced by copy_members(), create_par_table(), create_spec_table(), energy_scale(), eval(), GModelSpectralTable(), init_members(), load_par(), load_spec(), nspectra(), par_index(), print(), read(), scale_energy(), set_par_pointers(), table_par(), update(), and write().

The documentation for this class was generated from the following files:

include/GModelSpectralTable.hpp
src/model/GModelSpectralTable.cpp

Public Member Functions

Protected Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation