model_image_sed_fitted_riaf Class Reference

Defines an image model associated with the SED-fitted RIAF models employed in Broderick, Fish, Doeleman and Loeb (2011) and Broderick et al. (2016). More...

#include "model/model_image_sed_fitted_riaf.h"

Inheritance diagram for model_image_sed_fitted_riaf:

Collaboration diagram for model_image_sed_fitted_riaf:

Public Member Functions
	model_image_sed_fitted_riaf (std::string sed_fit_parameter_file=Themis::utils::global_path("src/VRT2/DataFiles/2010_combined_fit_parameters.d"), double frequency=230e9)
virtual size_t	size () const
	Returns the number of the parameters the model expects.
virtual std::string	model_tag () const
	Return a string that contains a unique identifying tag for use with the ThemisPy plotting features.
void	use_small_images ()
void	set_image_resolution (int Nray)
virtual void	set_mpi_communicator (MPI_Comm comm)
Public Member Functions inherited from model_image
virtual void	generate_model (std::vector< double > parameters)
	A one-time generate function that permits model construction prior to calling the visibility_amplitude, closure_phase, etc. for each datum. Takes a vector of parameters.
virtual void	generate_complex_visibilities ()
	A one-time generate function that will generate the complex visibilities and store them. This must be called after generate_model has been called.
virtual std::complex< double >	visibility (datum_visibility &d, double accuracy)
	Returns complex visibility in Jy computed from the image given a datum_visibility_amplitude object, containing all of the accoutrements. While this provides access to the actual data value, the two could be separated if necessary. Also takes an accuracy parameter with the same units as the data, indicating the accuracy with which the model must generate a comparison value. Note that this can be redefined in child classes.
virtual double	visibility_amplitude (datum_visibility_amplitude &d, double accuracy)
	Returns visibility ampitudes in Jy computed from the image given a datum_visibility_amplitude object, containing all of the accoutrements. While this provides access to the actual data value, the two could be separated if necessary. Also takes an accuracy parameter with the same units as the data, indicating the accuracy with which the model must generate a comparison value. Note that this can be redefined in child classes.
virtual double	closure_phase (datum_closure_phase &d, double accuracy)
	Returns closure phase in degrees computed from the image given a datum_closure_phase object, containing all of the accoutrements. While this provides access to the actual data value, the two could be separated if necessary. Also takes an accuracy parameter with the same units as the data, indicating the accuracy with which the model must generate a comparison value. Note that this can be redefined in child classes.
virtual double	closure_amplitude (datum_closure_amplitude &d, double accuracy)
	Returns closure amplitude computed from the image given a datum_closure_phase object, containing all of the accoutrements. While this provides access to the actual data value, the two could be separated if necessary. Also takes an accuracy parameter with the same units as the data, indicating the accuracy with which the model must generate a comparison value. Note that this can be redefined in child classes.
void	output_image (std::string fname, bool rotate=false)
void	get_image (std::vector< std::vector< double > > &alpha, std::vector< std::vector< double > > &beta, std::vector< std::vector< double > > &I) const
	Provides direct access to the constructed image. Sets a 2D grid of angles (alpha, beta) in radians and intensities in Jy per steradian.
void	get_visibilities (std::vector< std::vector< double > > &u, std::vector< std::vector< double > > &v, std::vector< std::vector< std::complex< double > > > &V) const
	Provides direct access to the complex visibilities. Sets a 2D grid of baselines (u,v) in lambda, and visibilites in Jy.
void	get_visibility_amplitudes (std::vector< std::vector< double > > &u, std::vector< std::vector< double > > &v, std::vector< std::vector< double > > &V) const
	Provides direct access to the visibility amplitudes. Sets a 2D grid of baselines (u,v) in lambda, and visibilites in Jy.
void	use_spline_interp (bool use_spline)
	Provides ability to use bicubic spline interpolator (true) instead of regular bicubic. Code defaults to false.
void	write_model_tag_file (std::string tagfilename="model_image.tag") const
	Write a unique identifying tag for use with the ThemisPy plotting features. This calls the overloaded version with the outstream, which is the only function that need be rewritten in child classes.
virtual void	write_model_tag_file (std::ofstream &tagout) const
	Write a unique identifying tag for use with the ThemisPy plotting features. For most child classes, the default implementation is suffcient. However, should that not be the case, this is the only function that need be rewritten in child classes.

Protected Attributes
MPI_Comm	_comm
Protected Attributes inherited from model_image
MPI_Comm	_comm
bool	_generated_model
	True when a model is generated with generate_model.
bool	_generated_visibilities
	True when model visibilities have been computed.
bool	_use_spline
	True when want to use bicubic spline interpolator.
double	_position_angle
	Position angle of the image, assumed to be the last element of the parameter list passed to generate_model. Assumed to be in radians and defined E of N.
std::vector< double >	_current_parameters
	Last set of parameters passed to generate_image, useful to determine if it is necessary to recompute the model. Useful, e.g., if we are varying only position angle, or if recomputing for a number of different data sets at the same set of parameters.
std::vector< std::vector< double > >	_alpha
	2D grid of horizonal pixel locations in radians, relative to the fiducial direction of the image (i.e., unrotated by the position angle).
std::vector< std::vector< double > >	_beta
	2D grid of vertical pixel locations in radians, relative to the fiducial direction of the image (i.e., unrotated by the position angle).
std::vector< std::vector< double > >	_I
	2D grid of intensities at pixel locations in Jy/str.
Protected Attributes inherited from model_visibility
MPI_Comm	_comm
Protected Attributes inherited from model_visibility_amplitude
MPI_Comm	_comm
Protected Attributes inherited from model_closure_phase
MPI_Comm	_comm
Protected Attributes inherited from model_closure_amplitude
MPI_Comm	_comm

Private Member Functions
virtual void	generate_image (vector< double > parameters, vector< vector< double > > &I, vector< vector< double > > &alpha, vector< vector< double > > &beta)
double	generate_renormalized_image (double density_factor, int Nrays, vector< double > parameters, vector< vector< double > > &I, vector< vector< double > > &alpha, vector< vector< double > > &beta)
void	open_error_streams ()
	Opens model-communicator-specific error streams for debugging output.

Private Attributes
VRT2::SgrA_PolintDiskModelParameters2010	_sdmp
	Interpolator object to return.
const double	_M
	Mass of Sgr A* in cm.
const double	_D
	Distance to Sgr A* in cm.
const double	_frequency
	Frequency at which model images in Hz.
int	_Nray
	Image resolution.
std::ofstream	_merr
	Model-specific error stream.

Detailed Description

Provides explicit implementation of the model_image object for the SED-fitted semi-analytic RIAF models. There are additional tuning parameters that may impact accuracy, e.g., distance to the image screen, number of rays to produce, etc. See the appropriate parts of model_image_sed_fitted_riaf.cpp for more.

Parameter list:

• parameters[0] ... Black hole spin parameter (-1 to 1).
• parameters[1] ... Cosine of the spin inclination relative to line of sight.
• parameters[2] ... Position angle (in model_image) in radians.

Warning

Requires the VRT2 library to be installed and the SED fitted disk parameter file, which may be found in "Themis/src/VRT2/DataFiles/2010_combined_fit_parameters.d".

Constructor & Destructor Documentation

model_image_sed_fitted_riaf	(	std::string	sed_fit_parameter_file = Themis::utils::global_path("src/VRT2/DataFiles/2010_combined_fit_parameters.d"),
		double	frequency = 230e9
	)

Constructor to make an SED-fitted RIAF model. Takes the disk parameter fit file and frequency at which to generate models. The mass of and distance to Sgr A* are hard-coded to be consistent with the SED fitting.

Here is the call graph for this function:

Member Function Documentation

void generate_image ( vector< double >  parameters, vector< vector< double > > &  I, vector< vector< double > > &  alpha, vector< vector< double > > &  beta  )

privatevirtual

Generates and returns rectalinear grid of intensities associated with the RIAF image model in Jy/str located at pixels centered on angular positions alpha and beta, both specified in radians and aligned with a fiducial direction. Note that the parameter vector has had the position removed.

Here is the call graph for this function:

double generate_renormalized_image ( double  density_factor, int  Nrays, vector< double >  parameters, vector< vector< double > > &  I, vector< vector< double > > &  alpha, vector< vector< double > > &  beta  )

private

Generates an image with all of the electron densities renormalized by density_factor, modeling a variation in the accretion rate. Returns the integrated intensity. Optionally returns I, alpha, beta.

Here is the call graph for this function:

Here is the caller graph for this function:

void set_image_resolution

(

int

Nray

)

Sets model_image_sed_fitted_riaf to generate production images with resolution Nray x Nray. The default is 128x128, which is probably larger than required in practice.

Here is the caller graph for this function:

virtual void set_mpi_communicator ( MPI_Comm  comm )

inlinevirtual

Defines a set of processors provided to the model for parallel computation via an MPI communicator. Only facilates code parallelization if the model computation is parallelized via MPI. The Following plot shows the scaling plot for the SED-fitted RIAF model. The green line shows the ideal case of linear scaling where the run time is inversely proportional to the number of MPI processes used. The purple line shows how the sampler scales with the number of MPI processes. Up to \(16\) MPI processes the scaling closely follows the linear scaling and deviation stays within \(\%20\). Using \(32\) MPI processes the deviation increases to \(\%50\).

SED-fitted RIAF scaling plot. The green line shows the linear scaling.

Reimplemented from model_image.

Here is the call graph for this function:

void use_small_images

(

)

Sets model_image_sed_fitted_riaf to generate 32x32 images (via set_image_resolution). This is usually used only for testing purposes.

Here is the call graph for this function:

---

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

• model/model_image_sed_fitted_riaf.h
• model/model_image_sed_fitted_riaf.cpp