signal Module Reference

Defines the base class of the type signal_class. More...

Data Types
type	signal_class
	Base class representing a signal. More...
interface	signal_construct
	Interface for the construction of signal_class types using signal_construct_magnetic, signal_construct_sxrem, signal_construct_intpol, signal_construct_thomson, signal_construct_extcurz, signal_construct_mse, signal_construct_ece, signal_construct_limiter, signal_construct_prior_gaussian, signal_construct_feedback, signal_construct_combination or signal_construct_diagnostic_netcdf. More...
type	signal_pointer
	Pointer to a signal object. Used for creating arrays of signal pointers. This is needed because fortran does not allow arrays of pointers directly. More...

Functions/Subroutines
subroutine	signal_construct_new (this, s_name, l_name, units, observed, sigma, weight, s_index, o_index)
	Construct new signal_class object. More...
subroutine	signal_construct_diagnostic_netcdf (this, mdsig_iou, observed, sigma, weight, s_index, o_index)
	Construct a signal_class containing a diagnostic object. More...
subroutine	signal_destruct (this)
	Deconstruct a signal_class object. More...
real(rprec) function, dimension(4)	signal_get_modeled_signal_cache (this, a_model, sigma, use_cache, last_value)
	Calculates the modeled signal. More...
real(rprec) function, dimension(4)	signal_get_modeled_signal_last (this, a_model, sigma, last_value)
	Calculates the modeled signal. More...
real(rprec) function	signal_get_observed_signal (this, a_model)
	Calculates the observed signal. More...
real(rprec) function	signal_get_g2 (this, a_model, use_cache, last_value)
	Calculates the g^2 contribution of a signal. More...
real(rprec) function	signal_get_e (this, a_model, use_cache, last_value)
	Calculates the e contribution of a signal. More...
real(rprec) function	signal_get_sigma2 (this)
	Calculates the total sigma^2 of a signal. More...
character(len=data_name_length) function	signal_get_type (this)
	Gets a discription of the signal type. More...
subroutine	signal_get_header (this, header)
	Gets a discription of the model and model sigma array indices. More...
real(rprec) function	signal_get_gp_i (this, a_model, i, flags)
	Gets the guassian process kernel for a signal and a position. More...
real(rprec) function	signal_get_gp_s (this, a_model, signal, flags)
	Gets the guassian process kernel for a signal and a signal. More...
real(rprec) function	signal_get_gp_x (this, a_model, x_cart, flags)
	Gets the guassian process kernel for a signal and cartesian position. More...
subroutine	signal_scale_and_offset (this, a_model, value)
	Apply scale and offset to the value. More...
character(len=data_short_name_length) function	signal_make_short_name (name, index)
	Generate a short name by appending an index. More...
subroutine	signal_write (this, iou, index, a_model)
	Write out the signal information to an output file. More...
subroutine	signal_write_header (this, iou)
	Write out the signal header information to an output file. More...
subroutine	signals_write_auxiliary (this, iou, index, a_model)
	Write out any auxiliary signal information to an output file. More...
subroutine	signal_write_step_data (this, a_model, result_ncid, current_step, index, signal_model_value_id, signal_sigma_value_id)
	Write out the signal data for a step to the result netcdf file. More...
subroutine	signal_sync_child (this, index, recon_comm)
	Syncronize a child signal state to the parent. More...

Detailed Description

Defines the base class of the type signal_class.

signal Sub Classes: magnetic, sxrem, intpol, thomson, extcurz, mse, ece, limiter, prior_gaussian, and feedback

Function/Subroutine Documentation

signal_construct_diagnostic_netcdf()

subroutine signal::signal_construct_diagnostic_netcdf	(	class (signal_class), intent(inout)	this,
		integer, intent(in)	mdsig_iou,
		real(rprec), intent(in)	observed,
		real(rprec), intent(in)	sigma,
		real(rprec), intent(in)	weight,
		integer, intent(in)	s_index,
		integer, intent(in)	o_index
	)

Construct a signal_class containing a diagnostic object.

Reads the s_name, l_name and units from a netcdf file. Used for constructing the magnetic objects.

Parameters

[in,out]	signal_class	An instance of a signal_class.
[in]	mdsig_iou	NetCDF id of the opened mdsig file.
[in]	observed	Observed value of the signal.
[in]	sigma	Observed sigma of the signal.
[in]	weight	Weight factor of the signal.
[in]	s_index	Index of the model signal scale factor.
[in]	o_index	Index of the model signal offset factor.

Returns

Apointer to a constructed signal_class object.

Definition at line 206 of file signal.f.

signal_construct_new()

subroutine signal::signal_construct_new	(	class (signal_class), intent(inout)	this,
		character(len=*), intent(in)	s_name,
		character(len=*), intent(in)	l_name,
		character(len=*), intent(in)	units,
		real(rprec), intent(in)	observed,
		real(rprec), intent(in)	sigma,
		real(rprec), intent(in)	weight,
		integer, intent(in)	s_index,
		integer, intent(in)	o_index
	)

Construct new signal_class object.

Allocates memory and initializes a signal_class object.

Parameters

[in,out]	this	An instance of a signal_class.
[in]	s_name	The short name of the signal.
[in]	l_name	The short name of the signal.
[in]	units	The units of the signal.
[in]	observed	Observed value of the signal.
[in]	sigma	Observed sigma of the signal.
[in]	weight	Weight factor of the signal.
[in]	s_index	Index of the model signal scale factor.
[in]	o_index	Index of the model signal offset factor.

Definition at line 145 of file signal.f.

signal_destruct()

subroutine signal::signal_destruct

(

type (signal_class), intent(inout)

this

)

Deconstruct a signal_class object.

Deallocates memory and uninitializes a signal_class object.

Parameters

[in,out]

this

A signal_class instance.

Definition at line 251 of file signal.f.

signal_get_e()

real (rprec) function signal::signal_get_e	(	class (signal_class), intent(inout)	this,
		type (model_class), pointer	a_model,
		logical, intent(in)	use_cache,
		real (rprec), dimension(4), intent(in)	last_value
	)

Calculates the e contribution of a signal.

Gets the e contribution of an individual signal. e = SQRT(W)*(O - M)/SQRT(sigma_o^2 + sigma_m^2)

Parameters

[in,out]	this	A diagnostic_class instance.
[in,out]	a_model	A model instance.
[in]	use_cache	If the signals have been already calculated this should be false. Otherwise set to true.
[in]	last_value	Last good value in case the signal did not change.

Returns

The e value.

Definition at line 460 of file signal.f.

signal_get_g2()

real (rprec) function signal::signal_get_g2	(	class (signal_class), intent(inout)	this,
		type (model_class), pointer	a_model,
		logical, intent(in)	use_cache,
		real (rprec), dimension(4), intent(in)	last_value
	)

Calculates the g^2 contribution of a signal.

Gets the g^2 contribution of an individual signal. g^2 = W*(O - M)^2/(sigma_o^2 + sigma_m^2) = e*e

Parameters

[in,out]	this	A diagnostic_class instance.
[in,out]	a_model	A model instance.
[in]	use_cache	If the signals have been already calculated this should be false. Otherwise set to true.
[in]	last_value	Last good value in case the signal did not change.

Returns

The g^2 value.

Definition at line 423 of file signal.f.

signal_get_gp_i()

real (rprec) function signal::signal_get_gp_i	(	class (signal_class), intent(in)	this,
		type (model_class), pointer	a_model,
		integer, intent(in)	i,
		integer, intent(in)	flags
	)

Gets the guassian process kernel for a signal and a position.

This method is meant to be overwritten by a subclass method. As a result, returns the get_gp_i method of the subclass instance signal_class was constructed with.

See also

sxrem::sxrem_get_gp_i

intpol::intpol_get_gp_i

thomson::thomson_get_gp_i

ece::ece_get_gp_i

prior_gaussian::prior_gaussian_get_gp_i

Parameters

[in]	this	A signal_class instance.
[in]	a_model	A model_class instance.
[in]	i	Index of the position for the kernel.
[in]	flags	State flags to send to the kernel.

Returns

Kernel value for the position and the signal.

Definition at line 635 of file signal.f.

signal_get_gp_s()

real (rprec) function signal::signal_get_gp_s	(	class (signal_class), intent(in)	this,
		type (model_class), pointer	a_model,
		class (signal_class), pointer	signal,
		integer, intent(in)	flags
	)

Gets the guassian process kernel for a signal and a signal.

This method is meant to be overwritten by a subclass method. As a result, returns the get_gp_s method of the subclass instance signal_class was constructed with. The second signal is cast to a void pointer so it can be send to the subclass instance.

See also

sxrem::sxrem_get_gp_s

intpol::intpol_get_gp_s

thomson::thomson_get_gp_s

ece::ece_get_gp_s

prior_gaussian::prior_gaussian_get_gp_s

Parameters

[in]	this	A signal_class instance.
[in]	a_model	A model_class instance.
[in]	signal	A signal_class instance for the second signal.
[in]	flags	State flags to send to the kernel.

Returns

Kernel value for the signal and the signal.

Definition at line 677 of file signal.f.

signal_get_gp_x()

real (rprec) function signal::signal_get_gp_x	(	class (signal_class), intent(in)	this,
		type (model_class), pointer	a_model,
		real (rprec), dimension(3), intent(in)	x_cart,
		integer, intent(in)	flags
	)

Gets the guassian process kernel for a signal and cartesian position.

This method is meant to be overwritten by a subclass method. As a result, returns the get_gp_x method of the subclass instance signal_class was constructed with.

See also

sxrem::sxrem_get_gp_x

intpol::intpol_get_gp_x

thomson::thomson_get_gp_x

ece::ece_get_gp_x

prior_gaussian::prior_gaussian_get_gp_x

Parameters

[in]	this	A signal_class instance.
[in]	a_model	A model_class instance.
[in]	x_cart	The cartesian position of to get the kernel at.
[in]	flags	State flags to send to the kernel.

Returns

Kernel value for position and the signal.

Definition at line 714 of file signal.f.

signal_get_header()

subroutine signal::signal_get_header	(	class (signal_class), intent(in)	this,
		character (len=data_name_length), dimension(7), intent(inout)	header
	)

Gets a discription of the model and model sigma array indices.

This method is meant to be overwritten by a subclass method. As a result, returns the get_header method of the subclass instance signal_class was constructed with.

See also

magnetic::magnetic_get_header

mse::mse_get_header

ece::ece_get_header

limiter::limiter_get_header

feedback::feedback_get_header

combination_get_header

gaussp_get_header

Parameters

[in]	this	A signal_class instance.
[in,out]	header	Buffer arrays to write header strings to.

Definition at line 591 of file signal.f.

signal_get_modeled_signal_cache()

real (rprec) function, dimension(4) signal::signal_get_modeled_signal_cache	(	class (signal_class), intent(inout)	this,
		type (model_class), pointer	a_model,
		real (rprec), dimension(:), intent(out)	sigma,
		logical, intent(in)	use_cache,
		real (rprec), dimension(4), intent(in)	last_value
	)

Calculates the modeled signal.

This method is checks if the cached vales should be used first otherwise calls the subclass method.

Parameters

[in,out]	this	A diagnostic_class instance.
[in]	a_model	A model instance.
[out]	sigma	The modeled sigma.
[in]	use_cache	If false calculate the signal and sigma and write them to the cache. Otherwise read from the cache without recalculating the signal.
[in]	last_value	Last good value in case the signal did not change.

Returns

The model value.

Definition at line 291 of file signal.f.

signal_get_modeled_signal_last()

real (rprec) function, dimension(4) signal::signal_get_modeled_signal_last	(	class (signal_class), intent(inout)	this,
		type (model_class), pointer	a_model,
		real (rprec), dimension(4), intent(out)	sigma,
		real (rprec), dimension(4), intent(in)	last_value
	)

Calculates the modeled signal.

This method is meant to be overwritten by a subclass method. As a result, returns the get_modeled_signal method of the subclass instance signal_class was constructed with unless the signal is read from the cached value. Read from the cached value when the signal doesn't need to be recalulated. If the signal is to be recalculated, write these to the cache.

See also

magnetic::magnetic_get_modeled_signal

sxrem::sxrem_get_modeled_signal

intpol::intpol_get_modeled_signal

thomson::thomson_te_get_modeled_signal

thomson::thomson_p_get_modeled_signal

thomson::thomson_ne_get_modeled_signal

extcurz::extcurz_get_modeled_signal

mse::mse_get_modeled_signal

ece::ece_get_modeled_signal

limiter::limiter_get_modeled_signal

prior_guassian::prior_guassian_get_modeled_signal

sxrem_ratio::sxrem_ratio_get_modeled_signal

combination::combination_get_modeled_signal

Parameters

[in,out]	this	A signal_class instance.
[in]	a_model	A model instance.
[out]	sigma	The modeled sigma.
[in]	last_value	Last good value in case the signal did not change.

Returns

The model value.

Definition at line 353 of file signal.f.

signal_get_observed_signal()

real (rprec) function signal::signal_get_observed_signal	(	class (signal_class), intent(in)	this,
		type (model_class), intent(in)	a_model
	)

Calculates the observed signal.

This method is meant to be overwritten by a subclass method. As a result, returns the get_modeled_signal method of the subclass instance signal_class was constructed with. All signals expect feed back signals use a fixed observed value.

See also

feedback::feedback_get_modeled_signal

Parameters

[in]	this	A diagnostic_class instance.
[in]	a_model	A model instance.

Returns

The observed value.

Definition at line 388 of file signal.f.

signal_get_sigma2()

real (rprec) function signal::signal_get_sigma2

(

class (signal_class), intent(in)

this

)

Calculates the total sigma^2 of a signal.

Gets the total sigma^2 of an individual signal. sigma^2 = sigma_o^2 + sigma_m^2

Parameters

[in]

this

A diagnostic_class instance.

Returns

The e value.

Note

this assumes that modeled sigma value has already been cached by a call to signal_get_modeled_signal

Definition at line 511 of file signal.f.

signal_get_type()

character (len=data_name_length) function signal::signal_get_type

(

class (signal_class), intent(in)

this

)

Gets a discription of the signal type.

This method is meant to be overwritten by a subclass method. As a result, returns the get_type method of the subclass instance signal_class was constructed with.

See also

magnetic::magnetic_get_type

sxrem::sxrem_get_type

intpol::intpol_get_type

thomson::thomson_get_type

extcurz::extcurz_get_type

mse::mse_get_type

ece::ece_get_type

limiter::limiter_get_type

prior_guassian::prior_guassian_get_type

sxrem_ratio::sxrem_ratio_get_type

combination::combination_get_type

Parameters

[in]

this

A signal_class instance.

Returns

A string describing the signal type.

Definition at line 555 of file signal.f.

signal_make_short_name()

character (len=data_short_name_length) function signal::signal_make_short_name	(	character (len=*), intent(in)	name,
		integer, intent(in)	index
	)

Generate a short name by appending an index.

Returns an appended string with no leading spaces. This check the number of digits in the index then chooses the appropriate formatting.

Parameters

[in]	name	A name of a signal to append to.
[in]	index	A index of a signal to append.

Returns

A string an appended string.

Definition at line 775 of file signal.f.

signal_scale_and_offset()

subroutine signal::signal_scale_and_offset	(	class (signal_class), intent(in)	this,
		type (model_class), intent(in)	a_model,
		real (rprec), intent(inout)	value
	)

Apply scale and offset to the value.

Modeled signals can contain an extra scaling and offset. Apply these to the value. These are obtained from the model.

Parameters

[in]	this	A intpol_class instance.
[in]	a_model	A model instance.
[in,out]	value	The value to scale and offset.

Definition at line 744 of file signal.f.

signal_sync_child()

subroutine signal::signal_sync_child	(	class (signal_class), intent(inout)	this,
		integer, intent(in)	index,
		integer, intent(in)	recon_comm
	)

Syncronize a child signal state to the parent.

Syncs data between a child and the parent process. If MPI support is not compiled in this subroutine reduces to a no op.

Parameters

[in,out]	this	A signal_class instance.
[in]	index	Reconstruction rank to sync.
[in]	recon_comm	MPI communicator for the reconstruction processes.

Definition at line 1031 of file signal.f.

signal_write()

subroutine signal::signal_write	(	class (signal_class), intent(inout)	this,
		integer, intent(in)	iou,
		integer, intent(in)	index,
		type (model_class), pointer	a_model
	)

Write out the signal information to an output file.

Signals information is formated as the index, signal type, s_name, g^2, weight, observed value, observed sigma, modeled value and modeled sigma.

Parameters

[in,out]	this	A signal_class instance.
[in]	iou	A input/output representing the file to write to.
[in]	index	A index of a signal.
[in,out]	a_model	The equilibrium model.

Note

This assumes that signals have already been calculated. The model isn't used by signal_get_g2 since it is reading from the cache but must be passed in anyway.

Definition at line 819 of file signal.f.

signal_write_header()

subroutine signal::signal_write_header	(	class (signal_class), intent(in)	this,
		integer, intent(in)	iou
	)

Write out the signal header information to an output file.

Signals information is formated as a the index, signal type, s_name, g^2, weight, observed value, observed sigma, modeled value and modeled sigma.

Parameters

[in]	this	A signal_class instance.
[in]	iou	A input/output representing the file to write to.

Note

This should only be called when the type of signal changes.

Definition at line 868 of file signal.f.

signal_write_step_data()

subroutine signal::signal_write_step_data	(	class (signal_class), intent(inout)	this,
		type (model_class), pointer	a_model,
		integer, intent(in)	result_ncid,
		integer, intent(in)	current_step,
		integer, intent(in)	index,
		integer, intent(in)	signal_model_value_id,
		integer, intent(in)	signal_sigma_value_id
	)

Write out the signal data for a step to the result netcdf file.

Writes out the observed, modeled and total sigma value of the signal.

Parameters

[in,out]	this	A diagnostic_class instance.
[in,out]	a_model	The equilibrium model.
[in]	result_ncid	A NetCDF id of the result file.
[in]	current_step	The surrent reconstruction step.
[in]	index	A index of a signal.
[in]	signal_model_value_id	NetCDF variable id of the model value.
[in]	signal_sigma_value_id	NetCDF variable id of the total sigma.

Note

This assumes that signals have already been calculated. The model isn't used by signal_get_g2 since it is reading from the cache but must be passed in anyway.

Definition at line 952 of file signal.f.

signals_write_auxiliary()

subroutine signal::signals_write_auxiliary	(	class (signal_class), intent(in)	this,
		integer, intent(in)	iou,
		integer, intent(in)	index,
		type (model_class), intent(in)	a_model
	)

Write out any auxiliary signal information to an output file.

This method is meant to be overwritten by a subclass method if need be. As a result, returns the write_auxiliary method of the subclass instance signal_class was constructed with. Otherwise uses the default. nothing is written.

See also

sxrem_ratio::sxrem_ratio_write_auxilary

prior_gaussian::prior_gaussian_write_auxiliary

combination::combination_write_auxiliary

Parameters

[in]	this	A signal_class instance.
[in]	iou	A input/output representing the file to write to.
[in]	index	A index of a signal.
[in]	a_model	The equilibrium model.

Definition at line 911 of file signal.f.