reconstruction Class Reference

Detailed Description

TokaMaker equilibrium reconstruction class.

Public Member Functions
	__del__ (self)
	Destroy reconstruction object.
	__init__ (self, tMaker_obj, in_filename='fit.in', out_filename='fit.out')
	Create equilibrium reconstruction object.
	add_flux_loop (self, loc, val, err)
	Add full poloidal flux loop constraint.
	add_Mirnov (self, loc, norm, val, err)
	Add Mirnov sensor constraint.
	add_pressure (self, loc, val, err)
	Add plasma pressure constraint.
	add_saddle (self, p1, p2, width, val, err)
	Add saddle loop constraint.
	destroy_constraints (self)
	Destroy constraints in TokaMaker.
	eval_error (self, vacuum=False, save_to_file=False)
	Evaluate error in current equilibrium for specified constraints without performing reconstruction.
	opt_get_dofs (self)
	Get DoF vector for general optimization routines using opt_error and opt_error_jacobian.
	plot_constraints (self, fig, ax, plot_flux_loops=True, plot_mirnovs=True, plot_saddles=True, plot_pressure=True, mirnov_scale=1.0, base_zorder=0)
	Plot constraint locations in R-Z plane for current constraints.
	plot_error (self, ax, error_mat=None, error_file=None, coil_ax=None, chi_ax=None)
	Plot constraint values and reconstructed signals for current equilibrium from reconstruction output file.
	read_fit_in (self)
	Read constraints from existing input file.
	reconstruct (self, vacuum=False, linearized_fit=False, maxits=100, eps=1.E-3, ftol=1.E-3, xtol=1.E-3, gtol=1.E-3)
	Reconstruct G-S equation with specified fitting constraints, profiles, etc.
	reset_constraints (self)
	Remove all current constraints.
	set_coil_currents (self, targets, errs)
	Set coil current constraints.
	set_DFlux (self, DFlux, err)
	Set diamagnetic flux constraint.
	set_Ip (self, Ip, err)
	Set plasma current constraint.
	setup_constraints (self)
	Set up constraints in TokaMaker for current equilibrium without performing reconstruction.
	setup_get_opt (self, fail_factor=1.E2, pp_target_weight=1.E5, opoint_target_weight=1.E2)
	Setup object for use with general optimization routines, such as scipy.optimize.minimize, which will call opt_error and, optionally, opt_error_jacobian with appropriate arguments.
	write_fit_in (self)
	Create reconstruction input file for specified constraints.

Static Public Member Functions
	opt_error (cofs, recon_obj, in_jac)
	Compute weighted constraint error for given DoF vector cofs and reconstruction object recon_obj.
	opt_error_jacobian (cofs, recon_obj, in_jac)
	Compute Jacobian matrix corresponding by differencing opt_error for given DoF vector cofs and reconstruction object recon_obj.

Public Attributes
	con_file = in_filename
	Name of constraint file (input for reconstruction).
	out_file = out_filename
	Name of reconstruction output file.
	settings = tokamaker_recon_settings()
	Reconstruction specific settings object.

Protected Attributes
list	_coil_current_cons = []
	Coil current constraints.
	_Dflux_con = None
	Diamagnetic flux constraint.
	_EQ_center = self._tMaker_obj.copy_eq()
int	_err_min = 1.E99
	_fail_error = error0*fail_factor
list	_flux_loops = []
	Flux loop constraints.
	_Ip_con = None
	Plasma current constraint.
list	_mirnovs = []
	Mirnov sensor constraints.
int	_ncons = 0
	Number of constraints in underlying Fortran object.
list	_pressure_cons = []
	Plasma pressure constraints.
list	_saddles = []
	Saddle loop constraints.
	_tMaker_obj = tMaker_obj
	Grad-Shafranov object for reconstruction.

Constructor & Destructor Documentation

__init__()

__init__	(		self,
			tMaker_obj,
			in_filename = 'fit.in',
			out_filename = 'fit.out' )

Create equilibrium reconstruction object.

Parameters

tMaker_obj	TokaMaker object used for computing G-S equilibria
in_filename	Filename to use for reconstruction input
out_filename	Filename to use for reconstruction outputs

__del__()

__del__

(

self

)

Destroy reconstruction object.

Member Function Documentation

add_flux_loop()

add_flux_loop	(		self,
			loc,
			val,
			err )

Add full poloidal flux loop constraint.

Parameters

loc	Location of flux loop in R-Z plane [2]
val	Value of flux loop constraint
err	Error in constraint

add_Mirnov()

add_Mirnov	(		self,
			loc,
			norm,
			val,
			err )

Add Mirnov sensor constraint.

Parameters

loc	Location of Mirnov in R-Z plane [2]
norm	Unit normal in R-Z plane [2]
val	Value of Mirnov constraint
err	Error in constraint

add_pressure()

add_pressure	(		self,
			loc,
			val,
			err )

Add plasma pressure constraint.

Parameters

loc	Location of measurement in R-Z plane [2]
val	Value of pressure constraint
err	Error in constraint

add_saddle()

add_saddle	(		self,
			p1,
			p2,
			width,
			val,
			err )

Add saddle loop constraint.

Parameters

p1	Location of first toroidal saddle leg in R-Z plane [2]
p2	Location of second toroidal saddle leg in R-Z plane [2]
width	Toroidal extent in radians
val	Value of saddle loop constraint
err	Error in constraint

destroy_constraints()

destroy_constraints

(

self

)

Destroy constraints in TokaMaker.

eval_error()

eval_error	(		self,
			vacuum = False,
			save_to_file = False )

Evaluate error in current equilibrium for specified constraints without performing reconstruction.

Parameters

vacuum	Perform vacuum reconstruction
save_to_file	Save error matrix to file specified in settings.outfile

Returns

Error matrix if save_to_file=False, otherwise None (error matrix will be saved to file)

opt_error()

opt_error ( cofs, recon_obj, in_jac )

static

Compute weighted constraint error for given DoF vector cofs and reconstruction object recon_obj.

This function is designed to be called by general optimization routines, such as scipy.optimize.minimize.

Parameters

cofs	DoF vector
recon_obj	TokaMaker.reconstruction.reconstruction object to be used
in_jac	Flag indicating whether this function is being called from a Jacobian calculation to make print statements during finite differencing

Returns

Weighted error vector for current DoFs

opt_error_jacobian()

opt_error_jacobian ( cofs, recon_obj, in_jac )

static

Compute Jacobian matrix corresponding by differencing opt_error for given DoF vector cofs and reconstruction object recon_obj.

Parameters

cofs	DoF vector
recon_obj	TokaMaker.reconstruction.reconstruction object to be used
in_jac	Ignored flag just to match signature with opt_error

Returns

Jacobian matrix for current DoFs

opt_get_dofs()

opt_get_dofs

(

self

)

Get DoF vector for general optimization routines using opt_error and opt_error_jacobian.

Returns

DoF vector

plot_constraints()

plot_constraints	(		self,
			fig,
			ax,
			plot_flux_loops = True,
			plot_mirnovs = True,
			plot_saddles = True,
			plot_pressure = True,
			mirnov_scale = 1.0,
			base_zorder = 0 )

Plot constraint locations in R-Z plane for current constraints.

Parameters

fig	Matplotlib figure to plot on
ax	Matplotlib axis to plot on
plot_flux_loops	Whether to plot flux loop constraints
plot_mirnovs	Whether to plot Mirnov sensor constraints
plot_saddles	Whether to plot saddle loop constraints
plot_pressure	Whether to plot plasma pressure constraints
mirnov_scale	Scaling factor for Mirnov sensor quiver plot (default 1.0, smaller values increase arrow size)
base_zorder	Base z-order for plotting constraints (default 0.0, higher values will plot on top of lower values)

plot_error()

plot_error	(		self,
			ax,
			error_mat = None,
			error_file = None,
			coil_ax = None,
			chi_ax = None )

Plot constraint values and reconstructed signals for current equilibrium from reconstruction output file.

Parameters

ax	Matplotlib axis to plot on
error_mat	Error matrix, if None eval_error will be called
error_file	File containing error data
coil_ax	Matplotlib axis to plot coil current constraints on
chi_ax	Matplotlib axis to plot chi-squared values on

read_fit_in()

read_fit_in

(

self

)

Read constraints from existing input file.

reconstruct()

reconstruct	(		self,
			vacuum = False,
			linearized_fit = False,
			maxits = 100,
			eps = 1.E-3,
			ftol = 1.E-3,
			xtol = 1.E-3,
			gtol = 1.E-3 )

Reconstruct G-S equation with specified fitting constraints, profiles, etc.

Parameters

vacuum	Perform vacuum reconstruction
linearized_fit	Use linearized solve for suitable terms
maxits	Maximum number of iterations
eps	Epsilong factor for finite difference derivative calculations
ftol	Stopping condition: termination occurs when both the actual and predicted relative reductions in the sum of squares are at most ftol
xtol	Stopping condition: termination occurs when the relative error between two consecutive iterates is at most xtol
gtol	Stopping condition: termination occurs when the cosine of the angle between fvec and any column of the jacobian is at most gtol in absolute value

Returns

Error flag

reset_constraints()

reset_constraints

(

self

)

Remove all current constraints.

set_coil_currents()

set_coil_currents	(		self,
			targets,
			errs )

Set coil current constraints.

Parameters

targets	Dictionary of coil current targets
errs	Dictionary of coil current constraint errors

set_DFlux()

set_DFlux	(		self,
			DFlux,
			err )

Set diamagnetic flux constraint.

Parameters

DFlux	Value of diamagnetic flux constraint
err	Error in constraint

set_Ip()

set_Ip	(		self,
			Ip,
			err )

Set plasma current constraint.

Parameters

Ip	Value of plasma current constraint
err	Error in constraint

setup_constraints()

setup_constraints

(

self

)

Set up constraints in TokaMaker for current equilibrium without performing reconstruction.

Returns

Error flag

setup_get_opt()

setup_get_opt	(		self,
			fail_factor = 1.E2,
			pp_target_weight = 1.E5,
			opoint_target_weight = 1.E2 )

Setup object for use with general optimization routines, such as scipy.optimize.minimize, which will call opt_error and, optionally, opt_error_jacobian with appropriate arguments.

This function should be called after setting up constraints with setup_constraints and before calling optimization routines.

Parameters

fail_factor	Factor to multiply initial error by for failed solves
pp_target_weight	Weight for global P' targets when treated as soft constraints and fit_Pscale is True
opoint_target_weight	Weight for R0 and Z0 targets when treated as soft constraints and fitR0 or fitZ0 is True

Returns

Initial DoF vector for optimization, error vector for initial DoFs

write_fit_in()

write_fit_in

(

self

)

Create reconstruction input file for specified constraints.

Member Data Documentation

_coil_current_cons

_coil_current_cons = []

protected

Coil current constraints.

_Dflux_con

_Dflux_con = None

protected

Diamagnetic flux constraint.

_EQ_center

_EQ_center = self._tMaker_obj.copy_eq()

protected

_err_min

int _err_min = 1.E99

protected

_fail_error

_fail_error = error0*fail_factor

protected

_flux_loops

_flux_loops = []

protected

Flux loop constraints.

_Ip_con

_Ip_con = None

protected

Plasma current constraint.

_mirnovs

_mirnovs = []

protected

Mirnov sensor constraints.

_ncons

int _ncons = 0

protected

Number of constraints in underlying Fortran object.

_pressure_cons

_pressure_cons = []

protected

Plasma pressure constraints.

_saddles

_saddles = []

protected

Saddle loop constraints.

_tMaker_obj

_tMaker_obj = tMaker_obj

protected

Grad-Shafranov object for reconstruction.

con_file

con_file = in_filename

Name of constraint file (input for reconstruction).

out_file

out_file = out_filename

Name of reconstruction output file.

settings

settings = tokamaker_recon_settings()

Reconstruction specific settings object.

---

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

• python/OpenFUSIONToolkit/TokaMaker/reconstruction.py