Data Structures
struct	atm_cfg_t

struct	atmr_cfg_t

struct	aper_cfg_t

struct	llt_cfg_t

struct	dither_cfg_t

struct	pywfs_cfg_t

struct	powfs_cfg_t

struct	wfs_cfg_t

struct	dm_cfg_t

struct	evl_cfg_t

struct	tomo_cfg_t

struct	fit_cfg_t

struct	lsr_cfg_t

struct	recon_cfg_t

struct	sim_cfg_t

struct	ncpa_cfg_t

struct	cn2est_cfg_t

struct	plot_cfg_t

struct	dbg_cfg_t

struct	gpu_cfg_t

struct	moao_cfg_t

struct	load_cfg_t

struct	save_cfg_t

struct	dist_cfg_t

struct	parms_t

struct	arg_t

Macros
#define	MAX_AMPS 5

Enumerations
enum	T_TYPE { T_PLOC =0, T_ALOC, T_XLOC, T_ATM }

enum	WFS_TYPE { WFS_SH =0, WFS_PY =1, WFS_TOT }

enum	GTYPE { GTYPE_G =0, GTYPE_Z =1, GTYPE_TOT }

enum	PTYPE { PTYPE_MF =1, PTYPE_COG =2, PTYPE_MAP =3, PTYPE_CORR =4, PTYPE_CORRS =5, PTYPE_TOT }

enum	NCPA_METHOD { NCPA_G =1, NCPA_I0 =2, NCPA_TOT }

enum	RECON_ALG { RECON_MVR =0, RECON_LSR =1, RECON_TOT }

enum	ALG_TOMO_FIT { ALG_CBS =0, ALG_CG =1, ALG_SVD =2, ALG_BGS =3, ALG_TOT }

enum	PCG_ALG { PCG_NONE =0, PCG_FD =1, PCG_TOT }

Functions
parms_t *	setup_parms (const char main, const char extra, int override)

void	setup_parms_gpu (parms_t parms, int gpus, int ngpu)

void	free_parms (parms_t *parms)

Detailed Description

Configuration parameters that remain constant during simulation.

Data Structure Documentation

◆ atm_cfg_t

struct atm_cfg_t

contains input parameters for the atmospheric turbulence.

Collaboration diagram for atm_cfg_t:

Data Fields
real	r0z	r0 at zenith
real	r0	derived from r0z for zenith angle za
real	dx	sampling of turbulence screens
real	hmax	maximum in ht
dmat *	L0	outer scale. One number or one per layer
dmat *	r0logpsds	[alpha beta]: temporal PSD of log(r0) is beta*f^alpha. f is in hz.
dmat *	r0logpsdt	[alpha beta]: spatial PSD of log(r0) is beta*f^alpha. f is in m.
dmat *	ht	height of each layer
dmat *	wt	weight of each layer (relative strength of $C_n^2$)
dmat *	ws	wind speed of each layer
dmat *	wddeg	wind direction of each layer
dmat *	size	size of atm in meter, [0 0]: automatic
lmat *	overx	maximum pixel distance in x direction the beam can be without wrapping
lmat *	overy	maximum pixel distance in y direction the beam can be without wrapping
int	nps	number of phase screens
int	wdrand	randomize wind direction
int	iground	index into the ground layer
lmat *	ipsr	corresponding reconstruction layer
int	nx	turbulence screen size along x
int	ny	turbulence screen size along y
lmat *	nxn	minimum turbulence screen size along to cover meta pupil
int	nxnmax	max of nxn
int	method	0: FFT Von Karman. 1: FFT Biharmonic. 2: Fractal method.
int	frozenflow	frozen flow. automatic if closeloop=1
int	ninit	Initial size of the screen in fractal method. >=2
int	share	0: disable sharing of atmosphere using file backend
int	r0evolve	Evolve r0 according to r0logpsd

◆ atmr_cfg_t

struct atmr_cfg_t

contains input parameters for the atmospheric reconstruction.

Collaboration diagram for atmr_cfg_t:

Data Fields
real	r0z	r0 at zenith
real	r0	derived from r0z for zenith angle za
real	L0	outer scale
real	hs	height of the high order guide star. derived
real	hmax	maximum of ht
dmat *	ht	height of each layer
dmat *	wt	weight of each layer (relative strength of $C_n^2$)
real	dx	baseline sampling (when os=1). matches to high order wfs.
lmat *	indps	Mapping atmr.ps to atm.ps
lmat *	os	over sampling factor of xloc over actuator spacing
int	nps	number of phase screens

◆ aper_cfg_t

struct aper_cfg_t

contains input parameters about the aperture, like the diameter, amplitude map, etc

Collaboration diagram for aper_cfg_t:

Data Fields
real	d	Telescope aperture diameter
real	din	Telescope inner blocking diameter
real	rot	pupil rotation
dmat *	misreg	Misregistration of the telescope pupil. 2x1 The intersection between the misregistered pupil (aper.d or fnamp) and a centered aper.d defines the actual pupil.
char *	fnamp
int	fnampuser	amplitude maps. expected to be square or rectangular mxn, with 0 at [m/2,n/2] (count from 0) User provided amplitude map (not default)
char *	pupmask	The pupil cold stop

◆ llt_cfg_t

struct llt_cfg_t

contains input parameters for laser launch telescope

Collaboration diagram for llt_cfg_t:

Data Fields
real	d	LLT clear aperture diameter
real	widthp	Gaussian beam width percentage of d
real	focus	RMS focus error in nm of LLT.
real	ttrat	Ratio of uplink jitter to science jitter due to M2 windshake.
real	fcfsm	corner frequency for offloading FSM to a common path pointing mirror in LLT. 0: disabled
real	dhs	Spacing of sublayers to simulate for LGS
char *	ttpsd	PSD of uplink beam jitter
char *	fnrange	File contains range to sodium layer
char *	fnprof	File contains sodium profile
char *	fnprep	File contains sodium profiled used for computing i0. if NULL: equal to fnprof
char *	fnsurf	Pupil Surface OPD error
char *	fnamp	Pupil amplitude map. overrides widthp
dmat *	ox	location x of LLT center wrt telescope aperture center
dmat *	oy	see ox.
dmat *	misreg
int	ttfr	Remove piston/tip/tilt and focus (if = 2) from ncpa
int	colprep	starting column to use in fn for ETF in preparation of matched filter
int	colsim	starting column to use in fn for ETF in simulation
int	coldtrat	change to next sodium profile during simulation every coldtrat time step
int	nhs	Number of sublayer to simulate for LGS
int	na_smooth	1: smooth sodium profile to coarser grid before computing etf
int	na_interp	1: Interpolate sodium profile and use FFT to build etf. 0: direct sum, slow
int	na_fit_maxit	Number of iterations. 0: auto, 1 for CMF, 3 for COG. see wfsgrad.c
real	na_fit_svdthres	threshold for SVD inverse in sodium fitting.
real	na_fit_dh	sampling in height in sodium fitting
real	na_thres	altitude error threshold to move trombone, in unit of meter.
lmat *	i	Index into llt for this iwfs.
int	nllt	number of launch telescopes in this powfs
real	epfsm	Integrator gain for LLT FSM offloading

◆ dither_cfg_t

struct dither_cfg_t

parameters for dithering

Collaboration diagram for dither_cfg_t:

Data Fields
int	mode
real	amp	Dither amplitude.
real	gpll	Gain of phase locked loop
real	gog	Gain for updating optical gain for cog
real	gdrift	Gain for drift control
real	glpf	LPF gain for i0,gx,gy update (matched filter)
int	npoint	Number of points in each dither peroid (4)
int	pllskip	Skip WFS frames for uplink loop to stable
int	pllrat	Number of WFS frames for updating PLL.
int	ogskip	Number of WFS frames to skip before computing averaged images
int	ograt	Number of WFS frames to update pixel processing algorithm (MF/CoG)
int	ogsingle	*Force using single gain update (when dither==1 for SHWFS)

◆ pywfs_cfg_t

struct pywfs_cfg_t

Pyramid WFS configurations. //Todo: move to pywfs.h after pywfs.h is moved to lib folder

Collaboration diagram for pywfs_cfg_t:

Data Fields
int	nside	Number of sides. can be 4 or 3. 4-side is traditional pyramid while 3-side is much easier to make.
int	raw	1: use normalized ints of each sub-pupil as gradient. 0: use difference between sub-pupils
int	ng	Number of gradients per 'subaperture'.
int	modulpos	Number of positions per modulation cycle
int	modulpos_i	For testing: Index of modulate position to use between 0 and modulpos
int	modulring	Number of rings within the maximum radius to modulate
int	sigmatch	Scale gradients by matching intensity (1: locally, 2: globally).
real	siglev	Nominal siglev per subaperture
real	hs	Height of guide star
real	hc	Conjugation height of WFS pupil
real	poke	How much to poke for mkg
real	modulate	Amount of modulation in radian
dmat *	psx	pyramid WFS pupil shift along x (in pixel). pupil ordering: -x+y, +x+y, -x-y, +x-y.
dmat *	psy	pyramid WFS pupil shift along y (in pixel).
real	flate	pyramid flat edge angular width
real	flatv	pyramid flat vertex angular width
real	pupelong	pyramid pupil (detector) elongation ratio (long axis / short axis).

◆ powfs_cfg_t

struct powfs_cfg_t

contains input parameters for each type of wfs (powfs).

Collaboration diagram for powfs_cfg_t:

Data Fields
dmat *	wvl	list of wavelength in ascending order.
dmat *	wvlwts	weights for each wavelength. can be overriden by wfs.wvlwts.
char *	saloc	saloc override file
real	misregx	offset of saloc from pupil illumination in unit of dsa
real	misregy	offset of saloc from pupil illumination in unit of dsa
char *	amp	amplitude override file
char *	piinfile	input averaged pixel intensities for matched filter. NULL to disable
char *	sninfile	Speckle noisy input file. NULL to disable. not used
real	hs	height of guide star
real	hc	conjugation height of WFS pupil
real	saat	subaperture area (normalized) threshold to drop subaperture.
real	safill2d	subaperture lenslet throughgput. value is used to alter amplitude map
real	saspherical	Subaperture spherical aberration in nm RMS at best focus.
real	safocuspv	Subaperture focus error in nm p/v
char *	neareconfile	File contains noise equivalent angle in radian^2. Contains cell array of nwfsx1.
real	nearecon	NEA used in reconstruction in milli-arcsecond, sim.dtref integration time. Will be scaled by powfs.dtrat and subaperture area before use.
real	neasim	NEA used in simulation. -1 to use nearecon
char *	neasimfile	read NEA used in simulation from file. Defined at sim.dt sampling rate, in radian. neasim must be -1
real	neaextra	Extra NEA to add in quadrature to the NEA determined by matched filter or CoG
real	neamin	Minimum NEA to limit the NEA determined by matched filter or CoG
real	bkgrnd	background in electron per pixel per LGS frame
real	bkgrndc	How much of the background in bkgrnd can be calibrated out. depends on variability.
char *	bkgrndfn	file contains sky background/rayleigh scatter input for each subaperture in each wfs.
char *	bkgrndfnc	How much of the background in bkgrndfn can be calibrated out. depends on variability.
dmat *	qe	File containing matrix of pixpsax*pixpsay specifying QE of each pixel. To simulate PCCD non uniform response
real	rne	read out noise in electron per pixel per frame
real	pixblur	pixel bluring due to leakage. relative to pixel size.
real	dsa	Size of subaperture in 1 dimension
real	dx	sampling of opd points in each subaperture. usually matches atmosphere sampling for LGS. may be coraser for NGS.
real	pixtheta	size of pixel pitch along x/y or azimuthal if radial ccd. Converted to radian from user input
real	radpixtheta	size of pixel pitch along radial direction. -1 for square pixel
real	fieldstop	size of field stop in arcsec.
real	pixoffx	offset of image center from center of detector
real	pixoffy	see pixoffx
real	sigscale	scale the signal level for simulation.
real	siglev	signal level at dtrat=1. will be override by wfs.siglev is specified.
dmat *	siglevs	in array format. 1x1 or nwfsx1, scaled from wfs.siglev by dtrat.
real	sigrecon	signal level for NEA computation
struct llt_cfg_t *	llt	configuration for LLT
char *	fnllt	filename of LLT configuration. empty means no llt.
pywfs_cfg_t *	pycfg	Set only for Pyramid WFS.
char *	pywfs	Pyramid WFS configuration
int	type	WFS type: 0: SHWFS, 1:Pyramid WFS
int	step	frame to start using WFS
int	trs	tip/tilt removal flag. True for LGS, False for NGS
int	frs
int	lo	Global focus removal flag. Optional for LGS, False for NGS whether this is a low order wfs. False for LGS, True for NGS
int	skip	skip in high order tomography, for split tomo (derived parameter)
int	psol	Compute pseudo open loop gradients (derived parameter)
lmat *	wfs	array of wfs belongs to this powfs
lmat *	wfsr	array of reconstruction wfs belongs to this powfs
lmat *	wfsind	wfsind[iwfs] gives the index of the wfs in this powfs group
int	nwfs	number of wfs belonging to this powfs
int	nwfsr	number of wfs for reconstruction belonging to this powfs
int	neaphy	use nea from physical optical precomputation in geometric simulations.
int	phyusenea	force using supplied noise equivalent angle in physical optics simulations
int	order	order of wavefront sensing along one dimension.
int	pixpsa	number of detector pixels along x/y (or azimuthal if radial CCD).
int	radpix	number of detector pixels along radial direction if radial CCD
int	radgx	1: gx/gy is along R/A coordinate. Only valid if radpix is set
int	notf	PSF is extended to this size before FFT into OTF. 0 for automatic
int	embfac	Embed subaperture atm OPD before fft. set to 2.
int	nwvl	Number of wavelength. 1 for LGS, 2 for NGS J+H sensing.
int	gtype_sim	wfs type if not using physical optics in reconstruction. 0: geometric 1: ztilt.
int	gtype_recon	wfs type if not using physical optics in simulation. 0: geometric 1: ztilt.
int	phytype_recon	physical optics type for reconstruction. 1: mtch, 2: tcog, 3: MAP
int	phytype_sim	physical optics type for simulation. -1 to follow phytype_recon
int	phytype_sim1	Save phytype_sim initial value
int	phytype_sim2	physical optics type after dithering update. -1 to follow phytype_sim
int	phystep	frame to start using physical optics. 0: means from frame 0. >0: need to compute GS0 to calculate geometric optics -1: never, doesn't need to compute DTF
int	usephy	whether physical optics is used at all during simulation.(derived parameter)
real	r0	Fried parameter for matched filter generation. Uses atm.r0, atm.L0 is not set
real	L0	Outerscale for matched filter generation. Uses atm.r0, atm.L0 is not set
int	mtchcr	use constrained matched filter (0: disable, 1: both axis. 2: radial/x only, 3: az/y only)
int	mtchcpl	use coupling between r/a measure error. useful for LGS with x-y ccd.
int	mtchstc	shift peak in the time averaged short exposure PSF to center using fft.
int	sigmatch	scale subaperture image to have the same intensity as i0. Keep false.
int	mtchadp	Using adaptive matched filter. When the number of pixels in the image brighter than half maximum is more than this value, use constraint. introduced on 2011-02-21.
int	mtchfft	Compute gx, gy using i0 with FFT derivative instead of PSF.
real	cogthres	CoG threshold, relative to max(im)
real	cogoff	CoG offset to remove, relative to max(im).
dmat *	ncpa	Description of NCPA; 2xn; first row is rms in meter, second row is zernike mode or negative for power law.
int	needGS0	need to compute GS0 (derived parameter)
int	noisy	noisy or not during simulation
int	ncpa_method	Method to correct ncpa. 0: do nothing. 1: derive gradient offset from OPD 2: derive gradient offset(CoG) or matched filter from i0 with bias OPD
int	pistatout	output time averaged short exposure image. 1: shift to center, 2: do not shift
int	pistatstart	time step to compute pistatout
int	pistatstc	1: shift to center using fft method. 0: use geometric gradients.
int	psfout	output time history of low order wfs PSF. never do this for LGS.
int	dtrat	ratio of sample period over sim.dt. Note that sim.dt is assumed to be the fast, high-order loop (e.g. LGS) and the slow or low-order TT WFS cannot run faster. powfs.dtrat must be an integer value.
int	idtrat	Index of dtrat into parms->sim.dtrats
int	i0scale	scale i0 to matched subaperture area.
int	moao	index into MOAO struct. -1: no moao
int	i0save	Save time averaged subaperture images.
char *	i0load	load i0,gx,gy from this folder.
real	gradscale	Scale CL gradients. For testing
int	dither	Turn on/off dithering to update centroid gain or matched filter
int	dither_mmd	Enable multi-mode dithering. Only effective when recon.modal is set
real	dither_amp	Dither amplitude.
real	dither_gpll	Gain of phase locked loop
real	dither_gog	Gain for updating optical gain for cog
real	dither_gdrift	Gain for drift control
real	dither_glpf	LPF gain for i0,gx,gy update (matched filter)
int	dither_npoint	Number of points in each dither peroid (4)
int	dither_pllskip	Skip WFS frames for uplink loop to stable
int	dither_pllrat	Number of WFS frames for updating PLL.
int	dither_ogskip	Number of WFS frames to skip before computing averaged images
int	dither_ograt	Number of WFS frames to update pixel processing algorithm (MF/CoG)
int	dither_ogsingle	*Force using single gain update (when dither==1 for SHWFS)
int	zoomshare	1: All LGS share the same trombone
real	zoomgain	gain of the trombone controller
real	zoomgain_drift	gain for the trombone controller with i0 drift input
int	zoomset	Set zoom position from the beginning
int	ng	number of gradients per subaperture. 2 for SHWFS. >2 for raw PWFS
real	apfsm	servo coefficient for for LGS uplink pointing loop.
real	epfsm	error gain for uplink pointing
real	alfsm	Additional latency (*sim.dt) of the uplink loop
real	zetafsm	Damping of FSM modeled as second harmonic oscillater (SHO).
real	f0fsm	Resonance frequency of FSM (SHO). 0: infinite.
int	idealfsm	ideal compensation for uplink pointing
int	commonfsm	Make FSM common for each powfs (LLT). Keep at 0.

◆ wfs_cfg_t

struct wfs_cfg_t

contains input parmaeters for each wfs

Collaboration diagram for wfs_cfg_t:

Data Fields
dmat *	wvlwts	Weights of signal value for each wavelength. if not specified in config, will use powfs.wvlwts
dmat *	sabad	coordinate of bad subaperture due to bad detector or lenslet array.
real	thetax	x direction
real	thetay	y direction
real	misreg_x	misregistration wrt telescope pupil. This is pure shift extracted from recon.distortion_tel2wfs.
real	misreg_y	misregistration wrt telescope pupil. This is pure shift extracted from recon.distortion_tel2wfs.
real	misreg_r	misregistration wrt telescope pupil. This is pure rotation extracted from recon.distortion_tel2wfs.
real	hc	conjugation height of WFS pupil is wfs.hc=powfs.hc+wfs.delta_hc (input)
real	hs
real	siglev	height of star is wfs.hs=powfs.hs+wfs.delta_hs (input) Total signal value for all wavelength. if not specified in config, will use powfs.siglev
real	sigsim	Signal value used for simulation. (derived parameter)
real	fitwt	Include wfs in fitting directions if corresponding wfs[iwfs].fitwt is greater than 0
int	powfs	powfs type

◆ dm_cfg_t

struct dm_cfg_t

contains input parameters for each deformable mirror.

Collaboration diagram for dm_cfg_t:

Data Fields
real	guard	extra DM actuator rings outside of aper.d
dmat *	stroke	Stroke of DM (surface). OPD goes to $\pm$ stroke. nactx2 array: min and max per actuator$
real	iastroke	Inter actuator stroke (surface)
dcell *	strokescale	describes polynomials that convert opd to voltage (first cell), and voltage to opd (second cell). The two operations has to be strict inverse of each other
real	dratio	telescope diameter to DM diameter ratio (beam angle magnification factor)
real	vmisreg	vertical misregistration
real	ht	height conjugation range
real	dx	actuator separation along x (derived from order)
real	ar	[in] aspect ratio: dy/dx
real	dy	actuator separation along y (derived from dx and ar
real	offset	Center-most actuator offset from origin =0 means there is a act on center. 1/2 means no one in the center.
real	iac	If !=0: use cubic influence function with this Inter-Actuator Coupling coefficient.
real	histbin	The bin width for histogram.
int	histn	Number of bins in histogram.
int	hist	Compute histogram of commands of each actuator
int	order	Nominal order of the DM within telescope clear subaperture
int	isground	Is this DM the ground DM (derived)
dmat *	actfloat	floating actuators. nx2 coordinate
dmat *	actstuck	stuck actuators. nx2 coordinate.
real	hyst	The hysteresis amount (ratio)
real	hyst_alpha	The DM hysteresis model alpha parameter
real	hyst_stroke	The surface stroke that the hysteresis is measured at

◆ evl_cfg_t

struct evl_cfg_t

contarins input parameters all evaluation directions.

Collaboration diagram for evl_cfg_t:

Data Fields
dmat *	thetax	x Coordinate of evaluation directions
dmat *	thetay	y Coordinate of evaluation directions
dmat *	wt	weight of each direction
dmat *	wvl	wavelength for PSF and strehl computation
dmat *	hs	height of each science object
real	dx	sampling of aperture for evaluation
int	nwvl	Number of wavelength
lmat *	psf	1: participate in psf evaluation.
lmat *	psfr	1: participate in psf reconstruction telemetry
int	npsf	how many directions we compute psf for
int	rmax	max radial mode for performance evaluation. 0: piston only 1: piston/tip/tilt.
int	nmod	Number of modes. derived from rmax. (nmax+1)*(nmax+2)/2
int	psfol	compute Open loop PSF. 1: on axis only. 2: all directions and average them.
int	psfhist	output history of the psf (a lot of storage)
int	psfmean	output time averaged psf
int	cov	save covairance of science OPD ,every this time step, for directions where evl.psf is 1
int	opdmean	save science OPD time average every `evlopdmean` time steps
lmat *	pttr	remove p/t/t from psf. 1 number for each evl.
int	psfisim	time step to start psfmean.
lmat *	psfsize	save this number of pixels of the center of the psf. 1 number for each wvl.
lmat *	psfgridsize	grid size for FFT to generate PSF. Becareful about FFT speed and enough padding. Determines the sampling of the generated PSF. 0 or negative for automatic. 1 number for each wvl.
int	nevl	Number of evaluation directions. (derived)
int	tomo	evaluate tomography performance.
int	indoa	index of the on axis evluation point.
int	moao	index into MOAO struct. -1: no MOAO

◆ tomo_cfg_t

struct tomo_cfg_t

contains input parameters for wavefront tomography.

Collaboration diagram for tomo_cfg_t:

Data Fields
real	tikcr	tikhonov regularization.
real	minwt	minimum layer weight allowed. if less than this will force to this.
real	iac	!=0: use cubic influence function with this Inter-Actuator Coupling coefficient.
real	cxxscale	scale the Cxx^-1 term.
real	svdthres	Threshold in SVD inversion
int	square	use square/rectangular grid instead of tighter irregular grid
int	cone	use cone coordinate in xloc: keep true
int	cxxalg	method to compute Cxx^-1. 0: bihormonic approx. 1: inverse psd. 2: fractal
int	guard	guard rings of reconstruction grid xloc
int	pos	over sampling factor of ploc over actuator spacing
int	nxbase	Each layer xloc grid size is tomo.os*tomo.nxbase is not zero. same for ploc.
int	piston_cr	single point piston constraint.
int	ahst_wt	Weight used to compute low order model removal in AHST 1: remove effect on NGS WFS (not good if WFS is outside of science FoV) 2: remove effect on Science 3: Identity weighting (bad)
int	ahst_idealngs	ideal correction on NGS modes. For skycoverage preprocessing.
int	ahst_focus	1: Make magnification mode free of focus in science (only effective when sim.mffocus=1
int	alg	Tomography algorithm to solve the linear equation. Todo: implement BGS, MG 0: Cholesky direct solve for the large matrix. (CBS) 1: CG or PCG. 2: SVD or EVD: Eigen value decompsition
int	bgs	1: use BGS, block Gaussia Seidel then use alg to solve each block.
int	precond	Tomography preconditioner. 0: No preconditioner. (CG) 1: Fourier Domain Preconditioner. (FDPCG)
int	maxit	max iterations. Usually 30 for CG, 3 for FDPCG in closed loop warm restart. x10 in open loop
int	cgwarm	Warm restart in CG.
int	assemble	force assemble tomography matrix in CG
int	predict	test predictive control.
int	ninit	like atm.ninit, the initial screen to generate from covariance directly
int	splitlrt	1: low rank terms also in LHS.

◆ fit_cfg_t

struct fit_cfg_t

contains input parameters for deformable mirror fitting.

Collaboration diagram for fit_cfg_t:

Data Fields
dmat *	thetax	x Coordinate of DM fitting directions.
dmat *	thetay	y Coordinate of DM fitting directions.
dmat *	wt	weight of each direction
dmat *	hs	height of target in each direction
real	tikcr	tikhonov regularization
real	svdthres	Threshold in SVD inversion
real	actthres	Threshold for slaving value of weakly coupled actuators
real	actthres2	Threshold for reducing jump across weakly coupled actuators
int	actslave	Enable slaving for non-active actuators. Useful in CBS method
int	actextrap	extrapolate actuator results to non-active actuators.
int	nfit	Number of DM fit directions
int	lrt_piston	Piston constraint low rank term in fit coefficient matrix
int	lrt_tt	differential tip/tilt constraint on two DMs or tt on upper dms.
int	alg	Fitting algorithm to solve the linear equation. 0: Cholesky direct solve for the large matrix. (CBS) 1: CG or PCG. 2: SVD or EVD: Eigen value decompsition
int	bgs	1: use BGS, block Gaussia Seidel then use alg to solve each block.
int	precond	Preconditioner. Not available.
int	maxit	max iterations. Usually 4 for CG
int	square	using square grid on DM and ploc.
int	assemble	force assemble fit matrix in CG
int	pos	over sampling of floc over aloc. for fitting. normally equal to tomo.pos
int	indoa	Index of on axis point.
int	cachedm	Cache DM command in intermediate plane
int	cachex	Cache X (xloc) in intermediate plane
int	cgwarm	Warm restart in CG.

◆ lsr_cfg_t

struct lsr_cfg_t

contains input parameters for the least square reconstructor.

Collaboration diagram for lsr_cfg_t:

Data Fields
real	tikcr	tikhonov regularization
real	svdthres	Threshold in SVD inversion
real	actthres	Threshold for slaving value of weakly coupled actuators
real	actthres2	Threshold for reducing jump across weakly coupled actuators
char *	fnreg	File containing a regularization term to add to LL.M
int	actextrap	extrapolate actuator results to non-active actuators .
int	actslave	Enable slaving for non-active actuators. Useful in CBS method
int	splitlrt	1: low rank terms also in LHS.
int	bgs	1: use BGS, block Gaussia Seidel then use alg to solve each block.
int	alg	algorithm to solve the linear equation. 0: Cholesky direct solve for the large matrix. (CBS) 1: CG or PCG. 2: SVD or EVD: Eigen value decompsition
int	maxit	max iterations. Usually 30 for CG
int	cgwarm	Warm restart in CG.

◆ recon_cfg_t

struct recon_cfg_t

contains input parameters for wavefront reconstruction.

Collaboration diagram for recon_cfg_t:

Data Fields
real	psdservo_gain	Gain used to update servo parameter
real	poke	How much WFE (meter) to apply to OPD for computing experimental interaction matrix
int	alg	algorithm for reconstruction. 0: MVR. 1: LSR. moved from sim.recon
int	glao	whether we are in GLAO mode where all WFS in each powfs are averaged
int	split	split reconstruction/tomography type. 0: integrated tomography 1: adhoc split tomography 2: minimum variance split tomography (only valid if recon.alg=0)
int	modal
int	nmod	-2: emulate zonal, -1: zernike, 0: zonal, 1: KL modes Maximum number of modes to control in modal controller
int	psol	Use pseudo open loop gradients for wavefront reconstruction
int	mvm	Use the various algorithms recon.alg to assemble a control matrix to multiply to gradients to get DM commands. If the algorithm needs PSOL gradient, we will have an auxillary matrix to multiply to the DM actuators and subtract from the result.
int	psd	Flag: compute PSDs of DM error signal averaged over aperture and field points (m^2/Hz).
int	psddtrat_hi	how many time step to sample for PSD computation.
int	psddtrat_lo	how many time step to sample for low order PSD computation.
int	psdnseg	how many overlapping partitions of the time history to compute PSD.
int	twfs_rmin	minimum zernike order (inclusive)
int	twfs_rmax	maximum zernike order (inclusive)
int	twfs_radonly	1: radial only, 0: all modes
char **	distortion_dm2wfs	Distortion from DM to each WFS model used in reconstruction. Affects GA
char **	distortion_dm2sci	Distortion from DM to each science model used in reconstruction. Affects HA
char **	distortion_tel2wfs	Distortion from Telescope to each WFS model used in reconstruction. Affects HXW

◆ sim_cfg_t

struct sim_cfg_t

contains input parameters for simulation, like loop gain, seeds, etc.

Collaboration diagram for sim_cfg_t:

Data Fields
real	dt	sampling period (s) for simulation.
real	dtref	sampling period (s) for setting siglev or nearecon.
real	za	zenith angle in radian
real	htel	Height of telescope. Used to adjust sodium profile range
int	start	time step to start simulation. 0
int	end	time step to stop simulation. exclusive
int	pause	Pause at the end of every time step
lmat *	seeds	simulation seeds
int	nseed	How many simulation seed
int	closeloop	closed loop or open loop
dmat *	wspsd	Telescope wind shake PSD input. Nx2. First column is freq in Hz, Second column is PSD in rad^2/Hz.
int	wsseq	sequence of wind shake time series.
dmat *	aphi	servo coefficient for high order dm. A is command. e is error signal. at time step n, the command is updated by A(n)=A(n-1)apdm(0)+A(n-2)ap(1)+...+e(n-2)*ep
dmat *	ephi	error gain for DM commands (high order)
real	f0dm	Natural frequency of the DMs.
real	zetadm	Damping of the DMs.
dmat *	aplo	servo coefficient for ngs modes.
dmat *	eplo	error gain for NGS modes (low order)
real	alhi	Additional latency (*sim.dt) of the high order loop besides 2 cycle delay.
real	allo	Additional latnecy (*sim.dt) of the low order loop
real	aptwfs	Twfs reference vector servo coefficient.
real	eptwfs	Twfs reference vector servo gain.
real	eptsph	Twfs reference vector servo gain for spherical mode
real	fcttm	cross over frequency of tip/tilt split. 0 to disable ttm.
real	fcfocus	cross-over frequency of the focus LPF.
real	fov	The diameter of total fov in arcsec
int	focus2tel	Offload focus to telescope
real	epfocus2tel
int	mffocus	method for focus blending between LGS and LO NGS 0: no focus blending. 1: Focus blending using CL gradients, for each LGS independently. 2: Focus blending using CL gradinets, for common LGS focus only (not preferred).
int	cachedm	cache dm shape on fine sampled grid matched WFS or Science grid
int	fuseint	fuse the high and low order integrators in split tomography
int	skysim	1: we are doing skycoverage preprocessing
int	evlol	evaluate open loop error only
int	noatm	disable atmosphere
int	idealfit	do ideal DM fitting from atmosphere directly.
int	idealtomo	ideal tomography without wfs (directly propagate from turbulence). conflicts with idealfit. combine with evl.tomo to evaluate its performance.
int	psfr	do PSF reconstruction telemetry
int	ecnn	Calculate WF covariance due to WFS noise cov Cnn.
int	wfsalias	Study the wfs aliasing effect by projecting turbulence onto the NULL space of DM.
int	idealwfs	Generates ideal WFS by sensing turbulence with DM range.
int	idealevl	Evaluate performance within DM range.
real	dtlo	low order wfs sampling period
real	dthi	high order wfs sampling period
int	dtrat_hi	ratio of sampling period over clock of high order wfs
int	dtrat_lo	highest dtrat of the lower order loop.
int	dtrat_lo2	lowest dtrat of the lower order loop.
int	dtrat_lof	lowest dtrat of the lower order focus loop.
int	dtrat_skip	dtrat (over sim.dt) for frame drop. Be careful when powfs.dtrat is not one.
int	noisy_hi	whether high order WFS is noisy
int	noisy_lo	whether low order WFS is noisy
real	lpfocushi	derived: lpfocus=2pifc*sim.dthi
real	lpfocuslo	derived: lpfocus=2pifc*sim.dtlo
real	lpttm	los path filter for ttm. derived: lpttm=2pifcttm*sim.dt
int	dmclip	derived: Need to clip actuator stroke
int	dmclipia	derived: Need to clip inter-actuator stroke
int	dmproj	derived: Need to projection atmosphere onto DMspace.
int	mvmport	Non zero: specify which port does the MVM server run on and connect to it for MVM reconstruction.
char *	mvmhost	Which host does the MVM server run
int	mvmsize	number of gradients to send each time. 0 is all.
int	mvmngpu	number of GPUs to use in server
char *	dmadd	Containing dm vector to simulate turbulence (added to integrator output). It should be cell array (time steps) of cell arry (DMs) of vectors. Can be empty

◆ ncpa_cfg_t

struct ncpa_cfg_t

Collaboration diagram for ncpa_cfg_t:

Data Fields
dmat *	thetax	Coordinate for NCPA calibration (arcsec)
dmat *	thetay	Coordinate for NCPA calibration (arcsec)
dmat *	wt	Weight for each point.
dmat *	hs	Height of star.
int	calib	calibrate NCPA. 1: with all DMs. 2: with only ground DM.
int	ttr	Remove average t/t from NCPA for WFS. Equivalent as repositioning WFS. default 1.
int	rmsci	1: do not include calibration residual in science path.
int	preload	preload integrator with DM sys flat
int	ndir	Number of points for NCPA calibration
char **	surf	OPD surfaces
int	nsurf	Number of OPD surfaces
char **	tsurf	Tilted surfaces, surface, not OPD
int	ntsurf	Number of tilted surfaces

◆ cn2est_cfg_t

struct cn2est_cfg_t

Parameters for Cn square estimation.

Collaboration diagram for cn2est_cfg_t:

Data Fields
dmat *	pair	If non empty, paris of WFS to use for cn2 estimation. Empty: disable cn2 estimation
int	step	do cn2 estimation every this time step
int	reset	reset the accumulated cn2 after every cn2step.
int	tomo	update tomography parameters if non zero
int	verbose	1:Print out estimated r0, cn2 during simulation.
int	keepht	>0: use the layer ht specified by atmr.ht. 2: also do slodar directly on these layers.
int	nhtomo	number of layers to feed into reconstructor. only effective if keepht=0
int	moveht	1: move the ht used for reconstructor to near strongest layers. only effective if keepht=0.
int	psol	Use pseudo open loop gradients. 0 to probe residual
real	hmax	maximum height to estimat
real	saat	subaperture area threashold to use in cn2 estimation

◆ plot_cfg_t

struct plot_cfg_t

contains input parameters for plotting during simulation. For debug purpose

Collaboration diagram for plot_cfg_t:

Data Fields
int	setup	Plot various information in setup process
int	atm	Plot the generated atmosphere
int	run	Plot information during simulation
int	opdx	Plot turbulence projected onto xloc.
int	psf	Plot PSF in linear (1) or log (2) mode
int	grad2opd	Plot gradients as reconstructed opd
int	all	Enables setup, atm, run
dmat *	opdmax	Set zlim for OPD drawing
dmat *	gmax	Set zlim for gradient drawing

◆ dbg_cfg_t

struct dbg_cfg_t

contains input parameters for debugging.

Collaboration diagram for dbg_cfg_t:

Data Fields
int	wamethod	method to compute wa for ngsmod removal.
int	mvstlimit	Limit number of modes controled on MVST
int	annular_W	Define the W0/W1 on annular aperture instead of circular
lmat *	tomo_maxit	if not empty, will study these maxits in open loop
int	tomo_hxw	1: Force use hxw always instead of ray tracing from xloc to ploc.
int	ecovxx	save the xx used to calculate ecov in psfr.
int	useopdr	use opdr in psf reconstruction
int	force	Force run even if Res_${seed}.done exists
int	cmpgpu	1: cpu code follows GPU implementation.
int	pupmask	Testing pupil mask for NGS WFS to be within LGS volume.
int	wfslinearity	Study the linearity of this wfs
int	nocgwarm	Disable warm restart in CG
int	test	Temporary any testing purpose
int	dmfullfov	let DM cover full FoV (sim.fov)
int	tomo	Comparing tomography in GPU and CPU
int	fit	Comparing DM fitting in GPU and CPU
int	gp_noamp	Use annular instead of ampground for GP
dmat *	atm	test special atmosphere. <0: fourier mode with spatial frequency 1/dbg.atm m^-1. >0: zernike mode
real	gradoff_scale	Scale the reference vector
int	gradoff_reset	reset gradoff after creating matched filter with dithering
dcell *	dmoff	DM offset for simulating turbulence on the DM. dimension: ndm*nstep
dcell *	gradoff	Introduced additional gradient offset. dimension: nwfs*nstep
int	twfsflag	use TWFS to control 0: all modes, 1: radial only
int	twfsrmax	TWFS maximum zernike radial order.
int	wfs_iac	Cubic spline coupling factor for turbulence fitting onto wfs grid.
int	fullatm	Always copy full atm to GPU.
int	lo_blend	Low order multi-rate control blending scheme.
real	eploscale	Scale of eplo
int	ahst_keepfocus	keep LGS focus in ngs mode removal
int	recon_stuck	Whether to handle stuck actuator in reconstruction.

◆ gpu_cfg_t

struct gpu_cfg_t

Configure GPU usage for different parts.

Collaboration diagram for gpu_cfg_t:

Data Fields
int	wfs	Use GPU for wavefront sensor
int	evl	Use GPU for performance evaluation
int	tomo	Use GPU for tomography
int	fit	Use GPU for DM fitting. Options: 1) assembled matrix 2) matrix free for RHS.
int	lsr	Use GPU for least square reconstruction
int	recon	Use GPU for reconstructor any of(tomo, fit, lsr)
int	psf	Use GPU for accumulating PSF.
int	moao	Use GPU for moao.

◆ moao_cfg_t

struct moao_cfg_t

contains input parameters for each MOAO type.

Collaboration diagram for moao_cfg_t:

Data Fields
real	dx	Spacing of MOAO DM act
int	order	Nominal order of this MOAO
int	used	This moao is used
int	actslave	Do we do actuator slaving
int	lrt_ptt	Piston/tip/tilt constraint
real	iac	If !=0: use cubic influence function with this Inter-Actuator Coupling coefficient.
real	stroke	Stroke of the MOAO DM
real	gdm	The gain of type I controller. a[n]=a[n-1]+e*g where g=o[n]-a[n-1]
real	ar	Aspect ratio dy/dx
real	guard
dmat *	actfloat	file containing floating actuators. nx2 coordinate
dmat *	actstuck	file containing stuck actuators. nx2 coordinate.

◆ load_cfg_t

struct load_cfg_t

contains input parameters for reusing of saved variables.

Collaboration diagram for load_cfg_t:

Data Fields
char *	atm	load atmosphere from. Contains cell array of square matrix
char *	locs	load aper_locs from
char *	aloc	load DM aloc from
char *	xloc	load xloc for recon from
char *	ploc	load ploc for recon from
char *	floc	load floc for recon from
char *	cxx	load laplacian from to do Cxx^-1 in tomo.
char *	HXW	load HXW from.
char *	GP	load GP from.
char *	GA	load GA from.
char *	mvm	load mvm from.
char *	mvmi	load mvmi from.
char *	mvmf	load mvmf from.
int	mvst	load MVST mvst_U and mvst_FU. see recon.c
int	GS0	if 1, load GS0 from powfsd_GS0.bin
int	tomo	if 1, load tomo matrix
int	fit	if 1, load fit matrix
int	W	if 1, load W0, W1
char *	ncpa	Load ncpa from this path. saved by save.ncpa

◆ save_cfg_t

struct save_cfg_t

contains input parameters for saving variables.

Collaboration diagram for save_cfg_t:

Data Fields
int	extra	Save extra results, namely clep, olep, cleNGSmp, etc
int	all	save absolutely everything. mainly for debugging
int	setup	save preparation matrices
int	recon	save reconstructor information. large
int	mvst	MVST computation intermediate matrices
int	ncpa	save NCPA surface OPD on aper and powfs
int	fdpcg	save FDPCG matrices
int	atm	save atmosphere
int	run	save run time informaton for each time step
int	opdr	save reconstructed OPD on XLOC for each time step
int	opdx	save ATM propagated to XLOC for each time step
int	dm	save computed DM actuator commands for each time step
int	evlopd	save science OPD every `evlopd` time steps
int	dither	save estimated matched filter from dithering
int	gradoff	save gradient reference vector
lmat *	wfsopd	save WFS OPD:
lmat *	ints	save WFS subaperture image
lmat *	grad	save WFS gradients
lmat *	gradnf	save WFS noise free gradients
lmat *	gradpsol	save WFS PSOL gradients
lmat *	gradgeom	save WFS geometric gradient during physical optics simu
int	wfsopdhi	save high order WFS OPD(derived)
int	wfsopdlo	save low order WFS OPD(derived)
int	intshi	save high orrder WFS integration(derived)
int	intslo	save low orrder WFS integration(derived)
int	gradhi	save WFS gradients for high order wfs (derived)
int	gradlo	save WFS gradients for low order wfs (derived)
int	gradgeomhi	save WFS geometric gradient during physical optics simulations.(derived)
int	gradgeomlo	save WFS geometric gradient during physical optics simulations.(derived)
int	gcovp	output cumulative gradient covariance average every gcovp step
int	ngcov	number of pairs of gradient covariance to compute
lmat *	gcov	size of 2*ngcov, specifying wfs for each pair
int	ecov	save covariance of DM error vector
int	mvmi	save TomoL output of mvm control matrix assembly for warm restart.
int	mvmf	save FitR output of mvm control matrix assembly
int	mvm	save computed mvm control matrix

◆ dist_cfg_t

struct dist_cfg_t

Collaboration diagram for dist_cfg_t:

Data Fields
char **	tel2wfs	Distortion from telescope pupil to each WFS
char **	dm2wfs	Distortion from DM to each WFS. Displacement due to altitude should not be included here
char **	dm2sci	Distortion from DM to science. Not specified for individual science

◆ parms_t

struct parms_t

is a wrapper of all _CFG_T data types.

Collaboration diagram for parms_t:

Data Fields
atm_cfg_t	atm	atmospheric parameters
atmr_cfg_t	atmr	information about reconstructed atm
aper_cfg_t	aper	aperture parameters
tomo_cfg_t	tomo	tomography parameters
fit_cfg_t	fit	DM fit parameters
lsr_cfg_t	lsr	LSR parameters
recon_cfg_t	recon	general reconstruction parameters
evl_cfg_t	evl	Performance evaluation parameters
powfs_cfg_t *	powfs	Array of wfs type
wfs_cfg_t *	wfs	Array of wfs
wfs_cfg_t *	wfsr	Array of wfs used in reconstruction. Has only 1 wfs per powfs in glao mode, otherwise same as wfs.
dm_cfg_t *	dm	Array of DM
moao_cfg_t *	moao	Array of MOAO
sim_cfg_t	sim	Simulation information
ncpa_cfg_t	ncpa	Surface and NCPA caibration parameters.
cn2est_cfg_t	cn2	Parameters for Cn2 estimation
plot_cfg_t	plot	Specify what to plot during simulation.
dbg_cfg_t	dbg	Specify debugging parameters
gpu_cfg_t	gpu	Specify GPU options.
load_cfg_t	load	Specify what matrices to load for debugging
save_cfg_t	save	Specify what to save to file for debugging
dist_cfg_t	distortion	Field distortion. Not misregistration
int	npowfs	Number of wfs types
int	nwfs	Number of wfs
int	nwfsr	Number of wfs used in reconstruction. =npowfs in glao, =nwfs otherwise
int	ndm	Number of DMs
int	nmoao	Number of different MOAO type
lmat *	fdlock	Records the fd of the seed lock file. if -1 will skip the seed
int	nlopowfs	Number of low order wfs types
lmat *	lopowfs	List of low order powfs
int	nhipowfs	Number of high order wfs types
lmat *	hipowfs	List of high order powfs
int	ntrpowfs	Number of tip/tilt removed wfs types
int	ntipowfs	Number of tip/tilt include wfs types
int	nphypowfs	Number of powfs with local/uplink tip/tilt loop
int	nlowfs	Number of low order wfs.
int	nhiwfs	Number of high order wfs
dmat *	dirs	Collect for beam directions
int	dither	Some WFS is doing dithering
int	ilgspowfs	Index of LGS WFS
int	nlgspowfs	Number of LGS WFS
int	itpowfs	Index of twfs
int	idmground	Index of ground dm. default to 0
int	step_lo	Enabling step for low order wfs
int	step_hi	Enabling step for high order wfs
real	hipowfs_hs	high order wfs minimum height
int	itwfssph	index of TWFS spherical mode

◆ arg_t

struct arg_t

arg_t is used for command line parsing.

Collaboration diagram for arg_t:

Data Fields
int	detach	Detach from the command line and run in background
int	over	Run simulation even if Res_${seed}.done exists
int	force	For start, bypassing scheduler
int *	gpus	Index of GPU to use. -1 to disable
int	ngpu	Number of entries in gpus
int	ngpu2	Number of GPUs to use. Ignore of gpus is set.
int	server	MAOS acting as server
char *	dirout	Result output directory
char *	conf	master .conf file. nfiraos.conf by default. -c to change
char *	confcmd	Additional configuration options supplied in command line.
char *	host	Run in another host
char *	execmd	concatenation of argv

Function Documentation

◆ setup_parms()

parms_t* setup_parms	(	const char *	mainconf,
		const char *	extraconf,
		int	over_ride
	)

This routine calles other routines in this file to setup the parms parameter struct parms and check for possible errors. parms is kept constant after returned from setup_parms.

◆ setup_parms_gpu()

void setup_parms_gpu	(	parms_t *	parms,
		int *	gpus,
		int	ngpu
	)

Additional setup_parms code to run when maos is running. It only contains GPU initialization code for the moment.

◆ free_parms()

void free_parms ( parms_t * parms )

Free the parms struct.

Data Structures

Macros

Enumerations

Functions