ComputeFstat.h File Reference

Author:: Reinhard Prix Header-file defining the API for the F-statistic functions.

#include <lal/LALRCSID.h>
#include <lal/LALComputeAM.h>
#include <lal/PulsarDataTypes.h>
#include <lal/DetectorStates.h>

Include dependency graph for ComputeFstat.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Data Structures

struct SSBtimes

Simple container for two REAL8-vectors, namely the SSB-timings DeltaT_alpha and Tdot_alpha, with one entry per SFT-timestamp. More...

struct MultiSSBtimes

Multi-IFO container for SSB timings. More...

struct AntennaPatternMatrix

Struct holding the "antenna-pattern" matrix $\mathcal{M}_{\mu\nu} \equiv \left( \mathbf{h}_\mu|\mathbf{h}_\nu\right)$ , in terms of the multi-detector scalar product. More...

struct MultiAMCoeffs

Multi-IFO container for antenna-pattern coefficients a^X(t), b^X(t) and atenna-pattern matrix M_mu_nu. More...

struct CmplxAntennaPatternMatrix

Struct holding the "antenna-pattern" matrix $\mathcal{M}_{\mu\nu} \equiv \left( \mathbf{h}_\mu|\mathbf{h}_\nu\right)$ , in terms of the multi-detector scalar product. More...

struct MultiCmplxAMCoeffs

Multi-IFO container for antenna-pattern coefficients a^X(t), b^X(t) and atenna-pattern matrix M_mu_nu. More...

struct Fcomponents

Type containing F-statistic proper plus the two complex amplitudes Fa and Fb (for ML-estimators). More...

struct ComputeFParams

Extra parameters controlling the actual computation of F. More...

struct ComputeFBuffer

Struct holding buffered ComputeFStat()-internal quantities to avoid unnecessarily recomputing things that depend ONLY on the skyposition and detector-state series (but not on the spins). More...

Defines

#define COMPUTEFSTATC_ENULL   1

#define COMPUTEFSTATC_ENONULL   2

#define COMPUTEFSTATC_EINPUT   3

#define COMPUTEFSTATC_EMEM   4

#define COMPUTEFSTATC_EXLAL   5

#define COMPUTEFSTATC_EIEEE   6

#define COMPUTEFSTATC_MSGENULL   "Arguments contained an unexpected null pointer"

#define COMPUTEFSTATC_MSGENONULL   "Output pointer is non-NULL"

#define COMPUTEFSTATC_MSGEINPUT   "Invalid input"

#define COMPUTEFSTATC_MSGEMEM   "Out of memory. Bad."

#define COMPUTEFSTATC_MSGEXLAL   "XLAL function call failed"

#define COMPUTEFSTATC_MSGEIEEE   "Floating point failure"

Enumerations

enum SSBprecision { SSBPREC_NEWTONIAN, SSBPREC_RELATIVISTIC, SSBPREC_LAST }

The precision in calculating the barycentric transformation. More...

Functions

NRCSID (COMPUTEFSTATH,"$Id: ComputeFstat.h,v 1.46 2008/10/01 21:29:47 reinhard Exp $")

int XLALComputeFaFb (Fcomponents *FaFb, const SFTVector *sfts, const PulsarSpins fkdot, const SSBtimes *tSSB, const AMCoeffs *amcoe, const ComputeFParams *params)

Revamped version of LALDemod() (based on TestLALDemod() in CFS).

int XLALComputeFaFbXavie (Fcomponents *FaFb, const SFTVector *sfts, const PulsarSpins fkdot, const SSBtimes *tSSB, const AMCoeffs *amcoe, const ComputeFParams *params)

Modified version of ComputeFaFb() based on Xavie's trick: need sufficiently oversampled SFTs and uses ZERO Dterms.

int XLALComputeFaFbCmplx (Fcomponents *FaFb, const SFTVector *sfts, const PulsarSpins fkdot, const SSBtimes *tSSB, const CmplxAMCoeffs *amcoe, const ComputeFParams *params)

Revamped version of XLALComputeFaFb() for the case where a and b are complex.

void LALGetBinarytimes (LALStatus *, SSBtimes *tBinary, const SSBtimes *tSSB, const DetectorStateSeries *DetectorStates, const BinaryOrbitParams *binaryparams, LIGOTimeGPS refTime)

For a given OrbitalParams, calculate the time-differences $\Delta T_\alpha\equiv T(t_\alpha) - T_0$ , and their derivatives $Tdot_\alpha \equiv d T / d t (t_\alpha)$ .

void LALGetMultiBinarytimes (LALStatus *status, MultiSSBtimes **multiBinary, const MultiSSBtimes *multiSSB, const MultiDetectorStateSeries *multiDetStates, const BinaryOrbitParams *binaryparams, LIGOTimeGPS refTime)

Multi-IFO version of LALGetBinarytimes().

void LALGetSSBtimes (LALStatus *, SSBtimes *tSSB, const DetectorStateSeries *DetectorStates, SkyPosition pos, LIGOTimeGPS refTime, SSBprecision precision)

For a given DetectorStateSeries, calculate the time-differences $\Delta T_\alpha\equiv T(t_\alpha) - T_0$ , and their derivatives $Tdot_\alpha \equiv d T / d t (t_\alpha)$ .

void LALGetAMCoeffs (LALStatus *, AMCoeffs *coeffs, const DetectorStateSeries *DetectorStates, SkyPosition skypos)

Compute the 'amplitude coefficients' $a(t), b(t)$ as defined in JKS98 for a series of timestamps.

void LALNewGetAMCoeffs (LALStatus *, AMCoeffs *coeffs, const DetectorStateSeries *DetectorStates, SkyPosition skypos)

Compute the 'amplitude coefficients' $a(t)\sin\zeta$ , $b(t)\sin\zeta$ as defined in JKS98 for a series of timestamps.

int XLALComputeAntennaPatternCoeffs (REAL8 *ai, REAL8 *bi, const SkyPosition *skypos, const LIGOTimeGPS *tGPS, const LALDetector *site, const EphemerisData *edat)

Compute single time-stamp antenna-pattern coefficients a(t), b(t).

void LALGetMultiSSBtimes (LALStatus *, MultiSSBtimes **multiSSB, const MultiDetectorStateSeries *multiDetStates, SkyPosition pos, LIGOTimeGPS refTime, SSBprecision precision)

Multi-IFO version of LALGetSSBtimes().

void LALGetMultiAMCoeffs (LALStatus *, MultiAMCoeffs **multiAMcoef, const MultiDetectorStateSeries *multiDetStates, SkyPosition pos)

Multi-IFO version of LALGetAMCoeffs().

void ComputeFStat (LALStatus *, Fcomponents *Fstat, const PulsarDopplerParams *doppler, const MultiSFTVector *multiSFTs, const MultiNoiseWeights *multiWeights, const MultiDetectorStateSeries *multiDetStates, const ComputeFParams *params, ComputeFBuffer *cfBuffer)

Function to compute (multi-IFO) F-statistic for given parameter-space point psPoint, normalized SFT-data (normalized by double-sided PSD Sn), noise-weights and detector state-series.

void ComputeFStatFreqBand (LALStatus *status, REAL8FrequencySeries *FstatVector, const PulsarDopplerParams *doppler, const MultiSFTVector *multiSFTs, const MultiNoiseWeights *multiWeights, const MultiDetectorStateSeries *multiDetStates, const ComputeFParams *params)

Function to compute a vector of Fstatistic values for a number of frequency bins.

int XLALWeighMultiAMCoeffs (MultiAMCoeffs *multiAMcoef, const MultiNoiseWeights *multiWeights)

Multiply AM-coeffs $a_{X\alpha}, b_{X\alpha}$ by weights $\sqrt(w_{X\alpha})$ and compute the resulting $A_d, B_d, C_d$ by simply *SUMMING* them, i.e.

void LALEstimatePulsarAmplitudeParams (LALStatus *status, PulsarCandidate *pulsarParams, const Fcomponents *Fstat, const LIGOTimeGPS *FstatRefTime, const CmplxAntennaPatternMatrix *Mmunu)

Parameter-estimation: based on large parts on Yousuke's notes and implemention (in CFSv1), extended for error-estimation.

void XLALDestroyMultiSSBtimes (MultiSSBtimes *multiSSB)

Destroy a MultiSSBtimes structure.

void XLALDestroyMultiAMCoeffs (MultiAMCoeffs *multiAMcoef)

Destroy a MultiAMCoeffs structure.

void XLALDestroyAMCoeffs (AMCoeffs *amcoef)

Destroy a AMCoeffs structure.

void XLALEmptyComputeFBuffer (ComputeFBuffer *cfb)

Destruction of a ComputeFBuffer *contents*, i.e.

int sin_cos_LUT (REAL4 *sinx, REAL4 *cosx, REAL8 x)

Calculate sin(x) and cos(x) to roughly 1e-7 precision using a lookup-table and Taylor-expansion.

int sin_cos_2PI_LUT (REAL4 *sin2pix, REAL4 *cos2pix, REAL8 x)

Variables

const SSBtimes empty_SSBtimes

const MultiSSBtimes empty_MultiSSBtimes

const AntennaPatternMatrix empty_AntennaPatternMatrix

const MultiAMCoeffs empty_MultiAMCoeffs

const Fcomponents empty_Fcomponents

const ComputeFParams empty_ComputeFParams

const ComputeFBuffer empty_ComputeFBuffer

Detailed Description

Author:: Reinhard Prix Header-file defining the API for the F-statistic functions.

Date:: 2005

This code is (partly) a descendant of an earlier implementation found in LALDemod.[ch] by Jolien Creighton, Maria Alessandra Papa, Reinhard Prix, Steve Berukoff, Xavier Siemens, Bruce Allen ComputSky.[ch] by Jolien Creighton, Reinhard Prix, Steve Berukoff LALComputeAM.[ch] by Jolien Creighton, Maria Alessandra Papa, Reinhard Prix, Steve Berukoff, Xavier Siemens

Id: ComputeFstat.h,v 1.46 2008/10/01 21:29:47 reinhard Exp

Definition in file ComputeFstat.h.

Define Documentation

#define COMPUTEFSTATC_ENULL 1

Definition at line 55 of file ComputeFstat.h.

#define COMPUTEFSTATC_ENONULL 2

Definition at line 56 of file ComputeFstat.h.

#define COMPUTEFSTATC_EINPUT 3

Definition at line 57 of file ComputeFstat.h.

#define COMPUTEFSTATC_EMEM 4

Definition at line 58 of file ComputeFstat.h.

#define COMPUTEFSTATC_EXLAL 5

Definition at line 59 of file ComputeFstat.h.

#define COMPUTEFSTATC_EIEEE 6

Definition at line 60 of file ComputeFstat.h.

#define COMPUTEFSTATC_MSGENULL "Arguments contained an unexpected null pointer"

Definition at line 62 of file ComputeFstat.h.

#define COMPUTEFSTATC_MSGENONULL "Output pointer is non-NULL"

Definition at line 63 of file ComputeFstat.h.

#define COMPUTEFSTATC_MSGEINPUT "Invalid input"

Definition at line 64 of file ComputeFstat.h.

#define COMPUTEFSTATC_MSGEMEM "Out of memory. Bad."

Definition at line 65 of file ComputeFstat.h.

#define COMPUTEFSTATC_MSGEXLAL "XLAL function call failed"

Definition at line 66 of file ComputeFstat.h.

#define COMPUTEFSTATC_MSGEIEEE "Floating point failure"

Definition at line 67 of file ComputeFstat.h.

Enumeration Type Documentation

enum SSBprecision

The precision in calculating the barycentric transformation.

Enumerator:

SSBPREC_NEWTONIAN	simple Newtonian: $\tau = t + \vec{r}\cdot\vec{n}/c$
SSBPREC_RELATIVISTIC	detailed relativistic: $\tau=\tau(t; \vec{n}, \vec{r})$
SSBPREC_LAST	end marker

Definition at line 167 of file ComputeFstat.h.

Function Documentation

NRCSID	(	COMPUTEFSTATH	,
		"$Id: ComputeFstat.	h,
		v 1.46 2008/10/01 21:29:47 reinhard Exp $"
	)

int XLALComputeFaFb	(	Fcomponents *	FaFb,
		const SFTVector *	sfts,
		const PulsarSpins	fkdot,
		const SSBtimes *	tSSB,
		const AMCoeffs *	amcoe,
		const ComputeFParams *	params
	)

Revamped version of LALDemod() (based on TestLALDemod() in CFS).

Compute JKS's Fa and Fb, which are ingredients for calculating the F-statistic.

Parameters:

params

addition computational params

Definition at line 446 of file ComputeFstat.c.

int XLALComputeFaFbXavie	(	Fcomponents *	FaFb,
		const SFTVector *	sfts,
		const PulsarSpins	fkdot,
		const SSBtimes *	tSSB,
		const AMCoeffs *	amcoe,
		const ComputeFParams *	params
	)

Modified version of ComputeFaFb() based on Xavie's trick: need sufficiently oversampled SFTs and uses ZERO Dterms.

Compute JKS's Fa and Fb, which are ingredients for calculating the F-statistic.

Parameters:

params

addition computational params

Definition at line 956 of file ComputeFstat.c.

int XLALComputeFaFbCmplx	(	Fcomponents *	FaFb,
		const SFTVector *	sfts,
		const PulsarSpins	fkdot,
		const SSBtimes *	tSSB,
		const CmplxAMCoeffs *	amcoe,
		const ComputeFParams *	params
	)

Revamped version of XLALComputeFaFb() for the case where a and b are complex.

Compute JKS's Fa and Fb, which are ingredients for calculating the F-statistic.

Parameters:

params

addition computational params

Definition at line 701 of file ComputeFstat.c.

void LALGetBinarytimes	(	LALStatus *	status,
		SSBtimes *	tBinary,
		const SSBtimes *	tSSB,
		const DetectorStateSeries *	DetectorStates,
		const BinaryOrbitParams *	binaryparams,
		LIGOTimeGPS	refTime
	)

For a given OrbitalParams, calculate the time-differences $\Delta T_\alpha\equiv T(t_\alpha) - T_0$ , and their derivatives $Tdot_\alpha \equiv d T / d t (t_\alpha)$ .

Note:: The return-vectors DeltaT and Tdot must be allocated already and have the same length as the input time-series DetStates.

Parameters:

tBinary	[out] DeltaT_alpha = T(t_alpha) - T_0; and Tdot(t_alpha)
tSSB	[in] DeltaT_alpha = T(t_alpha) - T_0; and Tdot(t_alpha)
DetectorStates	[in] detector-states at timestamps t_i
binaryparams	[in] source binary orbit parameters
refTime	SSB reference-time T_0 of pulsar-parameters

Definition at line 1499 of file ComputeFstat.c.

void LALGetMultiBinarytimes	(	LALStatus *	status,
		MultiSSBtimes **	multiBinary,
		const MultiSSBtimes *	multiSSB,
		const MultiDetectorStateSeries *	multiDetStates,
		const BinaryOrbitParams *	binaryparams,
		LIGOTimeGPS	refTime
	)

Multi-IFO version of LALGetBinarytimes().

Get all binary-timings for all input detector-series.

Parameters:

multiBinary	[out] SSB-timings for all input detector-state series
multiSSB	[in] SSB-timings for all input detector-state series
multiDetStates	[in] detector-states at timestamps t_i
binaryparams	[in] source binary orbit parameters
refTime	SSB reference-time T_0 for SSB-timing

Definition at line 1637 of file ComputeFstat.c.

void LALGetSSBtimes	(	LALStatus *	status,
		SSBtimes *	tSSB,
		const DetectorStateSeries *	DetectorStates,
		SkyPosition	pos,
		LIGOTimeGPS	refTime,
		SSBprecision	precision
	)

For a given DetectorStateSeries, calculate the time-differences $\Delta T_\alpha\equiv T(t_\alpha) - T_0$ , and their derivatives $Tdot_\alpha \equiv d T / d t (t_\alpha)$ .

Note:: The return-vectors DeltaT and Tdot must be allocated already and have the same length as the input time-series DetStates.

Parameters:

tSSB	[out] DeltaT_alpha = T(t_alpha) - T_0; and Tdot(t_alpha)
DetectorStates	[in] detector-states at timestamps t_i
pos	source sky-location
refTime	SSB reference-time T_0 of pulsar-parameters
precision	relativistic or Newtonian SSB transformation?

Definition at line 1716 of file ComputeFstat.c.

void LALGetAMCoeffs	(	LALStatus *	status,
		AMCoeffs *	coeffs,
		const DetectorStateSeries *	DetectorStates,
		SkyPosition	skypos
	)

Compute the 'amplitude coefficients' $a(t), b(t)$ as defined in JKS98 for a series of timestamps.

The input consists of the DetectorState-timeseries, which contains the detector-info and the LMST's corresponding to the different times.

In order to allow re-using the output-structure AMCoeffs for subsequent calls, we require the REAL4Vectors a and b to be allocated already and to have the same length as the DetectoStates-timeseries.

Note:: This is an alternative implementation to LALComputeAM() with the aim to be both simpler and faster. The difference being that we don't implicitly re-derive the final expression here but simply try to implement the final expressions (12), (13) in JKS98 in the most economical way possible.

Parameters:

coeffs	[out] amplitude-coeffs {a(t_i), b(t_i)}
DetectorStates	timeseries of detector states
skypos	{alpha,delta} of the source

Definition at line 1157 of file ComputeFstat.c.

void LALNewGetAMCoeffs	(	LALStatus *	status,
		AMCoeffs *	coeffs,
		const DetectorStateSeries *	DetectorStates,
		SkyPosition	skypos
	)

Compute the 'amplitude coefficients' $a(t)\sin\zeta$ , $b(t)\sin\zeta$ as defined in JKS98 for a series of timestamps.

The input consists of the DetectorState-timeseries, which contains the detector-info and the LMST's corresponding to the different times.

In order to allow re-using the output-structure AMCoeffs for subsequent calls, we require the REAL4Vectors a and b to be allocated already and to have the same length as the DetectoStates-timeseries.

Note:: This is an alternative implementation to both LALComputeAM() and LALGetAMCoeffs(), which uses the geometrical definition of $a\sin\zeta$ and $b\sin\zeta$ as detector response coefficients in a preferred polarization basis. (It is thereby more general than the JKS expressions and could be used e.g., with the response tensor of a bar detector with no further modification needed.)

Parameters:

coeffs	[out] amplitude-coeffs {a(t_i), b(t_i)}
DetectorStates	timeseries of detector states
skypos	{alpha,delta} of the source

Definition at line 1319 of file ComputeFstat.c.

int XLALComputeAntennaPatternCoeffs	(	REAL8 *	ai,
		REAL8 *	bi,
		const SkyPosition *	skypos,
		const LIGOTimeGPS *	tGPS,
		const LALDetector *	site,
		const EphemerisData *	edat
	)

Compute single time-stamp antenna-pattern coefficients a(t), b(t).

This is a simplified (ugly) wrapper to LALNewGetAMCoeffs() to allow the computing a single-timestamp antenna-pattern, without having to worry about the whole DetectorStateSeries complexity...

Parameters:

ai	[out] antenna-pattern function a(t)
bi	[out] antenna-pattern function b(t)
skypos	[in] skyposition {alpha, delta}
tGPS	[in] GPS time t
site	[in] detector
edat	[in] ephemeris-data

Definition at line 1426 of file ComputeFstat.c.

void LALGetMultiSSBtimes	(	LALStatus *	status,
		MultiSSBtimes **	multiSSB,
		const MultiDetectorStateSeries *	multiDetStates,
		SkyPosition	skypos,
		LIGOTimeGPS	refTime,
		SSBprecision	precision
	)

Multi-IFO version of LALGetSSBtimes().

Get all SSB-timings for all input detector-series.

NOTE: contrary to LALGetSSBtimes(), this functions *allocates* the output-vector, use XLALDestroyMultiSSBtimes() to free this.

Parameters:

multiSSB	[out] SSB-timings for all input detector-state series
multiDetStates	[in] detector-states at timestamps t_i
skypos	source sky-position [in equatorial coords!]
refTime	SSB reference-time T_0 for SSB-timing
precision	use relativistic or Newtonian SSB timing?

Definition at line 1824 of file ComputeFstat.c.

void LALGetMultiAMCoeffs	(	LALStatus *	status,
		MultiAMCoeffs **	multiAMcoef,
		const MultiDetectorStateSeries *	multiDetStates,
		SkyPosition	skypos
	)

Multi-IFO version of LALGetAMCoeffs().

Get all antenna-pattern coefficients for all input detector-series.

NOTE: contrary to LALGetAMCoeffs(), this functions *allocates* the output-vector, use XLALDestroyMultiAMCoeffs() to free this.

Parameters:

multiAMcoef	[out] AM-coefficients for all input detector-state series
multiDetStates	[in] detector-states at timestamps t_i
skypos	source sky-position [in equatorial coords!]

Definition at line 1900 of file ComputeFstat.c.

void ComputeFStat	(	LALStatus *	status,
		Fcomponents *	Fstat,
		const PulsarDopplerParams *	doppler,
		const MultiSFTVector *	multiSFTs,
		const MultiNoiseWeights *	multiWeights,
		const MultiDetectorStateSeries *	multiDetStates,
		const ComputeFParams *	params,
		ComputeFBuffer *	cfBuffer
	)

Function to compute (multi-IFO) F-statistic for given parameter-space point psPoint, normalized SFT-data (normalized by double-sided PSD Sn), noise-weights and detector state-series.

NOTE: for better efficiency some quantities that need to be recomputed only for different sky-positions are buffered in cfBuffer if given.

In order to 'empty' this buffer (at the end) use XLALEmptyComputeFBuffer()
You CAN pass NULL for the cfBuffer if you don't want to use buffering (slower).

NOTE2: there's a spaceholder for binary-pulsar parameters in psPoint, but this it not implemented yet.

Parameters:

Fstat	[out] Fstatistic + Fa, Fb
doppler	parameter-space point to compute F for
multiSFTs	normalized (by DOUBLE-sided Sn!) data-SFTs of all IFOs
multiWeights	noise-weights of all SFTs
multiDetStates	'trajectories' of the different IFOs
params	addition computational params
cfBuffer	CF-internal buffering structure

Definition at line 195 of file ComputeFstat.c.

void ComputeFStatFreqBand	(	LALStatus *	status,
		REAL8FrequencySeries *	fstatVector,
		const PulsarDopplerParams *	doppler,
		const MultiSFTVector *	multiSFTs,
		const MultiNoiseWeights *	multiWeights,
		const MultiDetectorStateSeries *	multiDetStates,
		const ComputeFParams *	params
	)

Function to compute a vector of Fstatistic values for a number of frequency bins.

This function is simply a wrapper for ComputeFstat() which is called repeatedly for every frequency value. The output, i.e. fstatVector must be properly allocated before this function is called. The values of the start frequency, the step size in the frequency and the number of frequency values for which the Fstatistic is to be calculated are read from fstatVector. The other parameters are not checked and they must be correctly set outside this function.

something to improve/cleanup -- the start frequency is available both from the fstatvector and from the input doppler point -- they could be inconsistent or the user of this function could misunderstand

Parameters:

fstatVector	[out] Vector of Fstat values
doppler	parameter-space point to compute F for
multiSFTs	normalized (by DOUBLE-sided Sn!) data-SFTs of all IFOs
multiWeights	noise-weights of all SFTs
multiDetStates	'trajectories' of the different IFOs
params	addition computational params

Definition at line 119 of file ComputeFstat.c.

int XLALWeighMultiAMCoeffs	(	MultiAMCoeffs *	multiAMcoef,
		const MultiNoiseWeights *	multiWeights
	)

Multiply AM-coeffs $a_{X\alpha}, b_{X\alpha}$ by weights $\sqrt(w_{X\alpha})$ and compute the resulting $A_d, B_d, C_d$ by simply *SUMMING* them, i.e.

$A_d \equiv \sum_{X,\alpha} \sqrt{w_{X\alpha} a_{X\alpha}^2$ etc.

NOTE: this function modifies the AMCoeffs *in place* ! NOTE2: if the weights = NULL, we assume unit-weights.

Definition at line 2085 of file ComputeFstat.c.

void LALEstimatePulsarAmplitudeParams	(	LALStatus *	status,
		PulsarCandidate *	pulsarParams,
		const Fcomponents *	Fstat,
		const LIGOTimeGPS *	FstatRefTime,
		const CmplxAntennaPatternMatrix *	Mmunu
	)

Parameter-estimation: based on large parts on Yousuke's notes and implemention (in CFSv1), extended for error-estimation.

This implementation follows closely the derivations found in http://www.lsc-group.phys.uwm.edu/cgi-bin/enote.pl?nb=puls5knownpulsardemod&action=view&page=12

Parameters:

pulsarParams	[out] estimated params {h0,cosi,phi0,psi} plus error-estimates
Fstat	Fstat-components Fa, Fb
FstatRefTime	reference-time for the phase of Fa, Fb
Mmunu	antenna-pattern A,B,C and normalization S_inv*Tsft

Definition at line 2255 of file ComputeFstat.c.

void XLALDestroyMultiSSBtimes ( MultiSSBtimes * multiSSB )

Destroy a MultiSSBtimes structure.

Note, this is "NULL-robust" in the sense that it will not crash on NULL-entries anywhere in this struct, so it can be used for failure-cleanup even on incomplete structs

Definition at line 1978 of file ComputeFstat.c.

void XLALDestroyMultiAMCoeffs ( MultiAMCoeffs * multiAMcoef )

Destroy a MultiAMCoeffs structure.

Note, this is "NULL-robust" in the sense that it will not crash on NULL-entries anywhere in this struct, so it can be used for failure-cleanup even on incomplete structs

Definition at line 2013 of file ComputeFstat.c.

void XLALDestroyAMCoeffs ( AMCoeffs * amcoef )

Destroy a AMCoeffs structure.

Note, this is "NULL-robust" in the sense that it will not crash on NULL-entries anywhere in this struct, so it can be used for failure-cleanup even on incomplete structs

Definition at line 2040 of file ComputeFstat.c.

void XLALEmptyComputeFBuffer ( ComputeFBuffer * cfb )

Destruction of a ComputeFBuffer *contents*, i.e.

the multiSSB and multiAMcoeff, while the buffer-container is not freed (which is why it's passed by value and not by reference...)

Definition at line 2063 of file ComputeFstat.c.

int sin_cos_LUT	(	REAL4 *	sinx,
		REAL4 *	cosx,
		REAL8	x
	)

Calculate sin(x) and cos(x) to roughly 1e-7 precision using a lookup-table and Taylor-expansion.

NOTE: this function will fail for arguments larger than |x| > INT4_MAX = 2147483647 ~ 2e9 !!!

return = 0: OK, nonzero=ERROR

Definition at line 2179 of file ComputeFstat.c.

int sin_cos_2PI_LUT	(	REAL4 *	sin2pix,
		REAL4 *	cos2pix,
		REAL8	x
	)

Definition at line 2191 of file ComputeFstat.c.

Variable Documentation

const SSBtimes empty_SSBtimes

Definition at line 95 of file ComputeFstat.c.

const MultiSSBtimes empty_MultiSSBtimes

Definition at line 96 of file ComputeFstat.c.

const AntennaPatternMatrix empty_AntennaPatternMatrix

Definition at line 97 of file ComputeFstat.c.

const MultiAMCoeffs empty_MultiAMCoeffs

Definition at line 98 of file ComputeFstat.c.

const Fcomponents empty_Fcomponents

Definition at line 99 of file ComputeFstat.c.

const ComputeFParams empty_ComputeFParams

Definition at line 100 of file ComputeFstat.c.

const ComputeFBuffer empty_ComputeFBuffer

Definition at line 101 of file ComputeFstat.c.

Generated on Tue Oct 7 02:40:53 2008 for LAL by

1.5.2


Data Structures
struct	SSBtimes
	Simple container for two REAL8-vectors, namely the SSB-timings DeltaT_alpha and Tdot_alpha, with one entry per SFT-timestamp. More...
struct	MultiSSBtimes
	Multi-IFO container for SSB timings. More...
struct	AntennaPatternMatrix
	Struct holding the "antenna-pattern" matrix $\mathcal{M}_{\mu\nu} \equiv \left( \mathbf{h}_\mu\|\mathbf{h}_\nu\right)$ , in terms of the multi-detector scalar product. More...
struct	MultiAMCoeffs
	Multi-IFO container for antenna-pattern coefficients a^X(t), b^X(t) and atenna-pattern matrix M_mu_nu. More...
struct	CmplxAntennaPatternMatrix
	Struct holding the "antenna-pattern" matrix $\mathcal{M}_{\mu\nu} \equiv \left( \mathbf{h}_\mu\|\mathbf{h}_\nu\right)$ , in terms of the multi-detector scalar product. More...
struct	MultiCmplxAMCoeffs
	Multi-IFO container for antenna-pattern coefficients a^X(t), b^X(t) and atenna-pattern matrix M_mu_nu. More...
struct	Fcomponents
	Type containing F-statistic proper plus the two complex amplitudes Fa and Fb (for ML-estimators). More...
struct	ComputeFParams
	Extra parameters controlling the actual computation of F. More...
struct	ComputeFBuffer
	Struct holding buffered ComputeFStat()-internal quantities to avoid unnecessarily recomputing things that depend ONLY on the skyposition and detector-state series (but not on the spins). More...
Defines
#define	COMPUTEFSTATC_ENULL 1
#define	COMPUTEFSTATC_ENONULL 2
#define	COMPUTEFSTATC_EINPUT 3
#define	COMPUTEFSTATC_EMEM 4
#define	COMPUTEFSTATC_EXLAL 5
#define	COMPUTEFSTATC_EIEEE 6
#define	COMPUTEFSTATC_MSGENULL "Arguments contained an unexpected null pointer"
#define	COMPUTEFSTATC_MSGENONULL "Output pointer is non-NULL"
#define	COMPUTEFSTATC_MSGEINPUT "Invalid input"
#define	COMPUTEFSTATC_MSGEMEM "Out of memory. Bad."
#define	COMPUTEFSTATC_MSGEXLAL "XLAL function call failed"
#define	COMPUTEFSTATC_MSGEIEEE "Floating point failure"
Enumerations
enum	SSBprecision { SSBPREC_NEWTONIAN, SSBPREC_RELATIVISTIC, SSBPREC_LAST }
	The precision in calculating the barycentric transformation. More...
Functions
	NRCSID (COMPUTEFSTATH,"$Id: ComputeFstat.h,v 1.46 2008/10/01 21:29:47 reinhard Exp $")
int	XLALComputeFaFb (Fcomponents FaFb, const SFTVector sfts, const PulsarSpins fkdot, const SSBtimes tSSB, const AMCoeffs amcoe, const ComputeFParams *params)
	Revamped version of LALDemod() (based on TestLALDemod() in CFS).
int	XLALComputeFaFbXavie (Fcomponents FaFb, const SFTVector sfts, const PulsarSpins fkdot, const SSBtimes tSSB, const AMCoeffs amcoe, const ComputeFParams *params)
	Modified version of ComputeFaFb() based on Xavie's trick: need sufficiently oversampled SFTs and uses ZERO Dterms.
int	XLALComputeFaFbCmplx (Fcomponents FaFb, const SFTVector sfts, const PulsarSpins fkdot, const SSBtimes tSSB, const CmplxAMCoeffs amcoe, const ComputeFParams *params)
	Revamped version of XLALComputeFaFb() for the case where a and b are complex.
void	LALGetBinarytimes (LALStatus , SSBtimes tBinary, const SSBtimes tSSB, const DetectorStateSeries DetectorStates, const BinaryOrbitParams *binaryparams, LIGOTimeGPS refTime)
	For a given OrbitalParams, calculate the time-differences $\Delta T_\alpha\equiv T(t_\alpha) - T_0$ , and their derivatives $Tdot_\alpha \equiv d T / d t (t_\alpha)$ .
void	LALGetMultiBinarytimes (LALStatus status, MultiSSBtimes multiBinary, const MultiSSBtimes multiSSB, const MultiDetectorStateSeries multiDetStates, const BinaryOrbitParams binaryparams, LIGOTimeGPS refTime)
	Multi-IFO version of LALGetBinarytimes().
void	LALGetSSBtimes (LALStatus , SSBtimes tSSB, const DetectorStateSeries *DetectorStates, SkyPosition pos, LIGOTimeGPS refTime, SSBprecision precision)
	For a given DetectorStateSeries, calculate the time-differences $\Delta T_\alpha\equiv T(t_\alpha) - T_0$ , and their derivatives $Tdot_\alpha \equiv d T / d t (t_\alpha)$ .
void	LALGetAMCoeffs (LALStatus , AMCoeffs coeffs, const DetectorStateSeries *DetectorStates, SkyPosition skypos)
	Compute the 'amplitude coefficients' as defined in JKS98 for a series of timestamps.
void	LALNewGetAMCoeffs (LALStatus , AMCoeffs coeffs, const DetectorStateSeries *DetectorStates, SkyPosition skypos)
	Compute the 'amplitude coefficients' $a(t)\sin\zeta$ , $b(t)\sin\zeta$ as defined in JKS98 for a series of timestamps.
int	XLALComputeAntennaPatternCoeffs (REAL8 ai, REAL8 bi, const SkyPosition skypos, const LIGOTimeGPS tGPS, const LALDetector site, const EphemerisData edat)
	Compute single time-stamp antenna-pattern coefficients a(t), b(t).
void	LALGetMultiSSBtimes (LALStatus , MultiSSBtimes multiSSB, const MultiDetectorStateSeries multiDetStates, SkyPosition pos, LIGOTimeGPS refTime, SSBprecision precision)
	Multi-IFO version of LALGetSSBtimes().
void	LALGetMultiAMCoeffs (LALStatus , MultiAMCoeffs multiAMcoef, const MultiDetectorStateSeries multiDetStates, SkyPosition pos)
	Multi-IFO version of LALGetAMCoeffs().
void	ComputeFStat (LALStatus , Fcomponents Fstat, const PulsarDopplerParams doppler, const MultiSFTVector multiSFTs, const MultiNoiseWeights multiWeights, const MultiDetectorStateSeries multiDetStates, const ComputeFParams params, ComputeFBuffer cfBuffer)
	Function to compute (multi-IFO) F-statistic for given parameter-space point psPoint, normalized SFT-data (normalized by double-sided PSD Sn), noise-weights and detector state-series.
void	ComputeFStatFreqBand (LALStatus status, REAL8FrequencySeries FstatVector, const PulsarDopplerParams doppler, const MultiSFTVector multiSFTs, const MultiNoiseWeights multiWeights, const MultiDetectorStateSeries multiDetStates, const ComputeFParams *params)
	Function to compute a vector of Fstatistic values for a number of frequency bins.
int	XLALWeighMultiAMCoeffs (MultiAMCoeffs multiAMcoef, const MultiNoiseWeights multiWeights)
	Multiply AM-coeffs $a_{X\alpha}, b_{X\alpha}$ by weights $\sqrt(w_{X\alpha})$ and compute the resulting by simply SUMMING them, i.e.
void	LALEstimatePulsarAmplitudeParams (LALStatus status, PulsarCandidate pulsarParams, const Fcomponents Fstat, const LIGOTimeGPS FstatRefTime, const CmplxAntennaPatternMatrix *Mmunu)
	Parameter-estimation: based on large parts on Yousuke's notes and implemention (in CFSv1), extended for error-estimation.
void	XLALDestroyMultiSSBtimes (MultiSSBtimes *multiSSB)
	Destroy a MultiSSBtimes structure.
void	XLALDestroyMultiAMCoeffs (MultiAMCoeffs *multiAMcoef)
	Destroy a MultiAMCoeffs structure.
void	XLALDestroyAMCoeffs (AMCoeffs *amcoef)
	Destroy a AMCoeffs structure.
void	XLALEmptyComputeFBuffer (ComputeFBuffer *cfb)
	Destruction of a ComputeFBuffer contents, i.e.
int	sin_cos_LUT (REAL4 sinx, REAL4 cosx, REAL8 x)
	Calculate sin(x) and cos(x) to roughly 1e-7 precision using a lookup-table and Taylor-expansion.
int	sin_cos_2PI_LUT (REAL4 sin2pix, REAL4 cos2pix, REAL8 x)
Variables
const SSBtimes	empty_SSBtimes
const MultiSSBtimes	empty_MultiSSBtimes
const AntennaPatternMatrix	empty_AntennaPatternMatrix
const MultiAMCoeffs	empty_MultiAMCoeffs
const Fcomponents	empty_Fcomponents
const ComputeFParams	empty_ComputeFParams
const ComputeFBuffer	empty_ComputeFBuffer