Precession class and calculations. More...

#include <Precession.h>

Public Types
enum	PrecessionTypes { NONE, IAU1976, B1950, IAU2000, IAU2000A, IAU2000B, STANDARD }
	Types of known precession calculations (at 1995/09/04 STANDARD == IAU1976), from 2004/01/01 will be IAU2000) More...
Public Member Functions
	Precession ()
	Default constructor, generates default J2000 precession identification.
	Precession (const Precession &other)
	Copy constructor (deep copy)
	Precession (PrecessionTypes type, Double catepoch=0)
	Constructor with epoch in Julian days.
Precession &	operator= (const Precession &other)
	Copy assignment (deep copy)
	~Precession ()
const Euler &	operator() (Double epoch)
	Return the precession angles (for IAU2000 including the IAU 2000 corrections) at the specified epoch (in MJD; TT for IAU2000).
const Euler &	derivative (Double epoch)
	Return derivative of precession (d^-1)
void	init ()
	Re-initialise Precession object.
void	init (PrecessionTypes type, Double catepoch=0)
void	refresh ()
	Refresh calculations.
Static Public Attributes
static const Double	INTV
	Default interval to be used for linear approximation (in days)
Private Member Functions
void	copy (const Precession &other)
	Make a copy.
void	fillEpoch ()
	Create correct default fixedEpoch and catalogue epoch data.
void	calcPrec (Double t)
	Calculate precession angles for time t.
Private Attributes
PrecessionTypes	method_p
	Method to be used.
Double	fixedEpoch_p
	Fixed epoch to be used (MJD)
Double	T_p
	Fixed epoch in centuries from base epoch.
Double	cent_p
	Length of century (depending on Bessel or Julian days)
Double	refEpoch_p
	Reference epoch;.
Double	checkEpoch_p
	Check epoch.
Polynomial< Double >	zeta_p [3]
	Polynomial coefficients for zeta,z,theta.
Double	pval_p [3]
	Cached calculated angles.
Double	dval_p [3]
	Cached derivatives.
Int	lres_p
	To reference results, and use a few in interim calculations, results are saced in a circular buffer.
Euler	result_p [4]
	Last calculation.
Static Private Attributes
static uInt	myInterval_reg
	Interpolation interval aipsrc registration.

Detailed Description

Precession class and calculations.

Intended use:

Public interface

Review Status

Reviewed By:: Tim Cornwell
Date Reviewed:: 1996/07/01
Test programs:: tMeasMath

Prerequisite

Measure class for use
Euler class for format
MeasData class for constants
MeasTable class for other data

Etymology

Precession

Synopsis

Precession forms the class for precession calculations. It is a simple container with the selected method, and the mean epoch. It acts as a cache for values and their derivatives, to enable fast calculations for time epochs close together (see the aipsrc variable measures.precession.d_interval.
The calculation method is selected from one of the following:

Precession::STANDARD (at 1995/09/04 the IAU1976 definition, at 2004/01/01 the IAU2000 defeinition))
Precession::NONE (precession of zero returned
Precession::IAU1976
Precession::B1950
Precession::IAU2000

Epochs can be specified as the MJD (with defined constants MeasData::MJD2000 and MeasData::MJD1950 or the actual MJD), leading to the following constructors:

Precession() default; assuming JD2000, STANDARD
Precession(method) assuming the correct default epoch of JD2000 or JD1950 depending on method
Precession(method,epoch) with epoch Double(MJD) (Note: not valid for IAU2000: in that case always JD2000 assumed)

Actual precession for a certain Epoch (TT for IAU2000) is calculated by the () operator as Precession(epoch), with epoch Double MJD. Values returned as an Euler . The derivative (d^-1) can be obtained as well by derivative(epoch).
A Precession can be re-initialed with a different method and/or zero epoch with the init() functions (same format as constructors). To bypass the full calculation actual returned values are calculated using the derivative if within about 2 hours (error less than about 10^-5 mas). A call to refresh() will re-initiate calculations from scratch.
The following details can be set with the Aipsrc mechanism:

measures.precession.d_interval: approximation interval as time (fraction of days is default unit) over which linear approximation is used (default is 0.1 day).

Example

    #include <measures/Measures.h>
        MVDirection pos(Quantity(10,"degree"),Quantity(-10.5,"degree"));
                                                // direction RA=10; DEC=-10.5
        Precession mine(Precession::IAU1976);   // define precession type
        RotMatrix rotat(mine(45837.0));         // rotation matrix for 84/05/17
        MVDirection new = rotat*pos;            // apply precession
        rotat = RotMatrix(mine(45839.0));       // interpolate new precession
                                                // assuming d_interval set large

Motivation

To calculate the precession angles. An alternate route could have been a global function, but having a simple container allows caching of some calculations for speed.
Using MJD (JD-2400000.5) rather than JD is for precision reasons.