PBMath1DIPoly.h

Classes

PBMath1DIPoly -- PBMath1DIPoly is a 1-D Inverse Polynomial Expansion for a Primary Beam (full description)

class PBMath1DIPoly : public PBMath1D

Interface

Public Members

PBMath1DIPoly(const Vector<Double>& coeff, Quantity maxRad, Quantity refFreq, Bool isThisVP=False, BeamSquint squint=BeamSquint(isThisVP(Quantity(0.0, "deg"), Quantity(0.0, "deg"), isThisVP::Ref(isThisVP::AZEL)), Quantity(1.0, "GHz")), Bool useSymmetricBeam=False)

PBMath1DIPoly& operator=(const PBMath1DIPoly& other)

~PBMath1DIPoly()

PBMathInterface::PBClass whichPBClass()

void summary(Int nValues=0)

Protected Members

void fillPBArray()

Description

Prerequisite

PBMathInterface class
PBMath1D class

Etymology

PBMath1DIPoly: derived from PBMath1D, implements an Inverse Polynomial PB and VP

Synopsis

See PBMath1D for a general synopsis of the 1D PB types.

VP(x) = sum_i coeff_i * x ^{2i}, out to maximumRadius (referenced to the reference frequency)

x is in arcminutes, referenced to the reference frequency

This move may make me a bit unpopular: since the PBMath1D constructors construct a Voltage Pattern, the coeficienct for a polynomial type must be in terms of the VP, not the PB. Hence, go back to the original PB data, take the square root, and fit the same type polynomial to the VP data.

Inverse polynomial VP model: fit polynomial to 1/(VP_data). Optionally, you can fit to 1/(VP_data) - 1, ignoring the constant polynomial term so that the PB is constrained to be 1.0 at the beam center.

Example

    coef(0) = 1.0;     / VLA model
    coef(1) = 7.22110e-04; 
    coef(2) = 1.21421e-07;
    coef(3) = 9.68612e-11; 
    coef(4) = 1.86268e-13;
    PBMath1DIPoly ipolyPB( coef, Quantity(43.0,"'"), Quantity(1.0,"GHz"));

Motivation

All of the 1-D PB types have everything in common except for the details of their parameterization. This lightweight class deals with those differences: construction, filling the PBArray from construction parameters, and flushing to disk. The inverse polynomial PB type is used for VLA and ATNF models.

To Do

constructor from a MS beam subtable
flush to MS beam subtable

Member Description

PBMath1DIPoly()

PBMath1DIPoly(const Vector<Double>& coeff, Quantity maxRad, Quantity refFreq, Bool isThisVP=False, BeamSquint squint=BeamSquint(isThisVP(Quantity(0.0, "deg"), Quantity(0.0, "deg"), isThisVP::Ref(isThisVP::AZEL)), Quantity(1.0, "GHz")), Bool useSymmetricBeam=False)

Instantiation from arguments; default = no squint squint is the offset from pointing center if the Stokes R beam useSymmetricBeam forces a fit to the squinted beam

PBMath1DIPoly& operator=(const PBMath1DIPoly& other)

Instantiation from a row in the Beam subTable PBMath1DIPoly(const Table& BeamSubTable, Int row, Bool useSymmetricBeam=False);

Copy constructor PBMath1DIPoly(const PBMath1DIPoly& other);

Assignment operator, by reference

~PBMath1DIPoly()

Clone the object CountedPtr clone();

PBMathInterface::PBClass whichPBClass()

Get the type of PB this is

void summary(Int nValues=0)

Flush the construction parameters to disk Bool flushToTable(Table& beamSubTable, Int iRow);

Summarize the construction data for this primary beam

void fillPBArray()

Fill in vp_p array from construction parameters