Compile-Time Variables

Defined within src/Main/GlobalVariables.h.

The following variables are defined at compile-time within the file and control important user-variables which must be set at compile time for performance reasons.

Efficiency Vector Parameters

These parameters primarily control the number of ordering of the elements within the Efficiency Vector.

const int Nt = 896769

Number of coarse time bins, can be 0 < Nt < TotalScanningTime.

const int Nm = 10

Number of magnitude bins.

const int healpix_order = 0

Order of the healpix mapping, can be any integer >= 0. Higher numbers = more pixels on the map.

const int needlet_order = -1

Maximum order of needlets used, can be any integer -1 <= needlet_order <= healpix_order. Higher numbers = finer detail within the map.

const int NVariancePops = 3

Number of variance populations (hyperparameter)

const int hyperOrder = 4

Order of the variance fitting, must be an even number.

Priors & Lengthscales

const double xtPrior = 6

Mean of prior on xt outside of gaps.

const double studentNu = 0.5

Student t nu parameter for zt prior.

const double sigmat = 3

Standard deviation of zt/xt prior (sort of)

const double lt_revs = 1

Temporal coupling lengthscale for zt (in Gaia revolution periods)

const double xmPrior = 3

Mean of gaussian prior on xml.

const double lm = 3

Magnitude coupling lengthscale (in mag-bins)

const double gapPriorAlpha = 0.5

Beta distribution alpha-value for the prior on the known gaps (in xt space)

const double gapPriorPeak = -6

Beta distribution peak value of xt.

const double gapPriorBeta = gapPriorAlpha * exp(-gapPriorPeak)

Given peak value and alpha value, derive appropriate beta parameter.

Initialisation

const double initialisationBounds = 0.3

Unless otherwise specified, the efficiency vector randomly initialises values between +/- this value.

const double xmInitialised = 3

The initialisation value of the zeroth-order (whole-sky) spatial mode.

Data Properties

const std::string TempDirName = "TempPositions"

The subdirectory within the output directory used to save non-converged positions.

const int DataLoadCount = 1e6

The maximum number of lines which can be loaded from a file. If the number is 0, or exceeds the number of lines within the file, the entire file is read.

const int magOffset = 0

offset of loaded files from 0.csv (assuming default file/bin association)

Misc. Likelihood Parameters

std::vector<bool> GapList

A list of length Nt. GapList[i] is true if t[i] corresponds to a time during which our.

Gap Catalogued indicates Gaia was deactivated.

enum VarianceScaling

The three kinds of scaling we have incorporated into the Variance Model.

Values:

enumerator NScaling

Scaling parameter is the (fixed) number of visitation of a star.

enumerator MScaling

Scaling parameter is the mean predicted number of detections (i.e. the sum of probabilities p_t * p_ml)

enumerator ActiveNScaling

Scaling parameter is the mean predicted number of times the star was visited whilst Gaia was turned on (i.e. the sum of probabilities p_t)

const VarianceScaling ScalingMode = ActiveNScaling

The chosen kind of scaling.

const double choleskyTolerance = 1e-4

The ‘significant value’ - values below this are assumed to be zero and not included in #LogLikelihood::cholesky_w.