class Gambit::MultiNest::LogLikeWrapper #

Class to connect multinest log-likelihood function and ScannerBit likelihood function.

#include <multinest.hpp>

Public Functions #

Public Functions Documentation #

function LogLikeWrapper #

LogLikeWrapper(
    scanPtr loglike,
    printer_interface & printer
)

Constructor.

LogLikeWrapper Constructor.

function LogLike #

double LogLike(
    double * Cube,
    int ndim,
    int 
)

Main interface function from MultiNest to ScannerBit-supplied loglikelihood function.

Main interface function from MultiNest to ScannerBit-supplied loglikelihood function This is the function that will be passed to Multinest as the loglike callback routine

Input arguments ndim = dimensionality (total number of free parameters) of the problem npars = total number of free plus derived parameters context = void pointer, any additional information

Input/Output arguments Cube[npars] = on entry has the ndim parameters in unit-hypercube on exit, the physical parameters plus copy any derived parameters you want to store with the free parameters

Output arguments lnew = loglikelihood

function dumper #

void dumper(
    int nSamples,
    int nlive,
    int nPar,
    double * physLive,
    double * posterior,
    double * ,
    double ,
    double ,
    double 
)

Main interface to MultiNest dumper routine.

Main interface to MultiNest dumper routine The dumper routine will be called every updInt*10 iterations MultiNest does not need to the user to do anything. User can use the arguments in whichever way he/she wants

Arguments:

nSamples = total number of samples in posterior distribution nlive = total number of live points nPar = total number of parameters (free + derived) physLive[1][nlive * (nPar + 1)] = 2D array containing the last set of live points (physical parameters plus derived parameters) along with their loglikelihood values TODO: Multinest uses the likelihood of the lowest live point as the “threshold” for iterating, i.e. it throws out the live point if it finds a better one. So we can use this number to update the GAMBIT ‘cutoff’ threshold when evaluating the likelihood function. posterior[1][nSamples * (nPar + 2)] = posterior distribution containing nSamples points. Each sample has nPar parameters (physical + derived) along with the their loglike value & posterior probability paramConstr[0][0] to paramConstr[0][nPar - 1] = mean values of the parameters paramConstr[0][nPar] to paramConstr[0][2nPar - 1] = standard deviation of the parameters paramConstr[0][nPar2] to paramConstr[0][3nPar - 1] = best-fit (maxlike) parameters paramConstr[0][nPar4] to paramConstr[0][4nPar - 1] = MAP (maximum-a-posteriori) parameters paramConstr[1][4nPar] = ???? maxLogLike = maximum loglikelihood value logZ = log evidence value logZerr = error on log evidence value context = void pointer, any additional information

Updated on 2023-06-26 at 21:36:52 +0000