bigGP (version 0.1.8)

SN2011fe: SN2011fe Supernova Dataset

Description

SN2011fe is a dataset of flux values and estimated standard errors, as a function of phase and wavelength, from the SN 2011fe supernova event. Data were collected over multiple nights (phases) and multiple wavelengths.

Format

The SN2011fe object is a data frame containing the following columns:

phase:

time of measurement in days.

wavelength:

wavelength of measurement in \(\mbox{\AA}\).

flux:

flux measurement in \(\mbox{erg}\ \mbox{s}^{-1}\ \mbox{cm}^{-2}\ \mbox{\AA}^{-1}\).

fluxerror:

estimated standard deviation of the error in measurement of the flux.

phaseindex:

1-based index value of the time of measurement [check this]

logwavelength:

log of wavelength.

The SN2011fe_newdata object is a data frame of prediction points on a fine grid of phases and wavelengths. The columns correspond to the phase and wavelength columns in SN2011fe but the initial 'p' stands for 'prediction'.

The SN2011fe_mle object is the output from maximum likelihood fitting of the parameters of a statistical model for the dataset, with the par element containing the MLEs.

The objects labeled '_subset' are analogous objects for a small subset of the dataset feasible to be fit without parallel processing.

The SN2011fe_initialParams object is a set of starting values for the maximum likelihood fitting.

The functions SN2011fe_meanfunc, SN2011fe_predmeanfunc, SN2011fe_covfunc, SN2011fe_crosscovfunc, and SN2011fe_predcovfunc are functions for calculating the various mean vectors and covariance matrices used in the statistical analysis of the dataset. Users will need to create analogous functions for their own kriging problems, so these are provided in part as templates.

Warning

Note that the SN2011fe_newdata set of prediction points was chosen to ensure that the points were not so close together as to result in numerically non-positive definite covariance matrices when simulating posterior realizations.

References

For more details on the dataset, see: R. Pereira, et al., 2013, "Spectrophotometric time series of SN 2011fe from the Nearby Supernova Factory," Astronomy and Astrophysics, accepted (arXiv:1302.1292v1), DOI: tools:::Rd_expr_doi("10.1051/0004-6361/201221008").

For more details on the statistical model used to fit the data, see:

Paciorek, C.J., B. Lipshitz, W. Zhuo, Prabhat, C.G. Kaufman, and R.C. Thomas. 2015. Parallelizing Gaussian Process Calculations in R. Journal of Statistical Software, 63(10), 1-23. tools:::Rd_expr_doi("10.18637/jss.v063.i10").

or

Paciorek, C.J., B. Lipshitz, W. Zhuo, Prabhat, C.G. Kaufman, and R.C. Thomas. 2013. Parallelizing Gaussian Process Calculations in R. arXiv:1305.4886. https://arxiv.org/abs/1305.4886.

See Also

krigeProblem-class

Examples

if (FALSE) {
doSmallExample <- TRUE

if(require(fields)) {
if(doSmallExample){
  SN2011fe <- SN2011fe_subset
  SN2011fe_newdata <- SN2011fe_newdata_subset
  SN2011fe_mle <- SN2011fe_mle_subset
  nProc <- 3
} else {
# users should select number of processors based on their system and the
# size of the full example
nProc <- 210 
}

n <- nrow(SN2011fe)
m <- nrow(SN2011fe_newdata)
nu <- 2
inputs <- c(as.list(SN2011fe), as.list(SN2011fe_newdata), nu = nu)

prob <- krigeProblem$new("prob", numProcesses = nProc, n = n, m = m,
predMeanFunction = SN2011fe_predmeanfunc, crossCovFunction = SN2011fe_crosscovfunc,
predCovFunction = SN2011fe_predcovfunc, meanFunction =
SN2011fe_meanfunc, covFunction = SN2011fe_covfunc,  inputs = inputs,
params = SN2011fe_mle$par, data = SN2011fe$flux, packages = c("fields"))

prob$calcLogDens()
}
}