spec.tbreak: Detect change points in spectrum across time

In astrostat/Automark: Automatic Change Point Detection in the Spectra of High-energy Astrophysical Sources

Description

Detect change points in spectrum across time. Between the change points, the spectra are fitted by spec. The number of change points is determined via the minimum description length principle.

Usage

spec.tbreak(x1, x2, y, A, delta.t, delta.w, display = TRUE,
opt = get.opt.spec(), cpus = 4, assign.emiss = NA,
emdl = TRUE, v1 = 1, v2 = 1, simple = TRUE,
max.B = 4, min.span = 5)

Arguments

x1	A vector containing time points (mid-points of time bins)
x2	A vector containing wavelengths (mid-points of wavelength bins)
y	A matrix containing the observed counts. Its dimension is c(number of wavelengths, number of time points)
A	A vector containing the effective areas
delta.t	Bin width of time
delta.w	Bin width of wavelength
display	If TRUE, progress will be shown.
opt	An option list returned by get.opt.spec. This is used in the homogeneous spectrum fitting between change points. If assign.emiss=NULL, nlambda will be automatically set to 10000 if it is set smaller than 10000. If this automatic change is not desired, one can set lam.ratio.len=-1. (lam.ratio.len is a useless parameter if assign.emiss=NULL)
cpus	Number of cores to use for parallel execution.
assign.emiss	Indices of wavelength bins that are allowed to have emission line. If NA, assign_emiss will be set to 1:length(y).
emdl	If TRUE, extended MDL, rather than MDL, is used.
v1	An ad-hoc parameter for tuning the strength of the penalty in fitting each time-homogenous spectrum between change points. That means, (v * penalty) will be the ultimate penalty used in the MDL / extended MDL. More penalty leads to a smoother spectrum and/or fewer emission lines. It is set to 1.0 by default.
v2	Similar to v1, but for tuning the strength of the penalty for change points across time. Larger v2 leads to fewer change points across time.
simple	If TRUE, a simplified output is returned. That is, results of the grid search (beta_cube) will be not be output.
max.B	The maximum number of time segments allowed across time.
min.span	The minimum number of time bins allowed between breaks.

Value

B	Number of time segments in the output fit
best.fit	List of time-homogeneous fits (in the output fit) corresponding to different time segments. E.g., best.fit[[1]] corresponds to the first time segment.
breaks	Indices of breakpoints. E.g., c(1,10) represents the first time segment spans time bins 1,2,...,9 and the second one spans the rest.
content	Information of the fitting process (debug use)

References

Raymond K. W. Wong, Vinay L. Kashyap, Thomas C. M. Lee and David A. van Dyk (2016). Detecting Abrupt Changes in the Spectra of High-energy Astrophysical Sources. The Annals of Applied Statistics, 10(2), 1107-1134.

See Also

spec, plotspec

Examples

library(Automark)       # load library
data(dat)               # load example data
## Not run: 
out <- spec.tbreak(dat$x1, dat$x2, dat$y, dat$A, dat$delta.t, dat$delta.w, cpus=5)
plotspec(dat$x2, out$best.fit[[1]])
plotspec(dat$x2, out$best.fit[[2]], np=F, col="red")

## End(Not run)

astrostat/Automark documentation built on Dec. 4, 2019, 12:30 a.m.