tuls: Spectral analysis of time-uncertain time series using...

In PeterFranke/tuts: Time Uncertain Time Series Analysis

Description

tuls computes multiple power estimates using the Lomb-Scargle algorithm and simulated realizations of timings of observations. Timings are simulated with normal distribution ti~N(ti.mu,ti.sd), and sorted in ascending order to ensure non-overlapping feature of observations.

Usage

tuls(y, ti.mu, ti.sd, ...)

Arguments

y	A vector of observations.
ti.mu	A vector of estimates of timings of observations.
ti.sd	A vector of standard deviations of timings.
...	A list of optional parameters. The list contains thinning oversamlping parameter, the default value is ofac=4, confidence interval, the default value is CI=0.99, and the number of simulated timings, wiith the default vale set to n.sim=1000.
n.sim	A number of simulations.

References

https://en.wikipedia.org/wiki/Least-squares_spectral_analysis

See Also

http://cran.r-project.org/package=Bchron

Examples

# Import or simulate the data (simulation is chosen for illustrative purposes):
DATA=simtuts(N=50,Harmonics=c(10,30,0), sin.ampl=c(10,10, 0), cos.ampl=c(0,0,0),
trend=0,y.sd=3, ti.sd=1)
y=DATA$observed$y.obs
ti.mu=DATA$observed$ti.obs.tnorm
ti.sd= rep(1, length(ti.mu))

# (Set parameters and ) run multiple Lomb-Scargle periodograms:
TULS=tuls(y=y,ti.mu=ti.mu,ti.sd=ti.sd)       # Optional parameters: ofac, CI, n.sim

# Plot the results:
plot(TULS)

# Obtain list of frequencies for which spectral power exceeds confidence interval prespecified in the tuts_ls class
summary(TULS)

PeterFranke/tuts documentation built on May 30, 2019, 6:24 a.m.