Description Usage Arguments Examples
tupoisbsf
performs spectral analysis of time-uncertain time series of count data
using bayesian frequency selection method.
1 |
y |
A vector of observations. |
ti.mu |
A vector of estimates/observed timings of observations. |
ti.sd |
A vector of standard deviations of timings. |
n.sim |
A number of simulations. |
CV |
TRUE/FALSE cross-validation indicator. |
... |
optional arguments: |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 | # Note: Most of models included in tuts package are computationally intensive. In the example
# below I set parameters to meet CRAN<e2><80><99>s testing requirement of maximum 5 sec per example.
# A more practical example would contain N=50 in the first line of the code and n.sim=10000.
#1. Import or simulate the data (simulation is chosen for illustrative purposes):
DATA=simtuts(N=7,Harmonics=c(4,0,0), sin.ampl=c(10,0, 0), cos.ampl=c(0,0,0),
trend=0,y.sd=2, ti.sd=0.2)
y=DATA$observed$y.obs
y=round(y-min(y))
ti.mu=DATA$observed$ti.obs.tnorm
ti.sd= rep(0.2, length(ti.mu))
#2. Fit the model:
n.sim=10
TUPOIS=tupoisbsf(y=y,ti.mu=ti.mu,ti.sd=ti.sd,freqs='internal',n.sim=n.sim,n.chains=2,
CV=TRUE,n.cores=2)
#3. Generate summary results (optional parameters are listed in brackets):
summary(TUPOIS) # Summary results (CI, burn).
summary(TUPOIS,burn=0.2) # Results after 20% of burn-in (CI).
#4. Generate plots and diagnostics of the model (optional parameters are listed in brackets):
plot(TUPOIS,type='periodogram') # spectral analysis (CI, burn).
plot(TUPOIS,type='predTUTS', CI=0.99) # One step predictions (CI, burn).
plot(TUPOIS,type='cv') # 5 fold cross validation (CI, burn).
plot(TUPOIS,type='GR') # Gelman-Rubin diagnostics (CI, burn).
plot(TUPOIS,type='mcmc') # MCMC diagnostics.
plot(TUPOIS,type='lambda') # Realizaitons of lambda.
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.