Description Usage Arguments Value Examples
Simulation experiment of change-point detection with unknown post-change distributions using the Bayesian stopping rule.
1 2 | sim.detect.bayes(GEN0, GEN1, th0, detect.bayes, nulower = NULL,
nuupper = NULL, ...)
|
GEN0 |
A function that take no argument and return an observation from the pre-change model. |
GEN1 |
A function that takes a value of the unknown parameter in the post-change model and return an observation. |
th0 |
Numeric: True value of the unknown parameter in the post-change model. |
detect.bayes |
A function that implements the Bayesian stopping rule.
Currently only
|
nulower, nuupper |
Optional nonnegative numerics: The earliest and latest
time of change-point based on prior belief. The default is |
... |
Additional arguments to be passed to |
A named numeric vector with components
is.FA
A numeric of 1 or 0 indicating whether a false alarm
has been raised.
DD
A positive numeric: The delay to
detection.
CT
A positive numeric: The computation time.
LER
A numeric in (0,1)
: The low ESS rate, i.e., the proportion of iterations that ESS drops below c*n
.
AAR
A numeric in (0,1)
: The average acceptance rate of the Metropolis-Hastings sampling in the move step. NaN
if ESS never drops below c*n
.
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 | ##Change from N(0,1) to 2*N(0,1)+1 occurs between 10 and 25.
##The mean and standard deviation of the post-change normal distribution are unknown.
GEN0=function() rnorm(1)
GEN1=function(th) th[2]*rnorm(1)+th[1]
ULP1=function(x,th) -(x-th[1])^2/2/th[2]^2
GLP1=function(x,th) -(x-th[1])/th[2]^2
LLP1=function(x,th) -1/th[2]^2
ULP0=function(x) ULP1(x,c(0,1))
GLP0=function(x) GLP1(x,c(0,1))
LLP0=function(x) LLP1(x,c(0,1))
par0=log(2*pi)/2
par1=function(th) log(2*pi)/2+log(th[2])
#using hyvarinen score
sim.detect.bayes(GEN0=GEN0,GEN1=GEN1,th0=c(1,2),detect.bayes=detect.bayes,
nulower=10,nuupper=25,alpha=0.1,thlower=c(-Inf,0),
GLP0=GLP0,LLP0=LLP0,GLP1=GLP1,LLP1=LLP1)
#using log score, normalizing constant unknown
#sim.detect.bayes(GEN0=GEN0,GEN1=GEN1,th0=c(1,2),detect.bayes=detect.bayes,
#nulower=10,nuupper=25,alpha=0.1,thlower=c(-Inf,0),
#score="log",ULP0=ULP0,ULP1=ULP1,lenx=1)
#using log score, normalizing constant known
sim.detect.bayes(GEN0=GEN0,GEN1=GEN1,th0=c(1,2),detect.bayes=detect.bayes,
nulower=10,nuupper=25,alpha=0.1,thlower=c(-Inf,0),
score="log",ULP0=ULP0,ULP1=ULP1,par0=par0,par1=par1)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.