Method MCMC/mhsample
MHSAMPLE.M The Metropolis - Hastings algorithm
The Metropolis-Hastings algorithm: Samples a given likelihood function.
Warning: The function mhsample does not performs sanity
checks on the inputs. It assumes that the given
data-sets are in frequency domain and correctly
defined.
CALL: b = MCMC.mhsample(pl)
INPUTS: pl - parameter list
OUTPUTS: b - pest object contatining estimated information
NOTE: The resulting pest object has its 'chain' field defined
with the MCMC chains. The chain is a (# of samples x 2 +
# of parameters) numerical matrix. In the first column
the log-likelihood values are stored, while in the second
the SNR. The rest of them contain the parameter values.
If the log-likelihood is a 'mfh' object, due to
Matlab limitations, the SNR column contains also the
log-likelihood values.
Parameters Description
MN/NK 2013
Method Details |
|
Access |
public |
Defining Class |
MCMC |
Sealed |
0 |
Static |
1 |
|
Method: MCMC/logDecision |
|
Method: MCMC/plotLogLikelihood |
 |
©LTP Team