
Sequential Metropolis-within-Gibbs (SMWG)
Acceptance Rate | The optimal acceptance rate is 44%, and is based on the univariate normality of each marginal posterior distribution. The observed acceptance rate may be suitable in the interval [15%,50%]. |
Applications | This algorithm is applicable with state-space models (SSMs), including dynamic linear models (DLMs). |
Difficulty | This algorithm is relatively easy for a beginner when the proposal variance has been tuned with the SAMWG algorithm. Otherwise, it may be tedious for the user to tune the proposal variance. |
Final Algorithm? | Yes. |
Proposal | Componentwise. |
The Sequential Metropolis-within-Gibbs (SMWG) algorithm is the non-adaptive version of the Sequential Adaptive Metropolis-within-Gibbs (SAMWG) algorithm, and is used for final sampling of state-space models (SSMs).



