Function used to assure that an object is of class 'BayesianSetup'. If you pass a function, it is coverted to an object of class 'BayesianSetup' (using createBayesianSetup) before it is returned.
checkBayesianSetup(bayesianSetup, parallel = F)
| bayesianSetup | either object of class bayesianSetup or a log posterior function |
|---|---|
| parallel | if bayesianSetup is a function, this will set the parallelization option for the class BayesianSetup that is created internally. If bayesianSetup is already a BayesianSetup, then this will check if parallel = T is requested but not supported by the BayesianSetup. This option is for internal use in the samplers |
The recommended option to use this function in the samplers is to have parallel with default NULL in the samplers, so that checkBayesianSetup with a function will create a bayesianSetup without parallelization, while it will do nothing with an existing BayesianSetup. If the user sets parallelization, it will set the approriate parallelization for a function, and check in case of an existing BayesianSetup. The checkBayesianSetup call in the samplers should then be followed by a check for parallel = NULL in sampler, in which case paralell can be set from the BayesianSetup