mprior-class {BMS} | R Documentation |
An object pertaining to a BMA model prior
An mprior
object holds descriptions and subfunctions pertaining to model priors. The BMA functions bms
and post-processing functions rely on this class.
There are currently five model prior structures built into the BMS package, generated by the following functions (cf. the appendix of vignette(BMS)
):
mprior.uniform.init
: creates a uniform model prior object.
mprior.fixedt.init
: creates the popular binomial model prior object with common inclusion probabilities.
mprior.randomt.init
: creates a beta-binomial model prior object.
mprior.pip.init
: creates a binomial model prior object that allows for defining individual prior inclusion probabilities.
mprior.customk.init
: creates a model prior object that allows for defining a custom prior for each model parameter size.
The following describes the necessary slots:
mp.mode
:A string with a human-readable identifier of the prior.
mp.msize
:A scalar holding the prior model size
mp.Kdist
:A vector holding the prior probabilities for each parameter size, from 0
to K
. (Not necessary for bms
, but for some post-processing functions.
pmp(ki, molddraw, ...):
A sub-function returning log-prior model probability depending on molddraw
(a logical/numeric indicating the positions of regressors included in the model) and model size k
(equivalent to sum(molddraw)
.
As for now, there are no methods defined with class "mprior" in the signature.
Martin Feldkircher and Stefan Zeugner
bms
for creating bma
objects.
Check the appendix of vignette(BMS)
for a more detailed description of built-in priors.
Check http://bms.zeugner.eu/custompriors.php for examples.