Name

bma_g1 — Bayesian model averaging

CALLING SEQUENCE

rbma_g = bma_g1(y,x,ndraw,burnin,mcmc,g,nvmax)

PARAMETERS

Input

• y = dependent variable vector

• x = explanatory variables

• ndraw = # of draws to carry out

• burnin = # of burn-in MCMC simulation

• mcmc = name of the MCMC algorithm (jump_g or mc3_g)

• g = value of g-prior (default = 1/max(nobs,k^2))

• nvmax = max number of variable allowed in each model

Output

• rbma_g = a results tlist with

  -rbma_g('meth') = 'bma g-prior'

  -rbma_g('nmod') = # of models visited during sampling

  -rbma_g('beta') = bhat averaged over all models

  -rbma_g('mprob') = posterior prob of each model

  -rbma_g('vprob') = posterior prob of each variable

  -rbma_g('model') = indicator variables for each model (nmod x k)

  -rbma_g('yhat') = yhat averaged over all models

  -rbma_g('resid') = residuals based on yhat averaged over models

  -rbma_g('sige') = averaged over all models

  -rbma_g('nobs') = nobs

  -rbma_g('nvar') = # of exogenous

  -rbma_g('y') = y data vector

  -rbma_g('x') = y data vector

  -rbma_g('visit') = visits to each model during sampling (nmod x 1)

  -rbma_g('time') = time taken for MCMC sampling

  -rbma_g('ndraw') = # of MCMC sampling draws

  -rbma_g('burnin')= # of burn-in MCMC simulation

  -rbma_g('gprior')= value of g-prior

DESCRIPTION

Computes bayesian model averaging under g-prior with selection of g-prior as proposed by Fernadez et alii (2001) g = 1/max(n,k^2). This function assumes that x and y are already a matrix and a vector.

EXAMPLE

 rbma_g = bma_g1(grocer_y,grocer_x,grocer_ndraw,grocer_burnin,grocer_tmcmc,grocer_g,grocer_nvmax): this example is taken from bma_g.
 
 
               

AUTHOR

Emmanuel Michaux 2006