summary.bma: Summary Statistics for a 'bma' Object

In BMS: Bayesian Model Averaging Library

Summary Statistics for a 'bma' Object

Description

Returns a vector with summary statistics for a 'bma' object

Usage

info.bma(bmao)

## S3 method for class 'bma'
summary(object, ...)

Arguments

bmao	same as object
object	a list/object of class 'bma' that typically results from the function bms (see bms for details)
...	further arguments passed to or from other methods

Details

info.bma is equivalent to summary.bma, its argument bmao conforms to the argument object

Value

A character vector summarizing the results of a call to bms

Mean no. of Regressors	the posterior mean of model size
Draws	the number of iterations (ex burn-ins)
Burnins	the number of burn-in iterations
Time	the time spent on iterating through the model space
No. of models visited	the number of times a model was accepted (including burn-ins)
Modelspace	the total model space 2^K
list(list("2^K"))	the total model space 2^K
Percentage visited	No. of models visited/Modelspace*100
Percentage Topmodels	number of times the best models were drawn in percent of Draws
Corr. PMP	the correlation between the MCMC frequencies of the best models (the number of times they were drawn) and their marginal likelihoods.
No. Obs.	Number of observations
Model Prior	a character conforming to the argument mprior of bms, and the expected prior model size
g-prior	a character corresponding to argument g of function bms
Shrinkage-Stats	Posterior expected value und standard deviation (if applicable) of the shrinkage factor. Only included if argument g.stats of function bms was set to TRUE

Note

All of the above statistics can also be directly extracted from the bma object (bmao). Therefore summary.bma only returns a character vector.

See Also

bms and c.bma for functions creating bma objects, print.bma makes use of summary.bma.

Check http://bms.zeugner.eu for additional help.

Examples

  data(datafls)

  m_fixed=bms(datafls,burn=1000,iter=3000,user.int=FALSE, )
  summary(m_fixed)
   
  m_ebl=bms(datafls,burn=1000,iter=3000,user.int=FALSE, g="EBL",g.stats=TRUE)
  info.bma(m_ebl)

BMS documentation built on Aug. 9, 2022, 5:08 p.m.