This is like the summary.lm() except Bayesian statistics reports different metrics like probability intervals (PIs) for the coefficients. 

### input: lm.bayes object
### output: a table with means and 95% PIs for the parameters b0,b1,...,bp, sigma^2 (one row each).  It returns (lower_bound, mean, upper_bound) where lower and upper bound are the 95% PIs.

summary.bayes=function(reg2)
{  N=nrow(reg2)
   k=(ncol(reg2))
   summ=matrix(0,k,3)
   for(i in 1:k)
   { summ[i,]= c(sort(reg2[,i])[.025*N], mean(reg2[,i]), sort(reg2[,i])[.975*N])
   }
   summ
}

### Example:   use lm.bayes to create a regression object: "reg2"
summary.bayes(reg2)