R help - Feb 2009 - How to calculate the marginals of the data for some model

Hi all,

I have modeled my observations/data,say, D=y(i) (i=1,2,...,20000), as a
univariate non-gausian distribution, gamma distribution y(i)~gamma(a,b) for
example, and I have also obtained N samples of a, b via WinBUGS(Gibbs
sampling), my question is: how to calculate the marginal density of D,
m(D|a,b)= integration of p(D|a,b)*p(a,b) or the log marginal
ln(m(D|a,b))=ln(f(D|a-bar,b-bar)+ln(p(a,b))-ln(p(a-bar,b-bar|D)) when we
don't have the closed-form equation to express p(a,b|D). Is there any
package in R to do the integration? Any suggestion will be highly
appreciated. Thanks,

Power

	[[alternative HTML version deleted]]