Given raw observational CMB data contaminated by instrumemtal noise
and galactic foregrounds, how does one get to data that is useful
for a cosmological analysis? While the answer is simple
algebraically, an exact solution is computationally
intractable. Gibbs sampling is one answer to this problem. Some aspects
of Gibbs sampling have already been treated by Eirik, Unni and Sigurd
this year. I hope to fill in the remaining gap: What, computationally,
does one have to do in order to sample from the joint posterior of the
CMB and the foregrounds.
I'll follow the usual recipe of first doing a basic introduction and
then talk about current developments.