Direct sampling or Gibbs?

Topic: Bayesian linear regression · more questions on this topic
‹ The inverse-Gamma shape trap Mean response or new response? ›

Write one algorithm using p(\sigma^2\mid{\boldsymbol{y}})p({\boldsymbol{\beta}}\mid\sigma^2,{\boldsymbol{y}}) and another alternating the two full conditionals. For each algorithm state whether the output pairs are iid or form a Markov chain, and identify the appropriate inverse-Gamma shape and sum of squares.

This exercise has no worked answer.