Metropolis-within-Gibbs

[@NuclearSpinach]

Possibly a stupid question: how do you get the full conditional proportionality first shown at 9:39? How do we "make it proportional to the target distribution of interest"?

[@NuclearSpinach]

Same question at 14:47

[@NuclearSpinach]

Is it as simple as p(theta k given everything else) = p(theta, alpha, beta | y) / p(theta_{-k}, alpha, beta | y)? I.e., is it simply using f(x|y) = f(x,y)/f(y), and here the constant would be f(y)?

[@jaradniemi]

@@NuclearSpinach Yes, that's it!

[@GaryMirams]

Really helpful - thanks. Any chance you could provide/link to the bladder cancer dataset and the code you discussed?

[@valerioguarrasi7203]

Where can we find the code and the dataset? Thankyou

[@caoyuanli1877]

Slide 5 shall be \alpha -1