I have a very basic question: What is the purpose of mode_compute in the estimation command? How is it related to characterizing the posterior distribution? And how is the mode related to the mean of the posterior distribution?

Thanks!

I have a very basic question: What is the purpose of mode_compute in the estimation command? How is it related to characterizing the posterior distribution? And how is the mode related to the mean of the posterior distribution?

Thanks!

Because the posterior tends to be asymptotically normal, characterizing the proposal distribution of the Metropolis-Hastings algorithm with a normal approximation to the posterior mode is often a good starting point. Mode-compute is about finding the posterior mode for this.