Here is a question where I had posted separate answers to my own question:
Can I semi-automate MCMC convergence diagnostics to set the burn-in length?
But then was requested to combine them. Which is this the preferred approach?
Here is a question where I had posted separate answers to my own question:
Can I semi-automate MCMC convergence diagnostics to set the burn-in length?
But then was requested to combine them. Which is this the preferred approach?
My opinion on this issue:
I think you should provide only one answer. I cannot think of a reason to do otherwise. You can always split up the answer into sections such as "Method 1", "Theoretical Solution", "Illustrative Code" etc.