It was briefly mentioned in Section XXXXXXXXXXthat the DSO created by the Stan command stan_model could be reused in multiple calls of the Stan sampling command. For this exercise, your goal is to...

It was briefly mentioned in Section 14.3.1 that the DSO created by the Stan command stan_model could be reused in multiple calls of the Stan sampling command. For this exercise, your goal is to implement in Stan the Monte Carlo approximation of power described in Section 13.2.4, p. 372. It is straight forward to implement it in Stan using the exactly analogous function definitions, but this approach, while logically correct, is inefficient because it will recompile the Stan model every time a new simulated data set is analyzed. The real exercise of this exercise is to modify the function that does the Stan analysis so that it can either compile the Stan model afresh or reuse an existing DSO. In your modified version, compile the Stan model outside the loop that runs the analyses on simulated data, and pass the compiled DSO as an argument into the function that runs an analysis on a simulated data set . Show your code, with explanatory comments.