How do I get this program to run faster?

A variety of ${\mathfrak{A}melia}$ options can significantly increase the speed of the program. In the order in which we suggest you try them, these are,

1. Install the dynamic library patch and set _AMpatch=1. This is by far the most important step (See section 9).

2. If you have fully observed covariates, Amelia will be faster if you identify them with _AMfully.

3. You could implement the EMis algorithm with priors. See description of _AMempri and related globals in reference material on amelia. This is useful for problem data sets.

4. Buying enough RAM so you do not need to take advantage of the virtual memory feature of Gauss is helpful. Virtual memory slows down the program considerably.

5. You can use the asymptotic normal approximation, and eliminate the importance sampling refinement, by setting _AMmthd=2.

6. You can set _AMdrTol to a larger number. The default is 0.00001; you can try 0.0001 or larger.

7. For especially hard problems, you may wish to change from EMis to EM (by setting _AMmthd=1). In theory, the cost of this choice is that your standard errors will be too small (by ignoring estimation error in $\mu$ and $\Sigma$) and some quantities of interest will be biased (such as inferences about quantities other than means); the extent of the problem in practical examples can be innocuous or serious, depending on application.