Thursday, February 16, 2006

Re: st: Re: slow

At 05:28 PM 2/16/2006, Michael Blasnik wrote: >I think you are stuck with my original option 3 -- get a faster >computer. I'm not that familiar with xtabond2, but it's probably a >good guess given what little I do know that they require a lot of >calculations. One other possibility- if you are allocating more >memory than is physically available you could speed things up quite >a bit by adding RAM. If the hard drive is being used a lot during >the do file, then that may be the problem.

The program does suggest one other possibility if you aren't doing it already: use Stata 9. The docs say "xtabond2 is written to take advantage of the Mata programming language first included in Stata 9.0. It works in Stata 7 and 8, but runs much faster in Stata 9 or later."

Also, some programs that are having convergence problems will run better if you use the "difficult" option. I don't know if xtabond2 is one of them or not.

As a sidelight, there is one other little thing in the docs I like:

"Citation xtabond2 is not an official Stata command. It is a free contribution to the research community, like a paper. Please cite it as such: Roodman, D. 2005. xtabond2: Stata module to extend xtabond dynamic panel data estimator. Center for Global Development, Washington."

I think that is a good idea, at least for a major estimation program like xtabond2. A lot of authors just seem to assume that some nameless Stata programmer received enormous amounts of money for this program they are running.

------------------------------------------- Richard Williams, Notre Dame Dept of Sociology OFFICE: (574)631-6668, (574)631-6463 FAX: (574)288-4373 HOME: (574)289-5227 EMAIL: Richard.A.Williams.5@ND.Edu WWW (personal): WWW (department):

* * For searches and help try: * * *


Links to this post:

Create a Link

<< Home

This page is powered by Blogger. Isn't yours?