[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Correct code for Poi's QUAIDS model

From   "Brian P. Poi" <>
Subject   Re: st: Correct code for Poi's QUAIDS model
Date   Fri, 2 May 2008 10:34:15 -0500 (CDT)

On Fri, 2 May 2008, wrote:


Can somebody mail me the correct code for Poi's (2002) QUAIDS model. I downloaded the code from but I suspect there are some mistakes in this code as I keep getting error message like 'conformability error' or 'the subcommand post has been moved from estimates to ereturn' when I run the code. I would like to be able to replicate his results to see how this works in Stata.

I have emailed Lavanya a zip file containing programs to fit a QUAIDS model.

I have never had anyone report a conformability error when running my QUAIDS code, though the error messages pertaining to the command -estimates post- moving to -ereturn post- is understandable, since the -estimates- commands were rewritten and renamed -ereturn- in Stata 8 or Stata 9. For those who want replicate the steps I took in that Stata Journal article, just change the commands -estimates post- and -estimates display- to -ereturn post- and -ereturn display-, respectively.

Also, I'm using Stata 10 and the code for Poi's example was written in Stata version 7 (all the .ado files say version 7).. could this cause a problem in the code running properly in State version 10?

For those who have access to Stata 10, I strongly recommend using -nlsur- instead of the programs in my 2002 Stata Journal article. The last example in the reference manual entry for -nlsur- shows how to fit the basic AIDS model, and it can be adapted to other demand systems as well. Using -nlsur- requires just one program instead of the small handful that the older method using -ml- requires, and -nlsur- tends to fit these models much more quickly.

-- Brian Poi

* For searches and help try:

© Copyright 1996–2022 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index