Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Steven Samuels <sjsamuels@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: Error after -predict- in -apc_ie- |
Date | Mon, 3 Jan 2011 22:46:52 -0500 |
-Wincent Ronggui Huang (ronggui.huang@gmail.com) wrote to me privately that advice that I had given in August about -predict- following - apc_ie- did not work. (-apc_ie- is the age-period-cohort "intrinsic estimation" command. It and apc_cglim ("generalized linear models with equality constraints on age-period-cohort effects") are commands in package -ap by Schulhofer-Wohl and Yang. ("net describe apc, from(http://fmwww.bc.edu/RePEc/bocode/a )".
I got mixed up and gave Wincent information about apc_cglim, not apc_ie. -predict- works after -apc_cglim-, but after -apc_ie- it raises an error. In the example do file that accompanies the package, the error message is:
". predict mu; variable age_0 not found r(111);"I apologize to Wincent for my mistake. I can only suggest that he contact the authors of -apc-.
Steve
On Jan 2, 2011, at 9:55 PM, Wincent wrote:May I ask how to use the three options. It generates an error like this,. use "apc_example_data.dta", clear . apc_ie death_f if age<=90, age(age) period(year) cohort(cohort) family(poisson) link(log) exposure(exp_f) agefx(A) cohortfx(C) periodfx(P) option agefx() not allowed r(198);
On 1 August 2010 20:24, Steve Samuels <sjsamuels@gmail.com> wrote:The -help- for -apc_ie- indicates that it takes -glm- options, and inspection of the source code reveals that it indeed calls -glm-. So all -glm- post-estimation command should work. The answer to your question is "yes", provided you use the agefx(), periodfx() and cohortfx() options (not mentioned in the -help-) and keep the generated variables. Run the second example in the accompanying do file (net describe apc, from(http://fmwww.bc.edu/RePEc/bocode/a) and do _not_ drop the generated variables _A* _P* and _C*. Then -predict- should work as expected. Steve
* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/