Xavier -
Interesting enough, I tried -nl- with cluster() and it didn't give an error
- but it gave the same standard errors as it did without the cluster()
option. This sounds as if Stata Corp. meant for -nl- to have a cluster()
option, but forgot to put it in!
One way to get around this would be to write your own ml-program with
normally distributed errors around the mean which is your nonlinear function
of the parameters. Stata's ml processor allows for robust standard errors
with clustering.
Al Feiveson
-----Original Message-----
From: VIVIAND Xavier [mailto:[email protected]]
Sent: Monday, September 16, 2002 10:46 AM
To: [email protected]
Subject: st: nl estimation with repeated data
Dear StataList,
I would like to fit an Emax model (E0-((Emax.C^n)/(EC50^n+C^n)) where
several data are available for each patient.
Unfortunately, the nl procedure does not include a cluster option.
Does someone know how to do it ?
Sincerely yours,
Xavier VIVIAND
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
