Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: atrho0() in mvprobit
From
Maarten Buis <[email protected]>
To
[email protected]
Subject
Re: st: atrho0() in mvprobit
Date
Wed, 31 Aug 2011 09:55:02 +0200
On Tue, Aug 30, 2011 at 11:48 PM, Zhi Su wrote:
> I estimate a five equation model using "mvprobit". And Stata return
> errors "could not calculate numerical derivatives flat or
> discontinuous region encountered".
> Then I am thinking to set the starting values for the off-diagonal
> elements of the rho matrix. So I use "mvprobit" to estimate four among
> these five equations and get rho21, rho31, rho41, rho32,rho42, rho43
> from e(rhoij) .
The maximum likelihood function does not maximize with respect to rho
but with respect to the Fisher's z transformation of rho, so your
starting values should contain those. See e.g.: M.L. Buis (2011)
"Stata tip 97: Getting at rhos and sigmas", The Stata Journal, 11(2),
pp. 315-317.
Hope this helps,
Maarten
--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany
http://www.maartenbuis.nl
--------------------------
*
* 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/