Quantcast

ivprobit not converge

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

ivprobit not converge

xueliansharon
Dear all,

I want to estimate an ivprobit model, but it doesn't converge, no matter I use the maximization tech as dfp, bfgs, nr or bhhh, so what can I do to make the model converge? And after some iterations, saying n iterations for instance, the value of log pseudolikelihood doesn't change, can I just set the number of "iterate" option to this number "n" to get the coefficients? Are the parameter estimates gotten by setting the "iterate" option meaningful when the model in fact doesn't converge?

Any help will be greatly appreciated.

Sharon

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ivprobit not converge

Maarten buis
--- On Wed, 18/8/10, xueliansharon wrote:
> I want to estimate an ivprobit model, but it doesn't
> converge, no matter I use the maximization tech as dfp,
> bfgs, nr or bhhh, so what can I do to make the model
> converge?

Stata is telling you that your data does not contain the
necessary information to estimate your model. So you need
to find out what part of your model caused your model to
be too demanding. I would start with the most minimal
model I can come up with. I don't have to believe that that
model makes sense, in fact I probably don't believe that
it does, but it gets the process started. Than step by
step I add variables, till my model no longer converges.
This gives me and idea what the problem is, and often
I see right than that either the model I tried to estimate
did not make sense or really could not be estimated with
my data. The things I am looking for than are near perfect
correlation with some of the other explanatory variables,
whether that variable is (nearly) a perfect linear
combination of other variables (e.g. period - age = cohort),
very small categories when it is a categorical variable,
association with the dependent variable that is so strong
that virtually noone failed after a certain value.

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/
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ivprobit not converge

xueliansharon
Thanks, Maarten.

But when I run the ivprobit model many times with the same codes and data set, it converges sometimes. And the parameter estimates are similar in those converged cases. So are these coefficients reliable?

Best,
Sharon
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ivprobit not converge

Maarten buis
--- On Fri, 20/8/10, xueliansharon wrote:
> But when I run the ivprobit model many times with the same
> codes and data set, it converges sometimes. And the
> parameter estimates are similar in those converged cases.
> So are these coefficients reliable?

I would be reluctant. The maximization algorithms used by
Stata are pretty good, so the fact that -ivprobit- sometimes
does not converge is for me reason enough to suspect that
your model tries to extract more information than the data
contains.

So I will stick to my earlier recomendation, which is to
start with a very simple model, and add complexities step
by step, till you see this non-converging behaviour, and
see if you understand why the complexity you added last
causes trouble. Once you understand why this problem
occurs, you can make an informed decision whether or not
you need to do something about it, whether or not you can
do something about it, and what that something might be.

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/
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Control over directories when installing Stata 11

DE SOUZA Eric-3
I haven't been able to find this in the manuals or FAQs.

On Monday, an administrator will install Stata 11 on my computer. During the installation is it possible to control where the various directories are installed and to name them.

This is what I have with Stata 10:
 sysdir
   STATA:  C:\Program Files\Stata10\
 UPDATES:  C:\Program Files\Stata10\ado\updates\
    BASE:  C:\Program Files\Stata10\ado\base\
    SITE:  C:\Program Files\Stata10\ado\site\
    PLUS:  c:\ado\plus\
PERSONAL:  c:\ado\personal\
OLDPLACE:  c:\ado\

I would like to avoid creating the PLUS, PERSONAL, OLDPLACE directories directly on the C:\drive.
Is there any disadvantage to creating them in the ¨Program Files\Stata11 directory?

Eric de Souza
College of Europe
Brugge (Bruges)
Belgium

*
*   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/
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Control over directories when installing Stata 11

DE SOUZA Eric-3
Something I must have done wrong got my previous post got my previous post on this topic attached to another thread. This is why I am posting it again.

I haven't been able to find this in the manuals or FAQs.

On Monday, an administrator will install Stata 11 on my computer. During the installation is it possible to control where the various directories are installed and to name them.

This is what I have with Stata 10:
sysdir
   STATA: C:\Program Files\Stata10\
 UPDATES: C:\Program Files\Stata10\ado\updates\
    BASE: C:\Program Files\Stata10\ado\base\
    SITE: C:\Program Files\Stata10\ado\site\
    PLUS: c:\ado\plus\
PERSONAL: c:\ado\personal\
OLDPLACE: c:\ado\

I would like to avoid creating the PLUS, PERSONAL, OLDPLACE directories directly on the C:\drive.

Is there any disadvantage to creating them in the ¨Program Files\Stata11 directory? Eric de Souza College of Europe Brugge (Bruges) Belgium

Eric de Souza
College of Europe
Brugge (Bruges)
Belgium


*
*   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/
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Control over directories when installing Stata 11

Eric A. Booth
<>

I haven't had any issues with pointing them elsewhere.  Yes, it is possible to control where any of these directories are located by using -set sysdir-.  I put the following line in my profile to point Stata to where I put my personal ado-files (on a mapped network drive):
**
sysdir set PERSONAL  "Y:\ebooth\projects\ado\personal\\"
**

~ Eric
__
Eric A. Booth
Public Policy Research Institute
Texas A&M University
[hidden email]
Office: +979.845.6754


On Aug 21, 2010, at 7:37 AM, DE SOUZA Eric wrote:

> Something I must have done wrong got my previous post got my previous post on this topic attached to another thread. This is why I am posting it again.
>
> I haven't been able to find this in the manuals or FAQs.
>
> On Monday, an administrator will install Stata 11 on my computer. During the installation is it possible to control where the various directories are installed and to name them.
>
> This is what I have with Stata 10:
> sysdir
>   STATA: C:\Program Files\Stata10\
> UPDATES: C:\Program Files\Stata10\ado\updates\
>    BASE: C:\Program Files\Stata10\ado\base\
>    SITE: C:\Program Files\Stata10\ado\site\
>    PLUS: c:\ado\plus\
> PERSONAL: c:\ado\personal\
> OLDPLACE: c:\ado\
>
> I would like to avoid creating the PLUS, PERSONAL, OLDPLACE directories directly on the C:\drive.
>
> Is there any disadvantage to creating them in the ¨Program Files\Stata11 directory? Eric de Souza College of Europe Brugge (Bruges) Belgium
>
> Eric de Souza
> College of Europe
> Brugge (Bruges)
> Belgium
>
>
> *
> *   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/




*
*   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/
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

RE: Control over directories when installing Stata 11

DE SOUZA Eric-3
Eric to Eric,

Thanks for the information. Somehow when moving from Stata 9 to Stata 10, I must have missed the moment of choice. (The last time I was given temporary local administrator rights whilst installing Stata. So I want to make sure that this time I can download from ssc and such like to a "writable" location given my rights.


Eric de Souza
College of Europe
Brugge (Bruges)
Belgium

-----Original Message-----
From: [hidden email] [mailto:[hidden email]] On Behalf Of Eric Booth
Sent: 21 August 2010 23:19
To: <[hidden email]>
Subject: Re: st: Control over directories when installing Stata 11

<>

I haven't had any issues with pointing them elsewhere.  Yes, it is possible to control where any of these directories are located by using -set sysdir-.  I put the following line in my profile to point Stata to where I put my personal ado-files (on a mapped network drive):
**
sysdir set PERSONAL  "Y:\ebooth\projects\ado\personal\\"
**

~ Eric
__
Eric A. Booth
Public Policy Research Institute
Texas A&M University
[hidden email]
Office: +979.845.6754


On Aug 21, 2010, at 7:37 AM, DE SOUZA Eric wrote:

> Something I must have done wrong got my previous post got my previous post on this topic attached to another thread. This is why I am posting it again.
>
> I haven't been able to find this in the manuals or FAQs.
>
> On Monday, an administrator will install Stata 11 on my computer. During the installation is it possible to control where the various directories are installed and to name them.
>
> This is what I have with Stata 10:
> sysdir
>   STATA: C:\Program Files\Stata10\
> UPDATES: C:\Program Files\Stata10\ado\updates\
>    BASE: C:\Program Files\Stata10\ado\base\
>    SITE: C:\Program Files\Stata10\ado\site\
>    PLUS: c:\ado\plus\
> PERSONAL: c:\ado\personal\
> OLDPLACE: c:\ado\
>
> I would like to avoid creating the PLUS, PERSONAL, OLDPLACE directories directly on the C:\drive.
>
> Is there any disadvantage to creating them in the ¨Program
> Files\Stata11 directory? Eric de Souza College of Europe Brugge
> (Bruges) Belgium
>
> Eric de Souza
> College of Europe
> Brugge (Bruges)
> Belgium
>
>
> *
> *   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/




*
*   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/

*
*   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/
Loading...