[Limdep Nlogit List] Generalised random effects model

Tue Jun 14 21:48:09 AEST 2016

Dear Professor,

Actually I using the commands below

FRONTIER;COST;Lhs=Inputs;Rhs=one,Outputs$
FRONTIER;COST;Lhs=Inputs;Rhs=one,Outputs
        ;Panel;CSN;Halton;Pts=100;Keep$

so I have already incorporated the ;cost command in the syntax and every time I am fitting the model I am running the first line  so without the GTRE setup. However the 141 warning is a general problem in all model specifications e.g TFE,TRE, LATENT CLASS, Pitt and Lee RE etc. 

Should I search for alternative measures of outputs? 

Many thanks in advance!

Best wishes,

Maria
Ms Maria PapadimitriouGraduate Teaching Assistant in EconomicsPhD Candidate in EconomicsLancaster University Charles Carter Building, floor D, Lancaster United Kingdom | T: +44 (0)7597745806  | Web page: http://www.lancs.ac.uk/lums/people/all/maria-papadimitriou/| E-mail address m.papadimitriou at lancaster.ac.ukCallSend SMSAdd to SkypeYou'll need Skype CreditFree via Skype

> Date: Tue, 14 Jun 2016 07:30:51 -0400
> From: wgreene at stern.nyu.edu
> To: limdep at limdep.itls.usyd.edu.au
> Subject: Re: [Limdep Nlogit List] Generalised random effects model
> 
> Maria.  The model may not be appropriate for your data.  Two
> program possibilities:
> (1) If you are analyzing cost data, add ;COST to the command.
> (2) You must fit the model without the CSN feature first.  I.e.,
> FRONTIER ; Lhs = y ; Rhs = X $
> FRONTIER ; Lhs = y ; Rhs = X ; Panel ; CSN ; ... $
> This is every time you fit the model.  It requires both commands.
> /Bill Greene
> 
> On Mon, Jun 13, 2016 at 11:20 AM, Maria Papadimitriou <
> mariapapadim at windowslive.com> wrote:
> 
> > Dear Nlogit Users,
> >
> > I am receiving an error message like this:
> >
> > Warning  141: Iterations:current or start estimate of sigma nonpositive
> > Line search at iteration  2 does not improve the function
> > Exiting optimization
> > With <   4 iterations, this may not be a good solution to
> > the optimization. The log-likelihood is flat.  Try refit-
> > ting with ;Output=3 and examining the derivatives.
> >
> > Also when I am trying to run the True Random Effects model I am receiving
> > again the same warning 141 and the comment below:
> >  Line search at iteration  3 does not improve the function
> > Exiting optimization
> > With <   4 iterations, this may not be a good solution to
> > the optimization. The log-likelihood is flat.  Try refit-
> > ting with ;Output=3 and examining the derivatives.
> >
> > Even
> >  in cases where I am using a latent class model,a Random effects with
> > Mundlak's treatment, a True Fixed effects model warning 141 exists.
> > Can
> >  you please give me some suggestions on what to do further to fix the
> > issue? Without being sure, maybe this is related to the nature of the
> > data (lots of zero values) or to model misspecification (For the moment I
> >  am using a linear function).
> >
> > Thank you a lot in advance!
> >
> > Ms Maria PapadimitriouGraduate Teaching Assistant in EconomicsPhD
> > Candidate in EconomicsLancaster University Charles Carter Building, floor
> > D, Lancaster United Kingdom | T: +44 (0)7597745806  | Web page:
> > http://www.lancs.ac.uk/lums/people/all/maria-papadimitriou/| E-mail
> > address m.papadimitriou at lancaster.ac.ukCallSend SMSAdd to SkypeYou'll
> > need Skype CreditFree via Skype
> >
> > From: mariapapadim at windowslive.com
> > To: limdep at limdep.itls.usyd.edu.au
> > Subject: Generalised random effects model
> > Date: Thu, 9 Jun 2016 15:11:00 +0300
> >
> >
> >
> >
> >
> >
> > Dear Nlogit
> > users,
> >
> >
> >
> > I am struggling to compose the code for a generalised random effects
> > model. I
> > would like to explore persistent and transient inefficiency in the Higher
> > Education Sector. Unfortunately I cannot find any existing routines for a
> > generalised random effects model in Nlogit (I am using the Nlogit 5
> > version).
> >  I
> > want to do something similar with what Kumbjakar et al (2011) have done in
> > their
> > model 6 or what Tsionas and Kumbhakar have applied in the JAE in (2012),
> > but
> > without the Bayesian framework.
> >  If anybody can help me with that it would be
> > extremely vital for me since I have stuck on the programming part. Also I
> > am
> > not sure if the version Nlogit 5 is supporting a generalised random effects
> > model or not.
> >
> >
> >
> >
> > I am
> > looking forward to hearing from you asap!
> >
> >
> >
> > Best wishes,
> >
> > Maria
> >
> >
> > Ms Maria PapadimitriouGraduate Teaching Assistant in EconomicsPhD
> > Candidate in EconomicsLancaster University Charles Carter Building, floor
> > D, Lancaster United Kingdom | T: +44 (0)7597745806  | Web page:
> > http://www.lancs.ac.uk/lums/people/all/maria-papadimitriou/| E-mail
> > address m.papadimitriou at lancaster.ac.ukCallSend SMSAdd to SkypeYou'll
> > need Skype CreditFree via Skype
> >
> >
> >
> > Get a signature like this: Click here!
> > _______________________________________________
> > Limdep site list
> > Limdep at limdep.itls.usyd.edu.au
> > http://limdep.itls.usyd.edu.au
> >
> 
> 
> 
> -- 
> William Greene
> Department of Economics
> Stern School of Business, New York University
> 44 West 4 St., 7-90
> New York, NY, 10012
> URL: http://people.stern.nyu.edu/wgreene
> Email: wgreene at stern.nyu.edu
> Ph. +1.212.998.0876
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au
> http://limdep.itls.usyd.edu.au