Olga,
You have to be clear about whether you are
doing ex ante or ex-post forecasting. How you
proceed depends on the approach you are using.
The protocol is quite different depending upon
which approach you are using.
- Best,
Bob Yaffee
Robert A. Yaffee, Ph.D.
Research Professor
Shirley M. Ehrenkranz
School of Social Work
New York University
home address:
Apt 19-W
2100 Linwood Ave.
Fort Lee, NJ
07024-3171
Phone: 201-242-3824
Fax: 201-242-3825
yaffee@nyu.edu
----- Original Message -----
From: Olga Gorbachev Melloni <olga_gorbachev@hotmail.com>
Date: Thursday, August 24, 2006 2:40 pm
Subject: Re: st: what is SIGMA2 _const in arima
> First, thank you very much for explaining what sigma2 _const means
> (it would
> have been great if STATA actually mentioned it in the manual with
> applicable
> equations...).
>
> I am sorry for being persistant, but what do you mean by your
> suggestion?
> "If you want to predict with a covariate, you first have to
> forecast the
> covariate into the forecast horizon, over which
> you'll forecast your y variable. The y variable will have missing
> values.
> You can then forecast over that horizon with the y variable."
>
> What I understand from it is:
>
> I can construct y_hat using the coefficients from the regression,
> then I can
> predict u_hat=y-y_hat
> then since I know coefficient on MA1, I can constract epsilons,
> assuming
> that u_hat_minyear=epsilon_minyear, or that epsilon_beforeminyear=0
>
> Is that what you meant?
>
> Thank you in advance,
>
> Olga Gorbachev Melloni
> Ph.D. student in Economics at Columbia University
> (212)866-6091
> 502 W122 Street, apt. 62
> NY, NY 10027
>
>
>
> I am sorry
>
> >From: Robert A Yaffee <bob.yaffee@nyu.edu>
> >Reply-To: statalist@hsphsun2.harvard.edu
> >To: statalist@hsphsun2.harvard.edu
> >Subject: Re: st: what is SIGMA2 _const in arima
> >Date: Thu, 24 Aug 2006 14:11:42 -0400
> >
> >Olga,
> > If you want to predict with a covariate, you first have to
> >forecast the covariate into the forecast horizon, over which
> >you'll forecast your y variable. The y variable will have
> >missing values. You can then forecast over that horizon with
> >the y variable.
> > - Regards,
> > Bob Yaffee
> >
> >
> >Robert A. Yaffee, Ph.D.
> >Research Professor
> >Shirley M. Ehrenkranz
> >School of Social Work
> >New York University
> >
> >home address:
> >Apt 19-W
> >2100 Linwood Ave.
> >Fort Lee, NJ
> >07024-3171
> >Phone: 201-242-3824
> >Fax: 201-242-3825
> >yaffee@nyu.edu
> >
> >----- Original Message -----
> >From: Olga Gorbachev Melloni <olga_gorbachev@hotmail.com>
> >Date: Thursday, August 24, 2006 1:59 pm
> >Subject: Re: st: what is SIGMA2 _const in arima
> >
> > > Unfortunately your suggestion (predict x, xb) doesn't work, since
> > > as I
> > > mentioned, I am using
> > >
> > > arima y x if cohort==1935, ma(2) condition
> > >
> > > specifying `condition' takes care of the fact that I have a
> panel,> > meaning
> > > repeated observations for each year of data. but when using
> > > predict x, xb
> > > after that regression doesn't work, (I tried it already) and
> adding> > to it
> > > predict x, xb condition doesn't either.
> > >
> > > Any ideas what will work?
> > >
> > >
> > >
> > > Olga Gorbachev Melloni
> > > Ph.D. student in Economics at Columbia University
> > > (212)866-6091
> > > 502 W122 Street, apt. 62
> > > NY, NY 10027
> > >
> > >
> > >
> > >
> > >
> > > >From: Robert A Yaffee <bob.yaffee@nyu.edu>
> > > >Reply-To: statalist@hsphsun2.harvard.edu
> > > >To: statalist@hsphsun2.harvard.edu
> > > >Subject: Re: st: what is SIGMA2 _const in arima
> > > >Date: Thu, 24 Aug 2006 13:52:25 -0400
> > > >
> > > >Olga,
> > > > To obtain the errors, after running the ARIMA, you type:
> > > >predict error, resid
> > > > To obtain the predicted values in an ARIMA, after running
> > > >the ARIMA you type:
> > > > predict pred
> > > >
> > > > Regards,
> > > > Bob Yaffee
> > > >
> > > >
> > > >Robert A. Yaffee, Ph.D.
> > > >Research Professor
> > > >Shirley M. Ehrenkranz
> > > >School of Social Work
> > > >New York University
> > > >
> > > >home address:
> > > >Apt 19-W
> > > >2100 Linwood Ave.
> > > >Fort Lee, NJ
> > > >07024-3171
> > > >Phone: 201-242-3824
> > > >Fax: 201-242-3825
> > > >yaffee@nyu.edu
> > > >
> > > >----- Original Message -----
> > > >From: Olga Gorbachev Melloni <olga_gorbachev@hotmail.com>
> > > >Date: Thursday, August 24, 2006 1:31 pm
> > > >Subject: st: what is SIGMA2 _const in arima
> > > >
> > > > > Hi,
> > > > > I am trying to estimate a model, in which errors follow an
> > > MA(1). I
> > > > > have
> > > > > panel data, and I assume that for each cohort, coefficients
> are> > the> > same. I
> > > > > ran:
> > > > >
> > > > > arima y x if cohort==1935, ma(1) condition
> > > > >
> > > > > I have two questions:
> > > > >
> > > > > 1. when the estimation finished its run, it used BFGS and
> for the
> > > > > estimate
> > > > > of the white-noise disturbance epsilon it showed instead of
> the> > > > usual
> > > > > '/sigma', I got SIGMA2 _const =0.10...
> > > > >
> > > > > what does this mean? I looked in the manual, but there is no
> > > > > mention of
> > > > > this.
> > > > >
> > > > > 2. When trying to predict the errors or the predicted
> values, I
> > > got> > an error
> > > > > message saying that the sample cannot include panels.
> > > > >
> > > > > Is there a way to get the predicted values and/or white-noise
> > > errors?> >
> > > > > thank you,
> > > > > olga
> > > > >
> > > > >
> > > > > *
> > > > > * 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/
> > >
> > >
> > > *
> > > * 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/
>
>
> *
> * 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/