[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
Christopher F Baum <baum@bc.edu> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
st: Re: dynamic forecasts |

Date |
Wed, 4 Aug 2004 09:05:40 -0400 |

On Aug 4, 2004, at 2:33 AM, Martin wrote:

Please read my London SUG talk (available in IDEAS, http://ideas.repec.org under Stata Users Group) and look at arima_X.do. You can do that via ssc type arima_X.do, but it helps to read the surrounding text.

Dear Stata users,

I am trying to calculate fitted values from a very simple dynamic model, in

this case one with a lagged dependent variable. The model is:

x[t] = a*x[t-1] +b*y[t] + u[t], for t >= 1

I first create the variable 'x1' which is the lagged value of x, and then

run:

reg x x1 y

Now I want to create the fitted values from this regression. The command

'predict xhat', would simply calculate:

xhat[t] = a*x1[t] + b*y[t] for t >= 1, which is not dynamic.

The equation I should be using is:

xhat[t] = a*x1[t] + b*y[t]), for t = 2 and

xhat[t] = a*xhat[t-1] + b*y[t], for t > 2

Does anyone know of a simple way of doing this in Stata? I realize I could

write a loop that calculates xhat observation by observation using the

parameters from the estimation, but that is a hassle when I have to run many

different specifications. I'm hoping Stata has a command similar to

'predict' that calculates the fitted values in simple dynamic models like

the one in this example. Any help will be appreciated.

Kit

*

* 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/

- Prev by Date:
**Re: st: Re: political geography** - Next by Date:
**st: SSC activity July 2004** - Previous by thread:
**st: Re: reproducing graphs** - Next by thread:
**st: SSC activity July 2004** - Index(es):

© Copyright 1996–2015 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |