EViews.com

gd gd stuff.

Hi, I have a very naive question as a graduate unfamiliar with this add in. I am trying to take a 4 variable VAR and form a forecast model to test the effect of changes in w, x, and/or z on y, and the effect of changes in w, y, and/or z on x. Any help would be greatly appreciated, I am just very unsure how to approach this Thanks

Hi,

is there anybody who can tell me why the forcasts produced by the add-in are in the long run slightly different from the dynamic solution created by a solve command?

Here are the relevant commands:

_var.makemodel(_model_03)
_model_03.scenario(n, a=bas) "bas"
_model_03.solve

_var.fcast(g,e) _f _se

Thanks!

In order to produce standard errors, the add-in does a stochastic solve.

f=na isn't an available option here? should/could it be?

best wishes!

Charlie

The model object doesn't have such an option.

You could always manually set the NA values.

Good morning everyone,
I have downloaded successfully the VARforecast addin. But when I write the command fcast, eviews tells me that fcast is not a valid view for VAR. Could anyone help me to solve this? I do appreciate your help. Many thanksss

Did you install it properly?

Did you install it properly?

Yes, I can forecast with the menu but not with the command. Could you please tell me how to fix that?

Did you install it properly?

Yes, I can forecast with the menu but not with the command. Could you please tell me how to fix that?

Dear Mr Eviews Gareth
I have reinstalled the proc, and now it works for me well with the command, Thank you very much

hi
it's possible to modify the Add-in to detect automaticaly the Time Series Functions, when I use the forecast like equation object (forecast differenced or undifferenced data).

best regards.

Hi all,

This si a bit of a naive question,
Is there a command line to change the name of the forecast and the forecast sample? I am trying to run a forecast evaluation and need to store several forecasts over different samples and it would be quite painful to do it by hand.

By default the var.fcast command seems to generate "_f" suffix forecast series int he same sample as the mdel is estimated.

Typing
smpl [smothign other than the estimation sample]
var.fcast

will not change the forecast sample, and

smpl [smothign other than the estimation sample]
var.fcast _f1

Will cause a massive error message

Many thanks for help
M

Have you checked the documentation?

I have installed this add-in, but I couldn't see it in Add-ins option. I try some times to reinstall it but it 's not succesful. Anybody can help me ? :(

Do you have a VAR open when you look at the Add-ins menu?