Unequal Variable Groups

[Dreeken]

Well, I seem to have hit another brick wall. Starting to feel stupid now...
Anyway, if I implement your previous suggestion into my 'test program' it works like a charm. The only issue is that I also want to incorporate dummy variables.
Now for slope-dummies there is no problem at all; I just create those by multiplying the dummy with a variable of choice and make sure that it gets placed into the right group. However, I get stuck when figuring out how to implement the intercept-dummies.

These dummies are based on the variables which aren't present for all the series, so it's quite related to the original problem. Thus, when I use the following the specification (for example)

Code: Select all

alfa_ts1 beta_ts1 gamma_ts1 beta_ts1*dummy1_ts1 c(1) c(2)*dummy_1_ts1

which could be created through using a program line like

Code: Select all

equation eq_alfa_ts{!i}.ls(n) ts{!i} c(1) c(2)*dummy1_ts{!i}

I get the error which indicates that there is a missing series, only this time for the dummy variables. How can I solve this?

Thanks a million for the help you already gave me! Hope I can get of your back quickly, sorry for the harrasment.

PS. I added my test program with the above described dummy specification so you can also take a look at that..

[EViews Gareth]

That example doesn't make sense - why do you have coefficients as variables?

[Dreeken]

Do you mean the c(1) and c(2)? Well, that was the way in which I was taught in my lecture to incorporate intercept dummies into an equation.
Or are you talking about something else?

[EViews Gareth]

Yes, I was talking about c(1) and c(2). They should not be in an equation specified by list.

[Dreeken]

Then in what way do you suggest that I use intercept dummy specifications?

[EViews Gareth]

Just include the dummy variable as you would any other variable.