EViews.com

Hello EViews chaps!

I have a question regarding storing of, and use of series in EViews. I have well over 2,000 series in EViews which I may or may not use for estimation of OLS equations.

The variable nomenclature is consistent, and allows me to get around the workfile page relatively adequately, but since I have so many variables it is difficult to simply browse through what I have. Ideally I would like a way that would allow me to file similar series in groups, reducing the complexity of the workfile page, and allowing one to see all the different types of variables which are present without seeing the individual series.

Perhaps it's best described using an example. Suppose that I had the following monthly dummy variables:

m_01, m_02, m_03, m_04, m_05,.....,m_12

In EViews currently the workfile space is cluttered with each of these individual variable names. I would like a way to group these into individual groups, and remove the variables from workfile page layout. I know how to create groups, but these are only copies of the variables themselves, so can't exist without the variables themselves.

Is there a way round this problem? Is it possible to create groups in one workfile page which link to series in another workfile page? This would still have to allow equations in the former page to access the individual series for estimation.

Essentially I would like the following way to store variables:

Workfile page 1:

var_1
var_2
.
.
.
var_2000

Workfile page 2:

Group - dummy variables
Group - economic variables
Group - sales variables
Group - weather data

Equation 1
Equation 2

Is this possible to do at all? Alternatively are there other ways to 'file' variables in EViews which ensure that workfile pages don't become cluttered with variables?

Thanks,

Ben

Unfortunately, in general, this is not possible.

With the case of your specific example, there are things you can do to reduce the clutter. For monthly dummies, assuming you have a dated workfile, there is no reason to have those dummies in the workfile at all. You can simply use one of EViews built in dummy tools whenever you want to use the dummies - for example @expand(@month)

Hi Gareth,

Thanks for your swift answer. On a similar note, a number of my variables are derivatives of other variables within my workfile. Instead of manually creating the variables and having them clutter up my file, is there a way to create them in an automated fashion for estimation purposes?

For example, if I have a variable which is a log of another variable I simply use log(variable_name) in an OLS equation. Is it possible to do this with a user-defined function, f()? So that instead of creating a series and storing it in the workfile, I simply put f(variable_name) in the OLS equation. This again would serve the same purpose as the original question and prevent variables from cluttering up the workspace.

Many thanks again.

Ben

Unfortunately not.

Hi Gareth,

Thanks for your quick answers - no problem, just wanted to check to see if there was a better way of filing variables than I am currently.

Thanks,

Ben