'Artur

perhaps I was not clear. Sometimes observations on say Y may be updated or revised. That means after I change those in Gretl any transformation such as ln Y  or difference Y  that I may have created before the update will still use the original Y data as they do not update automatically. When I try to delete these variables, because they have been used in the session, Gretl will not allow me delete them in order to re-create them using the updated Y.

I have figured a workaround which is to define a variable  YNEW=Y  and then create the transforms needed using YNew. But that is not very elegant!

Brian

Yes you cannot delete them, but re-assigning new values to the existing series should work without problems. If you're working with different samples and are afraid of mixing up stuff by accident you could perhaps set all values to missing first. Or you could close and re-open the dataset which should get rid of the "in use" problem.

-s