v6 and v7.1

[cel]

Hi

i'm using two version (7.1 Mar 21 2011 build) and (6, Sep 3 Build).

I now find that when i estimate a pre-saved eqn system under 6 i get fine results, but when i open up the identical wf1 file and do the identical estimation (in fact i save the estimation window so it's just re-running the same object), i get the error message: "No Valid observations in ... ", which is clearly crazy since it's easy to see that all the data is still there.

Previously i actually got both versions to work, but now only 6 will work. can i send anyone my wf1 to check this really annoying problem.

Thanks.

[startz]

Is your equation nonlinear? If so, the initial values in the C vector matter.

[cel]

yes, correct: the system is non linear.

but notwithstanding in the wf1 file there is a
param c(1) ...
line that is identical in both cases (i.e, the file ran under v7.1 and v6).

[startz]

The no valid observations sometimes happens because particular parameter values cause a divide by zero or something similar. But it sounds like you should have that covered.

What you're seeing does sound a little odd. It's a bit of a long shot, but you might want to update your EViews 7 to the current version.

[EViews Gareth]

If updating both copies to the latest version doesn't fix it, please post the workfile.