positive or non negative argument to function expected

[ensbsh]

hi,
I''m trying to do the autocorrelation test of serial correlation LM test
the dependent variable is mni and i have included in the independent variables mni(-1). When i try to perform the test with 2 lag in the lag specification the answer is positive or non negative argument to function expected when i try with 1 lag in the lag specification it says insufficient degrees of freedom. Can you please help me understand the problem, also the same thing happens when i try to do the Ramsey reset test.

[startz]

You are more likely to find someone can help you if you post
(1) the exact steps that are giving you problem
(2) the EViews workfile in which the problem occurs.

[ensbsh]

Ok thanks.
My problem is that after building the equation based on the data from the above excel file, i tried to test if there was serial correlation. Because my series contains the a variable with 1 lag i used the serial correlation Lm test in the residual diagnostics. When i type the lag specification: 2 , the error message that appears says: positive or non negative argument to function expected!The same thing happens when i try to do the Ramsey reset test to determine the form of the model. What does this mean and how can i fix it! If you can help i would be very grateful. The data is frim the excel file above named without macro.

[EViews Gareth]

You are more likely to find someone can help you if you post
(1) the exact steps that are giving you problem
(2) the EViews workfile in which the problem occurs.

[ensbsh]

Ok so i did that, but i didn't see a tweet related to my problem
Can't you help me please? The data is in the above excel file! Thanks

[startz]

You are more likely to find someone can help you if you post
(1) the exact steps that are giving you problem
(2) the EViews workfile in which the problem occurs.

[Goshka]

otchet_ufa2.wf1

used eviews 7

(213.13 KiB) Downloaded 294 times

I had the same problem. Here is the deal.
My sample was 179 initually. Then I cut it (with help of "Sample" button) to 173. After that i decided to estimate my residuals. I made a list of residuals (out of equation; Proc->make residual series) and a dummy (dt)(1 for 1/2 of my obs amount, and 0 for another 1/2 ) var for "Equality Tests by Classification".
The dummy i made was like this:
smpl if n01>86
series dt=1
smpl if n01<=86
genr dt=0
Then i did:
Resid01/View/Desriptive statistics&Tests/Equality Tests by Classification/dt/Variance
And then i got this error:
"positive or non negative argument to function expected"
The main problem is that while i was making dummy, my sample of 173 obs went back to 179 automatically; just after i press ENTER after the "smpl if n01>86" command.

[startz]

When you looked at your residuals just before doing the test, did you see a bunch of different values or were they all the same?

[Goshka]

When you looked at your residuals just before doing the test, did you see a bunch of different values or were they all the same?

i really did. but it is the question of my task. I just have to do this procedure

[startz]

You might want to post your workfile, and post the exact commands you have given.

[Goshka]

When you looked at your residuals just before doing the test, did you see a bunch of different values or were they all the same?

i missuderstood you first time, sorry. There was a bunch of different values

[startz]

It sounds like you are doing things correctly. You might want to update your copy of EViews if that hasn't been done recently. Past that, post your workfile and the exact commands you have used.