Panel data error

[Hossein ka]

Hi ,
I am estimating a panel data and when I do not include d.FXV( that is Exchange rate in my model), I get the following message. Would you please let me know what this message means? Thank you




And here's the message for r(498):

[William Lisowski]

On my browser, at least, neither of the two messages you posted display.

We can better help you if we know what commands you have tried and what Stata told you to indicate that there was a problem. Please review the Statalist FAQ linked to from the top of the page, as well as from the Advice on Posting link on the page you used to create your post. See especially sections 9-12 on how to best pose your question. It's particularly helpful to copy your commands as well as your output from your Stata Results window and paste them into your Statalist post using CODE delimiters, as described in section 12 of the FAQ, rather than present screen shots as pictures. For example, the following:

[code]
// sample code
sysuse auto, clear
describe
[/code]

will be presented in the post as the following:

// sample code
sysuse auto, clear
describe

[Hossein ka]

Thanks William for your comment. Attached please find the data and below the command:

 xtset Country Year, yearly 
  gen lnRD = ln(RD)
  gen lnWTIC = ln(WTIC)
  gen lnEDU = ln(EDU)
  xtpmg d.FDI d.lnRD d.MFP d.lnEDU d.lnWTIC d.MKT d.LAG d.OPEN d.FXV d.iRATE d.INFL d.TAX, lr(l.FDI lnRD lnEDU lnWTIC MFP MKT LAG OPEN FXV iRATE INFL TAX) ec(ec) 

I am still getting the following error:
"expression (-_b[lnRD]/_b[L.FDI]) evaluates to missing"

[William Lisowski]

I installed the software xtpmg from the Stata Journal and downloaded your data. Below is my attempt at reproducing your results (Stata 14.1 for Macintosh).

. import excel "/Users/lisowskiw/Research/Stata sandbox/160327/Book1.xlsx", sheet("Sheet1") firstrow

. xtset Country Year, yearly 
       panel variable:  Country (strongly balanced)
        time variable:  Year, 1994 to 2012
                delta:  1 year

. gen lnRD = ln(RD)

. gen lnWTIC = ln(WTIC)

. gen lnEDU = ln(EDU)

. xtpmg d.FDI d.lnRD d.MFP d.lnEDU d.lnWTIC d.MKT d.LAG d.OPEN d.FXV d.iRATE d.INFL d.TAX, ///
>    lr(l.FDI lnRD lnEDU lnWTIC MFP MKT LAG OPEN FXV iRATE INFL TAX) ec(ec) 

numerical derivatives are approximate
flat or discontinuous region encountered
[10 more like those]
Iteration 0:   log likelihood = -13.911434  (not concave)
Iteration 1:   log likelihood = -8.2276538  (not concave)
Iteration 2:   log likelihood = -7.9254988  (not concave)
Iteration 3:   log likelihood =  4.0591691  (not concave)
Iteration 4:   log likelihood =  6.2226751  (not concave)
Iteration 5:   log likelihood =  6.6365418  (not concave)
Iteration 6:   log likelihood =  6.6494126  (not concave)
...
Iteration 59:  log likelihood =  16.123657  (not concave)
Iteration 60:  log likelihood =  16.229481  (not concave)

and at that point I gave up. Is this the sort of behavior you saw when you ran it?

I have to say, it does not look like the sample runs shown in the Stata Journal paper associated with the software. If this is an accurate reproduction of what you saw, then I suspect xtpmg is trying to tell us that the algorithm is not well suited to the data. I'm not familiar with the software, but perhaps someone else will be able to advice. Or you could correspond with the name given for support in the software description at

net sj 7-2 st0125

[Hossein ka]

Thanks for your time. When I run the program in the Stata, there is not any iteration records and just the following error is appeared:
"expression (-_b[lnRD]/_b[L.FDI]) evaluates to missing".

[Matias Cocina]

Hi Hossein.
I'm having the same problem with the xtpmg command. Did you ever get to solve this? If so, any clues of what might be causing the issue?
Of course I'm using different data, but I was wondering if you might have some general insights about which features of your data were triggering the message.
Thanks
Matias

[Helge Arends]

Hi Matias, I have the same problem with the xtpmg command. Were you able to solve it?

Best
Helge

[River Huang]

Thanks for your time. When I run the program in the Stata, there is not any iteration records and just the following error is appeared:
"expression (-_b[lnRD]/_b[L.FDI]) evaluates to missing".

I have never encountered your problem. But according to my limited experience, I'd suggest that you start with a simple model (do not include so many variables in the first place), and you might find some preliminary results.

[Tafadzwa Luke Mupingashato]

Greetings was this problem ever solved kindly assist?