A problem with iteration log likelihood not concave in xtpmg command

[Brahim KHOUILED]

Hi everyone,
When I run the PMG model by xtpmg command, I have a problem the iteration log likelihood not concave like this

Iteration 0: log likelihood = -3467.8035 (not concave)
Iteration 1: log likelihood = -3467.4309 (not concave)
Iteration 2: log likelihood = -3467.4142 (not concave)
Iteration 3: log likelihood = -3467.4078 (not concave)
Iteration 4: log likelihood = -3467.4078 (not concave)
Iteration 5: log likelihood = -3467.4078 (not concave)
Iteration 6: log likelihood = -3467.4078 (not concave)
Iteration 7: log likelihood = -3467.4078 (not concave)
Iteration 8: log likelihood = -3467.4078 (not concave)
Iteration 9: log likelihood = -3467.4078 (not concave)
Iteration 10: log likelihood = -3467.4078 (not concave)
Iteration 11: log likelihood = -3467.4078 (not concave)
Iteration 12: log likelihood = -3467.4078 (not concave)
Iteration 13: log likelihood = -3467.4078 (not concave)
Iteration 14: log likelihood = -3467.4078 (not concave)
Iteration 15: log likelihood = -3467.4078 (not concave)
Iteration 16: log likelihood = -3467.4078 (not concave)
Iteration 17: log likelihood = -3467.4078 (not concave)
Iteration 18: log likelihood = -3467.4078 (not concave)
Iteration 19: log likelihood = -3467.4078 (not concave)
Iteration 20: log likelihood = -3467.4078 (not concave)
Iteration 21: log likelihood = -3467.4078 (not concave)
Iteration 22: log likelihood = -3467.4078 (not concave)
Iteration 23: log likelihood = -3467.4078 (not concave)
Iteration 24: log likelihood = -3467.4078 (not concave)
Iteration 25: log likelihood = -3467.4078 (not concave)
Iteration 26: log likelihood = -3467.4078 (not concave)
Iteration 27: log likelihood = -3467.4078 (not concave)
Iteration 28: log likelihood = -3467.4078 (not concave)
Iteration 29: log likelihood = -3467.4078 (not concave)
Iteration 30: log likelihood = -3467.4078 (not concave)
Iteration 31: log likelihood = -3467.4078 (not concave)
Iteration 32: log likelihood = -3467.4078 (not concave)
Iteration 33: log likelihood = -3467.4078 (not concave)
Iteration 34: log likelihood = -3467.4078 (not concave)
Iteration 35: log likelihood = -3467.4078 (not concave)
Iteration 36: log likelihood = -3467.4078 (not concave)
Iteration 37: log likelihood = -3467.4078 (not concave)
Iteration 38: log likelihood = -3467.4078 (not concave)
Iteration 39: log likelihood = -3467.4078 (not concave)
Iteration 40: log likelihood = -3467.4078 (not concave)
Iteration 41: log likelihood = -3467.4078 (not concave)
Iteration 42: log likelihood = -3467.4078 (not concave)
Iteration 43: log likelihood = -3467.4078 (not concave)
Iteration 44: log likelihood = -3467.4078 (not concave)
Iteration 45: log likelihood = -3467.4078 (not concave)
Iteration 46: log likelihood = -3467.4078 (not concave)
Iteration 47: log likelihood = -3467.4078 (not concave)
Iteration 48: log likelihood = -3467.4078 (not concave)
Iteration 49: log likelihood = -3467.4078 (not concave)
Iteration 50: log likelihood = -3467.4078 (not concave)
Iteration 51: log likelihood = -3467.4078 (not concave)
Iteration 52: log likelihood = -3467.4078 (not concave)
Iteration 53: log likelihood = -3467.4078 (not concave)
Iteration 54: log likelihood = -3467.4078 (not concave)
Iteration 55: log likelihood = -3467.4078 (not concave)
Iteration 56: log likelihood = -3467.4078 (not concave)
Iteration 57: log likelihood = -3467.4078 (not concave)
Iteration 58: log likelihood = -3467.4078 (not concave)
Iteration 59: log likelihood = -3467.4078 (not concave)
Iteration 60: log likelihood = -3467.4078 (not concave)
Iteration 61: log likelihood = -3467.4078 (not concave)
Iteration 62: log likelihood = -3467.4078 (not concave)
Iteration 63: log likelihood = -3467.4078 (not concave)
Iteration 64: log likelihood = -3467.4078 (not concave)
Iteration 65: log likelihood = -3467.4078 (not concave)
Iteration 66: log likelihood = -3467.4078 (not concave)
Iteration 67: log likelihood = -3467.4078 (not concave)
Iteration 68: log likelihood = -3467.4078 (not concave)
Iteration 69: log likelihood = -3467.4078 (not concave)
Iteration 70: log likelihood = -3467.4078 (not concave)
Iteration 71: log likelihood = -3467.4078 (not concave)
Iteration 72: log likelihood = -3467.4078 (not concave)
Iteration 73: log likelihood = -3467.4078 (not concave)
Iteration 74: log likelihood = -3467.4078 (not concave)
Iteration 75: log likelihood = -3467.4078 (not concave)
Iteration 76: log likelihood = -3467.4078 (not concave)
Iteration 77: log likelihood = -3467.4078 (not concave)
Iteration 78: log likelihood = -3467.4078 (not concave)
Iteration 79: log likelihood = -3467.4078 (not concave)
Iteration 80: log likelihood = -3467.4078 (not concave)
Iteration 81: log likelihood = -3467.4078 (not concave)
Iteration 82: log likelihood = -3467.4078 (not concave)
Iteration 83: log likelihood = -3467.4078 (not concave)
Iteration 84: log likelihood = -3467.4078 (not concave)
Iteration 85: log likelihood = -3467.4078 (not concave)
Iteration 86: log likelihood = -3467.4078 (not concave)
Iteration 87: log likelihood = -3467.4078 (not concave)
Iteration 88: log likelihood = -3467.4078 (not concave)
Iteration 89: log likelihood = -3467.4078 (not concave)
Iteration 90: log likelihood = -3467.4078 (not concave)
Iteration 91: log likelihood = -3467.4078 (not concave)
Iteration 92: log likelihood = -3467.4078 (not concave)
Iteration 93: log likelihood = -3467.4078 (not concave)
Iteration 94: log likelihood = -3467.4078 (not concave)
Iteration 95: log likelihood = -3467.4078 (not concave)
Iteration 96: log likelihood = -3467.4078 (not concave)
Iteration 97: log likelihood = -3467.4078 (not concave)
Iteration 98: log likelihood = -3467.4078 (not concave)
Iteration 99: log likelihood = -3467.4078 (not concave)
Iteration 100: log likelihood = -3467.4078 (not concave)
Iteration 101: log likelihood = -3467.4078 (not concave)
Iteration 102: log likelihood = -3467.4078 (not concave)
Iteration 103: log likelihood = -3467.4078 (not concave)
Iteration 104: log likelihood = -3467.4078 (not concave)
Iteration 105: log likelihood = -3467.4078 (not concave)
Iteration 106: log likelihood = -3467.4078 (not concave)
Iteration 107: log likelihood = -3467.4078 (not concave)
Iteration 108: log likelihood = -3467.4078 (not concave)
Iteration 109: log likelihood = -3467.4078 (not concave)
Iteration 110: log likelihood = -3467.4078 (not concave)
Iteration 111: log likelihood = -3467.4078 (not concave)
Iteration 112: log likelihood = -3467.4078 (not concave)
Iteration 113: log likelihood = -3467.4078 (not concave)
Iteration 114: log likelihood = -3467.4078 (not concave)
Iteration 115: log likelihood = -3467.4078 (not concave)
Iteration 116: log likelihood = -3467.4078 (not concave)
Iteration 117: log likelihood = -3467.4078 (not concave)
Iteration 118: log likelihood = -3467.4078 (not concave)

etc ....

Any suggestion please?

[Phil Bromiley]

Welcome to Stata list. You will increase your chances of useful answer by following the FAQ on asking questions-provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

There is obviously something wrong with your model. However, since you don't tell us what your model or command statement looks like, no one can really help you.

With user written procedures, assistance and someone on the listserv using that procedure. This is not a procedure I have heard discussed much.

You should start by trying to run a very simple model and then build up. Beyond that generic advice, there is nothing I can tell you is I have no clue what you actually ran.

[Brahim KHOUILED]

thank you for the response, Phil
I'm trying to build a model that studies the effect of the exchange rate on the balance of payments by the PMG estimator.
Note that I applied some of the procedures in this link, but it did not work
https://www.statalist.org/forums/for...ial-regression
I hope you help me!

[Brahim KHOUILED]

I fixed the problem using the option "technique".
Best.

[Clara Joan Joachim]

Brahim KHOUILED hello, could you please explain on how did you use "technique" option? thanks!

[Brahim KHOUILED]

Hi Clara Joan Joachim,
Use one of these techniques: nr or bfgs or dfp, for example:

xtpmg d.x1 d.x2, lr(l.x1 x2) ec(ect) replace technique(bfgs) pmg

Best regards.