Bert has already suggested that your code is too spaghetti to let us follow it.
I tried and got a msg that PKF isn't available for R 3.3.3.
Do you actually have a test that the function can be evaluated? That's often
the main issue.
And it does help to try a few parameter sets to see if you hav
Whether you have converged to a global optimum or not depends on the
nature of the likelihood surface. Have you tried different starting
values? As for the warning, I leave that to Prof. Nash, as he *is*
(way) more knowledgeable. However, I suspect the answer is yes, it is
concerning. Are you at a
I suspect, as you hinted, there's little to no hope that anyone will
be willing or able to navigate your code. **Usually** (whatever that
means!) these sorts of problems can be traced back to
overparameterization -- too few data, which could also mean a lot of
"correlated" data, chasing too many
Hi, I´m having troubel using nlminb this is the warning that shows up.
Warning: Cholesky factorization 'dpotrf' exited with status: 1
Variance of the prediction error can not be computed.
Warning: Cholesky factorization 'dpotrf' exited with status: 1
Determinant of the variance of the prediction e
4 matches
Mail list logo
