matrix not positive definite
Posted on

Forums
Another very basic question, but it has been bugging me and i hope someone will answer so I can stop pondering this one.
So you run a model and get the message that your covariance matrix is not positive definite.
Why isn't the optimiser algorithm set up to avoid areas of the solution space where the eigenvalues are negative?
So you run a model and get the message that your covariance matrix is not positive definite.
Why isn't the optimiser algorithm set up to avoid areas of the solution space where the eigenvalues are negative?
The optimizer will do its
Because we allow people to specify arbitrary algebras, we can't really make guesses about the parameter space, and even in standard analysis cases, strange things can happen when the parameters are far from the minimum.
Often the error happens because of a user specification--for example, the error commonly comes up when users start all their free parameters at the same number. We can warn the user about this, but OpenMx tries not to make changes to things like starting values without the user's consent.
I think early detection is unlikely to be a viable solution, but it might be possible to add some recovery routines to handle this case when it happens however. For example, it might be possible for the user to request that OpenMx try different starting values if the first ones don't work, for example, or to jitter the parameters when a non-positive-definite point is found to see if nearby regions are viable.
Log in or register to post comments
In reply to The optimizer will do its by tbrick
At the end of the day, often
Nevertheless, as Tim says, good starting values (try getting the means & variances right, and the covariances in the right ball park, or near zero) can often keep the optimizer from straying into non-positive definite regions.
Log in or register to post comments
In reply to At the end of the day, often by neale
I guess what I still don't
Still a bit puzzled as to how optimisation works, I guess.
Log in or register to post comments
In reply to I guess what I still don't by Dorothy Bishop
Unfortunately we have only
Log in or register to post comments