[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: Re: messy ml problem
Are you employing the trick that Stata uses to bound the correlation
coefficient in the biprobit.ado code: estimate a bounded function of r,
and then back-transform when the r is needed?
Kit Baum, Boston College Economics
On Feb 5, 2005, at 2:33 AM, Jason wrote:
I am using the function `binorm' to calcluate the density of the two
binary variables. `binorm' takes three inputs, (conditional) means for
two variables and a correlation coefficient (variances are normalized
one). Now there three arguments are, in my model, complicated
of elements from the 4 x 4 covariance matrix of the full system.
The problem is Stata can't find any starting values for this procedure.
I've looked at the correlation coefficient that is being generated and
often it exceeds 1 or is smaller than -1, in which case binorm
output. I have tried imposing restrictions on various parameters when
running `ml search' to see if I can get a correlation coefficient in
unit interval, but that didn't work either.
Any suggestions for how to proceed?
* For searches and help try: