Postby EViews Glenn » Thu Oct 29, 2015 12:30 pm
One of those bugs that is a combination of things. It's not really a bug in the equation estimation part, but in our handling of an error that occurs because the nonlinear estimation failed.
First your estimation is failing because the starting values for your particular equation are poor (getting starting values when you have exponentials is always a bit tricky as you have to guard against overflow) and the optimizer can't figure out how to start. Normally that doesn't cause problems, but in your case it requires us writing an overflow value into the output table, and we're not handing that operation properly. I've cleaned up the write and a fix will appear in the next patch.
For now, set your starting coefficient values to something safer like 0.1 and you'll be good to go.
Sorry about the inconvenience.