This post has NOT been accepted by the mailing list yet.
I was hoping you could help me. I am trying to run a cross random effects model in Stata 13.
But so far, I have not been able to pass from the error message "initial values not feasible" .
I have tried to change the start values to the 4 possible options (zero, iv, constantonly and fixedonly), I have also tried to change the startgrid option (although I am sure what that is) and finally I tried running a simpler model to get some initial values and then use them as initial estimates for a more complex model. But I still cannot fix the problem.
This is the code I have used to get initial estimates from a simpler model:
*Obtain initial values from a simpler model
ologit happy female c.age##c.age
meologit happy female c.age##c.age|| coh5: || year: , from (x)
Can someone please help me? Are there any other options to solve this? Also, I read that using the command evaltype(gf0) might help, but what does that mean?