Getting Error in Regression Modelling

[jagathisp]

Hi All,

I am receiving the below error in PROC MODEL

ERROR: Unable to determine a feasible staring point for the estimation. Check your restrictions and starting values for

       compatibility.

Could you please help me to resolve this issue?

Thanks

JP

[SteveDenham]

Most likely the constraints you have placed on the model yield no valid solutions.  Try relaxing any constraints you might have.

Steve Denham

[sqdqd]

Hi,

I know that you asked the question in 2014, but I had similar problems and found a solution that worked for me.

restrict alpha + beta + gamma = 1;

=> gave me an error

restrict sum(alpha,beta,gamma) = 1;

=> no error occurred