Not Converging
Posted: Wed Dec 07, 2011 1:11 pm
Hello,
I've specified a model with eight independent variables (excluding intercept) and a dependent variable.
When I'm specifying the random intercept model it does converge. However when I'm trying to estimate a random slope with the model won't converge.
N=1666 and there are two levels (brand and individual). For almost all the variables I got the following warning, only for two variables I'm able to make them 'random'.
I get the following warning:
"SSP Matrix for fixed part has gone negative indefinite"
Anyone knows what causing this or how to solve it?
Thank you
I've specified a model with eight independent variables (excluding intercept) and a dependent variable.
When I'm specifying the random intercept model it does converge. However when I'm trying to estimate a random slope with the model won't converge.
N=1666 and there are two levels (brand and individual). For almost all the variables I got the following warning, only for two variables I'm able to make them 'random'.
I get the following warning:
"SSP Matrix for fixed part has gone negative indefinite"
Anyone knows what causing this or how to solve it?
Thank you