Spencer Jones
2006-Aug-10 19:35 UTC
[R] Negatie Binomial Regression: "Warning while fitting theta: alternation limit reached"
I am fitting a negative binomial regression model to some count data. I chose the negative binomial b/c the variance is greater than the mean. Anyways, when I fit the model I get the following warning: "Warning while fitting theta: alternation limit reached" The estimate that I end up with is very large (1070), and the standard error is even larger (1276). Does this indicate that I should not be using glm.nb? Is there anyway to remedy the problem,i.e. increase the number of iterations allowed? thanks, Spencer [[alternative HTML version deleted]]