Hy everyone :o
I have a question about SAS insight. I'm trying to fit the gamma regression model, and I need Anscombe residuals. Since I can get them only in SAS insight, I've tried to fit the model there, but... I have two variables, response is continuous variable "settled amount" and explanatory variable is categorical, "injury code", with seven levels, 1-6 and 9.
When I'm fitting the model with PROC GENMOD I get (I've used log link, and reference level is 1)
log y = 10.1575 + 0.9404 x_2 + 1.3265 x_3 + 1.6345 x_4 + 2.1624 x_5 + 1.3331 x_6 - 0.6371 x_9
but when I'm using insight, I only get
log y = 10.2808 + 0.1395 x
because it treats injury code like a continuous variable.
So, my question is, is there any way for letting it know that injury code is categorical variable? :)
I have a question about SAS insight. I'm trying to fit the gamma regression model, and I need Anscombe residuals. Since I can get them only in SAS insight, I've tried to fit the model there, but... I have two variables, response is continuous variable "settled amount" and explanatory variable is categorical, "injury code", with seven levels, 1-6 and 9.
When I'm fitting the model with PROC GENMOD I get (I've used log link, and reference level is 1)
log y = 10.1575 + 0.9404 x_2 + 1.3265 x_3 + 1.6345 x_4 + 2.1624 x_5 + 1.3331 x_6 - 0.6371 x_9
but when I'm using insight, I only get
log y = 10.2808 + 0.1395 x
because it treats injury code like a continuous variable.
So, my question is, is there any way for letting it know that injury code is categorical variable? :)
