proc genmode regression equation

[NinaWestern]

I am using proc genmod with tweedie distribution and log link to analyze positively skewed outcome.

Trying to figure out the actual regression equation to include it in the paper.

I am struggling with how to reflect the link function in the equation.

Here is the SAS syntax: 

proc genmod data=sasuser...;
class sex CAS;
model int= sex age_cent age_sq sex*age_cent CAS rstr_cent rstr_sq sex*rstr_cent/dist=tweedie link=log;
run;

Here sex is coded 0 and 1, CAS has 4 categories (1,2,3,4), and the last category is used as a reference.

The model includes the quadratic effects of age and Rstr.

Wondering if the following equation is correct:

Or should it be without Ln's in the formula? 

[Reeza · Accepted solution]

This may not answer you question specifically but lets you check your answer.

You can use a CODE statement in PROC GENMOD, in many cases, and it will generate the code to score new data. Looking at that code can help you check if the formula is correct when typing it out. 

https://blogs.sas.com/content/iml/2014/02/19/scoring-a-regression-model-in-sas.html

---

@NinaWestern wrote:

I am using proc genmod with tweedie distribution and log link to analyze positively skewed outcome.

Trying to figure out the actual regression equation to include it in the paper.

I am struggling with how to reflect the link function in the equation.

Here is the SAS syntax: 

proc genmod data=sasuser...;
class sex CAS;
model int= sex age_cent age_sq sex*age_cent CAS rstr_cent rstr_sq sex*rstr_cent/dist=tweedie link=log;
run;

Here sex is coded 0 and 1, CAS has 4 categories (1,2,3,4), and the last category is used as a reference.

The model includes the quadratic effects of age and Rstr.

Wondering if the following equation is correct:

Or should it be without Ln's in the formula? 

---