proc hppanel data=creditend1;
id fyear sic;
class ratings idvc ids iloss;
model ratings = logat ROA leverage idvc ids iloss / fixone;
run;
quit;
What I want is a regression.
I have a pooled cross-sectional dataset whose regresee is also categorical(1 to22, credit ratings) and 3 of regressor are not regular ones: logAT, ROA, and leverage while other 3 are dummies: idvc, ids, and iloss.
With this dataset, I want to regress ratings on these 6 variables while simultaneously holding fixed effects from year and sic(industry).
1) As far as I know, this is pooled cross-sectional, I need to use proc hppanel.
2) then need to have the fixed effect, so used id fyear, sic ~~ / fixone;
3) however, with PROC HPPNAEL, I cannot do have categorical variables... as I did in the CLASS statement.
Does anybody know what would be a proper regression and code for what I am trying now?
It seems that If I need to do this, I may be better to learn SAS Macro. but I cannot now( Yeah. I can start learning but cannot solve this by Macro right now).
Please share your thoughts and experience with this matter so I must super appreciate you!
Jerry
