[R] ungrouped data for multinomial regression

Rodrigo Briceño rjavierbriceno at gmail.com
Sat Jun 28 14:48:13 CEST 2008

Thanks for helping me out with the thing of quotation marks. I finally
could save time running programs at the university and home.
One additional question for all of you. I'm starting to be familiar
with multinomial regression models. Actually I'm doing some exercise
with the pneumo data that appears in Faraway library. The database has
24 observations and I want to study the probability of getting sick
according with the number of years working at a coal mine.
Here is the issue. For modelling that response what I need to include
in the model is something similar at what is produced by:
xtabs(Freq~year+status). This table provided me the proportions of
people sick according with years worked and severity of the disease.

The question is: do this imply that I need to transform my data
previously to run the regression? how?

Thanks in advance.

More information about the R-help mailing list