I am trying to use the Stata kitchensink package for a model with more than 10 independent variables. Unfortunately, the package does not allow models to include more than 10 independent variable.
So, I wonder if you could modify the package code to remove this limitation?
I am attaching a copy of the package.
use R instead (r-project.org). Its free, has only the limitations related to the software and hardware you are using and a large user support base...
Tarik - in line 8 of the ado file, you will see:
syntax varlist(min=3 max=11 numeric) [if] [in] [, LEVel(real 90) LOGit aic]
The max is set to 11, which is one dependent variable and 10 independent variables. Change this max value to a higher value to use more than 10. However, you should note that the number of regression models run is exponential in the number of independent variables. So increasing beyond 10 can be computationally expensive
Thanks Luke.
i did change that line, but kept getting an error message saying too many variables.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics