Relative importance of predictors in cox regression: How to get the partial chi-square and degrees of freedom?

[Eran Itskovich]

Hi,

I'm running a Cox regression using -stcox-. I want to compare the relative importance of the independent variables, to see which one explains more of the variance in the dependent variable.
I came across several articles that calculate the relative importance as the partial chi-squared statistic minus the degrees of freedom of each variable.
Can someone please explain how can I get these values (partial chi-square and df) for each variable after running -stcox-?

Thanks!

[Felix Bittmann]

I suggest using domin since it does all this automatically for you.

https://ageconsearch.umn.edu/record/...es/Luchman.pdf

[Eran Itskovich]

Thanks Felix Bittmann!

Unfortunately, the dataset I'm using is on a stand-alone computer with no internet access, so it's a bit of a problem to install user-written programs.
Is there any way to get the partial chi-square and df using some post-estimation command?

Best,
Eran