topic Cox Regression variable selection in Discussions

Cox Regression variable selection

jmpquestions123 — Fri, 09 Jun 2023 00:38:09 GMT

Hello everyone, is there a way to select variables for Cox regression using a stepwise model in JMP? The stepwise regression model from the Fit Model only has a "Y" category, whereas for Cox-regression I would imagine that in order to choose the best variables I would need an option to add both a "Y" category (event), as well as an "X" category (Time to event)? If not would I put the event in for Y?

TLDR: Is there a model selection option for Cox-regression?

Re: Cox Regression variable selection

Mark_Bailey — Mon, 30 Aug 2021 13:48:26 GMT

I assume that you mean the Cox regression that assume proportional hazards. Click the button in the top right corner of the launch dialog and select Proportional Hazard for the fit personality.

The Y role is for the time to event. The Effects are the predictors / factors / features you want in the linear model.

A step-wise procedure is not available in this platform, though. Do you have a lot of effects to test?