Subscribe Bookmark RSS Feed

SAS JMP: Mixed Step wise Selection

ShriHanuman

Occasional Contributor

Joined:

Feb 25, 2017

Hi,

Can I run Mixed Step Wise Selection on SAS JMP and if yes can this be done if my dependent variable is categorical?.

1 ACCEPTED SOLUTION

Accepted Solutions
markbailey

Staff

Joined:

Jun 23, 2011

Solution

Select Analyze > Fit Model.

Select the response column and click Y, Response.

Select the predictor colulmns and click Add or use one of the Macros if you want to include higher order effects.

Click the button next to Personality and select Stepwise.

Click Run.

Click Minimum BIC and select P-Value Threshold.

Click Forward and select Mixed.

Click Go.

Learn it once, use it forever!
6 REPLIES
markbailey

Staff

Joined:

Jun 23, 2011

Yes, you can use mixed selection with a cateogorical response, but you can only use the p-value threshold stopping rule.

Learn it once, use it forever!
ShriHanuman

Occasional Contributor

Joined:

Feb 25, 2017

Could you please elaborate the process to do so.
markbailey

Staff

Joined:

Jun 23, 2011

Solution

Select Analyze > Fit Model.

Select the response column and click Y, Response.

Select the predictor colulmns and click Add or use one of the Macros if you want to include higher order effects.

Click the button next to Personality and select Stepwise.

Click Run.

Click Minimum BIC and select P-Value Threshold.

Click Forward and select Mixed.

Click Go.

Learn it once, use it forever!
markbailey

Staff

Joined:

Jun 23, 2011

I suggest that you try select Help > Books > Fitting Linear Models. This book has an entire chapter devoted to the use of the Stepwise platform.

Learn it once, use it forever!
ShriHanuman

Occasional Contributor

Joined:

Feb 25, 2017

Hi, thnx for the reply... When I do the steps mentioned above. I get
Alert: Failed to converge(step-halving-limit) error. What am I missing?
markbailey

Staff

Joined:

Jun 23, 2011

Hard to say, given so little information.

  • What is the nature of your response? Ordinal? Nominal? Number of levels?
  • Does your response exhibit a separation problems with any of the predictors?
  • How many predictors?
  • Did you try to fit the same model with Nominal Logistic or Ordinal Logistic already? Do you get the same error?

You might need to send your case to support@jmp.com for help.

Learn it once, use it forever!