BookmarkSubscribeRSS Feed
Mike90
Quartz | Level 8

I am getting this message in the score report of a decision tree model.  Three variables are not being used for this reason.  These variable have between 150 and 250 levels.  I can drop a lot of rare levels, but I need to know what the cutoff is, and if it can be changed.  I read there's a 512 level cutoff for logistic regression.

 

Googling for this exact error message returns 2 results, neither of which is about resolving this problem.

 

Thanks.

 

SAS Enterprise Miner 14.3.

 

 

3 REPLIES 3
Reeza
Super User

@Mike90 wrote:

 

 

Googling for this exact error message returns 2 results, neither of which is about resolving this problem.

 

 


Post what you've tried so we don't suggest those. 

So, we can assume you've set/changed the maximum levels in the project settings?

Mike90
Quartz | Level 8

I went through the settings, and I don't see where to set this.

 

I see Options -> Preferences, and there are options to the left of any selected node. .

 

2 irrelevant Google matches for the exact error message isn't a lot of help.

 

I did try increasing exhaustive candidate splits on the decision tree to 15,000.

 

 

 

 

Reeza
Super User

Project Macro Variables?

EM_TRAIN_LEVELS?

 

http://support.sas.com/kb/20/054.html

 

I'm not sure what you mean by 2 irrelevant results, searching "sas maximum levels 512" returned a lot of results, not all relevant as usual. The first one is the note above which includes:

 

To access these properties, click the project name in the Project Panel (upper left area). The Project Start Code property and the Project Macro Variables property are displayed in the Properties panel (middle left area). For information, see the chapter "SAS Enterprise Miner User Interface Help". To access the chapter in SAS Enterprise Miner, select Help ► Contents ► User Interface ► SAS Enterprise Miner User Interface Help.

  • Project Macro Variables. Enter the EM_TRAIN_MAXLEVELS value that you want to use.

hackathon24-white-horiz.png

2025 SAS Hackathon: There is still time!

Good news: We've extended SAS Hackathon registration until Sept. 12, so you still have time to be part of our biggest event yet – our five-year anniversary!

Register Now

How to choose a machine learning algorithm

Use this tutorial as a handy guide to weigh the pros and cons of these commonly used machine learning algorithms.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 3 replies
  • 2655 views
  • 0 likes
  • 2 in conversation