BookmarkSubscribeRSS Feed
🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
Hi All,

This is my 1st time using E-miner. I have a data set and have specified variables and targets.

When I run the decision tree I get an error, its saying "Maximum Levels of 512 Exceeded"

How can I rectify this and what does it mean.

Any help is much appreciated.

cheers
1 ACCEPTED SOLUTION

Accepted Solutions
WayneThompson
SAS Employee
Thanks for using SAS EM. Please see this usage note about max levels exceed 512. http://support.sas.com/kb/20/054.html


Also as suggested I would recommend being careful about using a variable like zip code or other nominal variables that have a really large number of discrete nominal values as predictors (inputs). You can reject this variable or using a binning technique.

View solution in original post

6 REPLIES 6
benbald
Calcite | Level 5
Hi WW, you have a Nominal variable that either needs to be set to Interval or Rejected altogether. For example, if you set Zip Code to Nominal there are too many levels in that field for it to be set to Nominal. You can also transform it using the SAS Code Node so that it has less levels.
deleted_user
Not applicable
Hi benbald,

I have a few nominal variables as inputs and the main one out of these are set as the "ID" as it is an unique identifier like a referrance number.
Maybe they are the ones that are causing the problem, but if I reject them the model im building wouldnt work.

How do you use the SAS code node to lessen the levels?

If I group these variables in to sub groups would that work?
WayneThompson
SAS Employee
Thanks for using SAS EM. Please see this usage note about max levels exceed 512. http://support.sas.com/kb/20/054.html


Also as suggested I would recommend being careful about using a variable like zip code or other nominal variables that have a really large number of discrete nominal values as predictors (inputs). You can reject this variable or using a binning technique.
benbald
Calcite | Level 5
Hey Wickey, ID variables should not be the cause of this error, only input nominal variables. You can bin the variable using the transform node or write code in a sas code node. For vars like Zip Code you can also truncate the value and use the first two or three digits, as long as the geographic area is regional because national would still cause too many levels.
Benbald
deleted_user
Not applicable
Thanks Wayne & Benbald,
I think I know whats causing the error, its the post code that has been set by a variable.

I have 1st try to group these but if they dont work, I will reject them.

Part of the Model is to identify impact based on the area of origination, I wont be able to do this the variable is completely rejected.

Thanks again for your help, its much appreciated.

kind regards
SlutskyFan
Obsidian | Level 7
"impact based on the area of origination"

This is unrelated to your original problem, (although I did learn the hard way and finally figured out the numeric variable errors issue recently myself), & I'm not sure what you are working on, but, would a spatial regression model or something that captures spatial correlation be useful in your work? Not sure how it can be implemented in Miner though, but curious.

FYI - thanks

sas-innovate-2024.png

Join us for SAS Innovate April 16-19 at the Aria in Las Vegas. Bring the team and save big with our group pricing for a limited time only.

Pre-conference courses and tutorials are filling up fast and are always a sellout. Register today to reserve your seat.

 

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
  • 6 replies
  • 6451 views
  • 0 likes
  • 4 in conversation