BookmarkSubscribeRSS Feed
Michael_Sz
Calcite | Level 5

My Situation:

I am using SAS Enterprise Miner within SAS OnDemand for Academics.

On an empty diagram I put an Input Data-node containing a matrix (filled with numbers with many missing values).

Connected to this node there is a MBR-node (with Method=Scan, Number of Neighbors=2)

The Problem:

When using the MBR-node I get the following errors in the log of the node:

 

ERROR: Domain error.

 

A few lines below:

ERROR: Termination due to Floating Point Exception

 

These errors seem to appear depending on the target variable that I chose (i.e., when changing the target variable I sometimes get no errors).

What could be causing these errors?

The only information I could find were this or this but neither do I use proc sql nor does my matrix consist of dates.

1 REPLY 1
gergely_batho
SAS Employee

Perhaps you don’t have valid (all variables non-missing) observations in your dataset?

Also check the type of the target variable.

Ready to join fellow brilliant minds for the SAS Hackathon?

Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.

Register today!
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
  • 1 reply
  • 984 views
  • 0 likes
  • 2 in conversation