Hi,
I try to run this path in Enterprise Miner 13.2 to get rid of correlated variables:
Dataset --> Metadata --> Transform variables --> Sample (becouse sample size>100000) --> Variable clustering
but I get a run-time error on variable clustering node prompting me to see at the log file where I find no description of the error..
any idea?
Thank you
That's right, the diagram and project logs wouldn't show the error message. So after running and getting the error window at a particular node, right-click on the node and select Results... and then click on View --> SAS Results --> Log For some nodes, the log might be truncated in the node itself but it will tell you at the bottom of the log where the full log is in those situations.
Just want to make sure you are looking at the correct log - are you opening the Results of the Variable Clustering node and viewing the log from there (View>SAS Results>Log) ? From there, you should be able to search for "ERROR:" to find more information.
Hi, thank for the reply
The Run Status window pops up and notifies me over a run time error before any results are produced. Log tab next to diagram tab is empty, Project Log (View --> Project Log) does not produce anything
That's right, the diagram and project logs wouldn't show the error message. So after running and getting the error window at a particular node, right-click on the node and select Results... and then click on View --> SAS Results --> Log For some nodes, the log might be truncated in the node itself but it will tell you at the bottom of the log where the full log is in those situations.
it is!
I though you meant the results windows that was supposed to pop up if everything had run successfully
So thanks a lot!
The problem is at the TRANSFORM VARIABLES node that preceded the VARIABLES CLUSTERING NODE. The settings for this node was: DEFAULT METHODS,INTERVAL INPUTS --> RANGE. Whenever I change the first setting to NONE the path runs successfully no matter what the other setting were. Some other methods produce the same problem (like BUCKET) while others don't (like LOG)
Thanks a lot
The Variable Clustering node only works on interval inputs, so when you use a binning transformation like Bucket, the interval inputs are converted to nominal, so it won't run. Though I'm not sure why that would happen with the Range transformation.
SAS Innovate 2025 is scheduled for May 6-9 in Orlando, FL. Sign up to be first to learn about the agenda and registration!
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.