SAS EM - Error when trying to run a decision tree model

[Oussa]

Hello folks,

I am having my E-Miner exam this Monday and was trying to do the practical exam   SAS® Certification Practice Exam:    Predictive Modeling Using SAS® Enterprise Miner related to the E-Miner course.

In the 4th section, it asks the following question; "Add a Decision Tree node after the Impute node with TARGET as the dependent variable and all other input variables as independent variables (main effects only). Configure the decision tree to use 1 for Number of Surrogate Rules and Largest for Method in Subtree. Do not change any other property of the Decision Tree node."

but when running the node I am getting an error, I wonder if I am missing an option, because when I connect the Decision Tree node  to the Data Partition node I don't get issues but when I connect it to the impute node I have errors, and in the exam solution it says "Change the c of the decision tree (do not change anything else)." what is the c ? Why I am having an error ?

Thank you in advance

[Reeza]

Note: I don't know the answer to your question, but I modified the subject line so it's clear what you're asking to help you get a faster answer. Good Luck.

[WendyCzika]

Did you see if the log (inside of the Decision Tree node results, not the diagram or project log) has any more details about the error (try searching for "ERROR:") ?

[Oussa]

Hi Wendy,

Yes I did, it's

"NOTE: The data set WORK.EM_USER_REPORT has 132 observations and 4 variables.
NOTE: DATA statement used (Total process time):
      real time           0.03 seconds
      cpu time            0.04 seconds
 
 
NOTE: Fileref X has been deassigned.
16510
16511  *------------------------------------------------------------*;
16512  * End TRAIN: Tree;
16513  *------------------------------------------------------------*;
 
*------------------------------------------------------------*
*
* ERROR: Run time error was encountered.  The system error returned was 1016.
* Please report unresolved problems to Technical Support.
*
*------------------------------------------------------------*"

this is what the log gives (in the results after running the node.

[DougWielenga]

In cases like this, it is important to see the settings and logs of the nodes that led up to the problem.  If you can, please respond and attach two model packages -- one from the Tree attached following the Data Partition node, and the other from the Tree attached to the Impute node.  

The irony is that the Tree does not need imputed values so the Tree should not be dependent on whether values have been imputed or not.  In fact, I would argue that the Tree node can only be hampered by imputed values since it can no longer distinguish between actual values and imputed values unless you keep all of the missing value indicator variables.   

You can create a model package following the steps outlined in SAS Note 46764 - Creating a model package (.SPK file) in SAS® Enterprise Miner(tm) available at  

    http://support.sas.com/kb/46/764.html

You might also consider reviewing SAS Note 41211: Techniques for testing, solving, or reporting problems that occur when using SAS® Enterprise Miner(tm) available at

     http://support.sas.com/kb/41/211.html

And if there is not enough detail in the log, consider reviewing SAS Note 58396: Specify the MPRINT option in SAS® Enterprise Miner(tm) Project Start Code to display additional diagnostic information in node logs available at

     http://support.sas.com/kb/58396

I look forward to hearing from you.

Cordially,

Doug