Decision Tree Node Rules truncated and not matching the Scoring Code

[MarianaL]

I have long text variables and the Node Rules (English Rules) are truncated. How do I remedy this and display the whole If any of ... then condition? Also, the Scoring code does not match the Node Rules, it has more variables than used in the Decision Tree Node Rules and picture. I can't use the scoring code for this reason, and the node rules are not complete. Why is the scoring code different than the Node Rules?

[PatrickHall]

Did you use interactive mode? If so, check the log from the interactive session for errors.

My understanding is that this situation can occur if an error happens after interactive training when you close the interactive session. Normally the PROC behind the node would re-run and update the rules and score code at this time to reflect your changes, but cannot do so if there is an error.

[MarianaL]

No, I have used the non-interactive DT, the auto tree

[WendyCzika]

Another case where the English Rules could appear different from the score code is when using surrogate rules.  The surrogate rules are not included in the English Rules.

[PatrickHall · Accepted solution]

Just putting this here in case it comes up again for someone else ...

As discussed over email - surrogate rules - i.e. looking for another variable to split on in the case of an observation containing a missing value for the primary splitting variable - are not displayed in the english rules. (Thanks to Susan Haller for pointing this out.)

Note that the Decision Tree node automatically handles missing values in split searches. The surrogate rules are just meant to give users more flexibility for how to handle missing values.