BookmarkSubscribeRSS Feed
🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
pvareschi
Quartz | Level 8

Would it be possible to clarify how Enterprise Miner treats/considers binary variables?

The official documentation "Enterprise Miner 15.1: Reference Help", at page 154, states: "A binary variable can be legitimately treated as nominal, ordinal, interval, or sometimes ratio".

Does it mean it depends on the context/node being used?

Additionally, does it matter if a binary variable is created with a numeric rather than character type?

1 ACCEPTED SOLUTION

Accepted Solutions
CatTruxillo
SAS Employee
Hi! The way that Enterprise Miner uses a binary variable depends on how it is defined in the metadata. When you create a data source, each variable has a Name, Label, Role, Level, etc. If a binary variable has Level=Binary, then it is treated as nominal. If the variable has Role = Target, then there is typically a reference level (or Event level) defined for the modeling node. All of these can be changed by the user.

I'm sure others will chime in with more details-in general, what EM does with your metadata is aligned with solid predictive modeling practice, although it is important to know what it is doing so that you can interpret your results correctly.

I hope this helps!
Cat

View solution in original post

1 REPLY 1
CatTruxillo
SAS Employee
Hi! The way that Enterprise Miner uses a binary variable depends on how it is defined in the metadata. When you create a data source, each variable has a Name, Label, Role, Level, etc. If a binary variable has Level=Binary, then it is treated as nominal. If the variable has Role = Target, then there is typically a reference level (or Event level) defined for the modeling node. All of these can be changed by the user.

I'm sure others will chime in with more details-in general, what EM does with your metadata is aligned with solid predictive modeling practice, although it is important to know what it is doing so that you can interpret your results correctly.

I hope this helps!
Cat