Impute Node Question - Module 3: Predictive Modeling using SAS Enterprise Miner 14

mandronicou — Sat, 25 Apr 2020 12:55:31 GMT

Hi,

[This is a general question]

When asked to create unique indicator variables whenever a missing value is imputed and use them as inputs - how would this affect my choice of the "Indicator Variable Source" under the Impute Note Score Properties? Do I choose "Missing" or "Imputed"?

In the SAS documentation https://documentation.sas.com/?docsetId=emref&docsetTarget=p1lmg1hmgbpz5cn1cxhmqfucnxlj.htm&docsetVersion=14.3&locale=en it just mentions that you specify the role, but there's another field in which we specify the role (i.e. "Indicator Variable Role").

Can you help with this?

Thanks,

Maria

Re: Impute Node Question - Module 3: Predictive Modeling using SAS Enterprise Miner 14

Cynthia_sas — Mon, 27 Apr 2020 01:36:28 GMT

Hi:
One of the instructors has commented:

"The default option imputed will create missing value indicator variable only when option perform imputation using the selected method.

If you select missing , the missing value indicator variable is also generated and used as input when you have missing values for a given variable but imputation is not performed (> 50 % missing values; or imputation method selected is none."

If I get more feedback, I'll post it.

Cynthia

topic Impute Node Question - Module 3: Predictive Modeling using SAS Enterprise Miner 14 in SAS Academy for Data Science

Impute Node Question - Module 3: Predictive Modeling using SAS Enterprise Miner 14

Re: Impute Node Question - Module 3: Predictive Modeling using SAS Enterprise Miner 14