Adding a missing indicator with Feature-engine

It is video we will add a missing indicator binary variable using Feature- engine’s missing indicator transformer.

The Missing indicator transformer from Feature-engine can add binary missing indicators to both numerical and categorical variables. It can finds and select the variables automatically. Alternatively, the user can indicate for which variables the indicators should be added.

The missing indicator transformer can add indicators to all variables, or only to those that show missing data in the train set. All these characteristics can be set up with the transformer parameters.

The added indicators are marked with the variable name, plus underscore, plus na. So it is easy to understand which indicator derives from which original variable.

Feature-engine’s missing indicator transformer can be used together with the Mean or median imputation transformer or the categorical variable transformer to both flag missing values in the original variables, as well as replacing them with estimates of the variable population, leaving the variables ready to use to train machine learning models.

You can learn more about Feature-engine in the following articles:

[TraininData]

To learn more about Feature-engine visit:

[TraininData]

To learn more about Feature Engineering visit: