Arbitrary number imputation with Feature-engine
In this video we show how to use Feature-engine to replace missing data in numerical variables by an arbitrary number. Using the ArbitraryNumberImputer from Feature-engine, the user is able to replace missing data in 1 or more variables. The imputer can select all numerical variables automatically, or the user can enter which variables to transform in a list as an argument to the transformer.
Feature-engine makes it easy to impute different variable groups with different numbers, and can be accommodated within a Scikit-learn pipeline, smoothing model deployment.
To learn more about Feature-engine visit:
https://www.trainindata.com/feature-e...
https://feature-engine.readthedocs.io...
To learn more about Feature Engineering visit:
https://www.udemy.com/feature-enginee...