Python Feature Engineering Cookbook

How it works...

In this recipe, we applied the square and cube root transformations to variables in the Boston House Prices dataset using NumPy and scikit-learn.

To apply the square or cube root transformations with NumPy, we created a copy of the original dataframe with pandas copy() and then used sqrt() or cbrt() on a slice of the dataset with the variables to be transformed. This procedure returned a pandas dataframe with the original variables, while LSTAT, NOX, DIS, and RM were transformed with the square root.

To apply the square root transformation with scikit-learn, we used FunctionTransformer(), which applies a user-defined function – in this case, np.sqrt – to a dataset and returns the result in a NumPy array. The transform() method ...