When instantiating polynomial features, there are three parameters to keep in mind:

• degree
• interaction_only
• include_bias

Degree corresponds to the degree of the polynomial features, with the default set to two.

interaction_only is a boolean that, when true, only interaction features are produced, meaning features that are products of degree distinct features. The default for interaction_only is false.

include_bias is also a boolean that, when true (default), includes a bias column, the feature in which all polynomial powers are zero, adding a column of all ones.

Let's set up a polynomial feature instance by first importing the class and instantiating with our parameters. At first, let's take a look at what features we get when setting interaction_only to False:

from sklearn.preprocessing import PolynomialFeatures

poly = PolynomialFeatures(degree=2, include_bias=False, interaction_only=False)

Now, we can fit_transform these polynomial features to our dataset and look at the shape of our extended dataset:

X_poly = poly.fit_transform(X)
X_poly.shape

(162501, 9)

Our dataset has now expanded to 162501 rows and 9 columns.

Let's place our data into a DataFrame, setting the column headers to the feature_names, and taking a look at the first few rows:

pd.DataFrame(X_poly, columns=poly.get_feature_names()).head()

This shows us: