Built-in Datasets

Soon you will train models, and this chapter gets their data ready. scikit-learn, the ML library, ships classic datasets with it, like these iris flowers:

One line loads them. from ... import takes a single function from a module, so you call it by its bare name:

from sklearn.datasets import load_iris

iris = load_iris(as_frame=True)
print(iris.frame.shape)
print(iris.target_names)

This prints:

(150, 5)
['setosa' 'versicolor' 'virginica']

as_frame=True asks for pandas, so iris.frame is a DataFrame and iris.feature_names lists its measurements. target holds numbers, which models need: iris.target_names[2] names class 2.