scikit-learn toy datasets

scikit-learn provides some built-in datasets that can be used for prototyping purposes because they don't require very long training processes and offer different levels of complexity. They're all available in the sklearn.datasetspackage and have a common structure: the data instance variable contains the whole input set X while the target contains the labels for classification or target values for regression. For example, considering the Boston house pricing dataset (used for regression), we have the following:

from sklearn.datasets import load_boston

boston = load_boston()
X = boston.data
Y = boston.target

print(X.shape)
(506, 13)

print(Y.shape)
(506,)

In this case, we have 506 samples with 13 features and a single target value. In this book, we're going to use it for regressions and the MNIST handwritten digit dataset (load_digits()) for classification tasks. scikit-learn also provides functions for creating dummy datasets from scratch: make_classification(...