>>> from pprint import pprint
>>> from river import datasets

>>> dataset = datasets.Phishing()

>>> for x, y in dataset:
...     pprint(x)
...     print(y)
...     break
{'age_of_domain': 1,
 'anchor_from_other_domain': 0.0,
 'empty_server_form_handler': 0.0,
 'https': 0.0,
 'ip_in_url': 1,
 'is_popular': 0.5,
 'long_url': 1.0,
 'popup_window': 0.0,
 'request_from_other_domain': 0.0}
True

>>> from river import compose
>>> from river import linear_model
>>> from river import metrics
>>> from river import preprocessing

>>> model = compose.Pipeline(
...     preprocessing.StandardScaler(),
...     linear_model.LogisticRegression()
... )

>>> metric = metrics.Accuracy()

>>> for x, y in dataset:
...     y_pred = model.predict_one(x)      # make a prediction
...     metric = metric.update(y, y_pred)  # update the metric
...     model = model.learn_one(x, y)      # make the model learn

>>> metric
Accuracy: 89.20%