Compact estimator representations

This example illustrates the use of the print_changed_only global parameter.

Setting print_changed_only to True will alternate the representation of estimators to only show the parameters that have been set to non-default values. This can be used to have more compact representations.

Out:

Default representation:
LogisticRegression(penalty='l1')

With changed_only option:
LogisticRegression(penalty='l1')
print(__doc__)

from sklearn.linear_model import LogisticRegression
from sklearn import set_config


lr = LogisticRegression(penalty='l1')
print('Default representation:')
print(lr)
# LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,
#                    intercept_scaling=1, l1_ratio=None, max_iter=100,
#                    multi_class='auto', n_jobs=None, penalty='l1',
#                    random_state=None, solver='warn', tol=0.0001, verbose=0,
#                    warm_start=False)

set_config(print_changed_only=True)
print('\nWith changed_only option:')
print(lr)
# LogisticRegression(penalty='l1')

Total running time of the script: ( 0 minutes 0.003 seconds)

© 2007–2020 The scikit-learn developers
Licensed under the 3-clause BSD License.
https://scikit-learn.org/0.24/auto_examples/miscellaneous/plot_changed_only_pprint_parameter.html