1.. Places parent toc into the sidebar 2 3:parenttoc: True 4 5.. include:: includes/big_toc_css.rst 6 7.. _data-transforms: 8 9Dataset transformations 10----------------------- 11 12scikit-learn provides a library of transformers, which may clean (see 13:ref:`preprocessing`), reduce (see :ref:`data_reduction`), expand (see 14:ref:`kernel_approximation`) or generate (see :ref:`feature_extraction`) 15feature representations. 16 17Like other estimators, these are represented by classes with a ``fit`` method, 18which learns model parameters (e.g. mean and standard deviation for 19normalization) from a training set, and a ``transform`` method which applies 20this transformation model to unseen data. ``fit_transform`` may be more 21convenient and efficient for modelling and transforming the training data 22simultaneously. 23 24Combining such transformers, either in parallel or series is covered in 25:ref:`combining_estimators`. :ref:`metrics` covers transforming feature 26spaces into affinity matrices, while :ref:`preprocessing_targets` considers 27transformations of the target space (e.g. categorical labels) for use in 28scikit-learn. 29 30.. toctree:: 31 :maxdepth: 2 32 33 modules/compose 34 modules/feature_extraction 35 modules/preprocessing 36 modules/impute 37 modules/unsupervised_reduction 38 modules/random_projection 39 modules/kernel_approximation 40 modules/metrics 41 modules/preprocessing_targets 42