dev/modules/compose.html


<!DOCTYPE html>


<html lang="en" data-content_root="../" >

  <head>
    <meta charset="utf-8" />
    <meta name="viewport" content="width=device-width, initial-scale=1.0" /><meta name="viewport" content="width=device-width, initial-scale=1" />
<meta property="og:title" content="7.1. Pipelines and composite estimators" />
<meta property="og:type" content="website" />
<meta property="og:url" content="https://scikit-learn/stable/modules/compose.html" />
<meta property="og:site_name" content="scikit-learn" />
<meta property="og:description" content="To build a composite estimator, transformers are usually combined with other transformers or with predictors(such as classifiers or regressors). The most common tool used for composing estimators i..." />
<meta property="og:image" content="https://scikit-learn.org/stable/_static/scikit-learn-logo-small.png" />
<meta property="og:image:alt" content="scikit-learn" />
<meta name="description" content="To build a composite estimator, transformers are usually combined with other transformers or with predictors(such as classifiers or regressors). The most common tool used for composing estimators i..." />

    <title>7.1. Pipelines and composite estimators &#8212; scikit-learn 1.7.dev0 documentation</title>
  
  
  <script data-cfasync="false">
    document.documentElement.dataset.mode = localStorage.getItem("mode") || "";
    document.documentElement.dataset.theme = localStorage.getItem("theme") || "";
  </script>
  <!--
    this give us a css class that will be invisible only if js is disabled
  -->
  <noscript>
    <style>
      .pst-js-only { display: none !important; }

    </style>
  </noscript>
  
  <!-- Loaded before other Sphinx assets -->
  <link href="../_static/styles/theme.css?digest=8878045cc6db502f8baf" rel="stylesheet" />
<link href="../_static/styles/pydata-sphinx-theme.css?digest=8878045cc6db502f8baf" rel="stylesheet" />

    <link rel="stylesheet" type="text/css" href="../_static/pygments.css?v=8f2a1f02" />
    <link rel="stylesheet" type="text/css" href="../_static/copybutton.css?v=76b2166b" />
    <link rel="stylesheet" type="text/css" href="../_static/plot_directive.css" />
    <link rel="stylesheet" type="text/css" href="https://fonts.googleapis.com/css?family=Vibur" />
    <link rel="stylesheet" type="text/css" href="../_static/jupyterlite_sphinx.css?v=2c9f8f05" />
    <link rel="stylesheet" type="text/css" href="../_static/sg_gallery.css?v=d2d258e8" />
    <link rel="stylesheet" type="text/css" href="../_static/sg_gallery-binder.css?v=f4aeca0c" />
    <link rel="stylesheet" type="text/css" href="../_static/sg_gallery-dataframe.css?v=2082cf3c" />
    <link rel="stylesheet" type="text/css" href="../_static/sg_gallery-rendered-html.css?v=1277b6f3" />
    <link rel="stylesheet" type="text/css" href="../_static/sphinx-design.min.css?v=95c83b7e" />
    <link rel="stylesheet" type="text/css" href="../_static/styles/colors.css?v=cc94ab7d" />
    <link rel="stylesheet" type="text/css" href="../_static/styles/custom.css?v=8f525996" />
  
  <!-- So that users can add custom icons -->
  <script src="../_static/scripts/fontawesome.js?digest=8878045cc6db502f8baf"></script>
  <!-- Pre-loaded scripts that we'll load fully later -->
  <link rel="preload" as="script" href="../_static/scripts/bootstrap.js?digest=8878045cc6db502f8baf" />
<link rel="preload" as="script" href="../_static/scripts/pydata-sphinx-theme.js?digest=8878045cc6db502f8baf" />

    <script src="../_static/documentation_options.js?v=473747f4"></script>
    <script src="../_static/doctools.js?v=9bcbadda"></script>
    <script src="../_static/sphinx_highlight.js?v=dc90522c"></script>
    <script src="../_static/clipboard.min.js?v=a7894cd8"></script>
    <script src="../_static/copybutton.js?v=97f0b27d"></script>
    <script src="../_static/jupyterlite_sphinx.js?v=96e329c5"></script>
    <script src="../_static/design-tabs.js?v=f930bc37"></script>
    <script data-domain="scikit-learn.org" defer="defer" src="https://views.scientific-python.org/js/script.js"></script>
    <script>DOCUMENTATION_OPTIONS.pagename = 'modules/compose';</script>
    <script>
        DOCUMENTATION_OPTIONS.theme_version = '0.16.1';
        DOCUMENTATION_OPTIONS.theme_switcher_json_url = 'https://scikit-learn.org/dev/_static/versions.json';
        DOCUMENTATION_OPTIONS.theme_switcher_version_match = '1.7.dev0';
        DOCUMENTATION_OPTIONS.show_version_warning_banner =
            true;
        </script>
    <script src="../_static/scripts/dropdown.js?v=d6825577"></script>
    <script src="../_static/scripts/version-switcher.js?v=a6dd8357"></script>
    <script src="../_static/scripts/sg_plotly_resize.js?v=2167d4db"></script>
    <link rel="canonical" href="https://scikit-learn.org/stable/modules/compose.html" />
    <link rel="icon" href="../_static/favicon.ico"/>
    <link rel="author" title="About these documents" href="../about.html" />
    <link rel="search" title="Search" href="../search.html" />
    <link rel="next" title="7.2. Feature extraction" href="feature_extraction.html" />
    <link rel="prev" title="7. Dataset transformations" href="../data_transforms.html" />
  <meta name="viewport" content="width=device-width, initial-scale=1"/>
  <meta name="docsearch:language" content="en"/>
  <meta name="docsearch:version" content="1.7" />
  </head>
  
  
  <body data-bs-spy="scroll" data-bs-target=".bd-toc-nav" data-offset="180" data-bs-root-margin="0px 0px -60%" data-default-mode="">

  
  <div id="pst-skip-link" class="skip-link d-print-none"><a href="#main-content">Skip to main content</a></div>
  
  <div id="pst-scroll-pixel-helper"></div>
  
  <button type="button" class="btn rounded-pill" id="pst-back-to-top">
    <i class="fa-solid fa-arrow-up"></i>Back to top</button>

  
  <dialog id="pst-search-dialog">
    
<form class="bd-search d-flex align-items-center"
      action="../search.html"
      method="get">
  <i class="fa-solid fa-magnifying-glass"></i>
  <input type="search"
         class="form-control"
         name="q"
         placeholder="Search the docs ..."
         aria-label="Search the docs ..."
         autocomplete="off"
         autocorrect="off"
         autocapitalize="off"
         spellcheck="false"/>
  <span class="search-button__kbd-shortcut"><kbd class="kbd-shortcut__modifier">Ctrl</kbd>+<kbd>K</kbd></span>
</form>
  </dialog>

  <div class="pst-async-banner-revealer d-none">
  <aside id="bd-header-version-warning" class="d-none d-print-none" aria-label="Version warning"></aside>
</div>

  
    <header class="bd-header navbar navbar-expand-lg bd-navbar d-print-none">
<div class="bd-header__inner bd-page-width">
  <button class="pst-navbar-icon sidebar-toggle primary-toggle" aria-label="Site navigation">
    <span class="fa-solid fa-bars"></span>
  </button>
  
  
  <div class=" navbar-header-items__start">
    
      <div class="navbar-item">

  
<a class="navbar-brand logo" href="../index.html">
  
  
    <img src="../_static/scikit-learn-logo-small.png" class="logo__image only-light" alt="scikit-learn homepage"/>
    <img src="../_static/scikit-learn-logo-small.png" class="logo__image only-dark pst-js-only" alt="scikit-learn homepage"/>
  
  
</a></div>
    
  </div>
  
  <div class=" navbar-header-items">
    
    <div class="me-auto navbar-header-items__center">
      
        <div class="navbar-item">
<nav>
  <ul class="bd-navbar-elements navbar-nav">
    
<li class="nav-item ">
  <a class="nav-link nav-internal" href="../install.html">
    Install
  </a>
</li>


<li class="nav-item current active">
  <a class="nav-link nav-internal" href="../user_guide.html">
    User Guide
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../api/index.html">
    API
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../auto_examples/index.html">
    Examples
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-external" href="https://blog.scikit-learn.org/">
    Community
  </a>
</li>

            <li class="nav-item dropdown">
                <button class="btn dropdown-toggle nav-item" type="button"
                data-bs-toggle="dropdown" aria-expanded="false"
                aria-controls="pst-nav-more-links">
                    More
                </button>
                <ul id="pst-nav-more-links" class="dropdown-menu">
                    
<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../getting_started.html">
    Getting Started
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../whats_new.html">
    Release History
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../glossary.html">
    Glossary
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../developers/index.html">
    Development
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../faq.html">
    FAQ
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../support.html">
    Support
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../related_projects.html">
    Related Projects
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../roadmap.html">
    Roadmap
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../governance.html">
    Governance
  </a>
</li>


<li class=" ">
  <a class="nav-link dropdown-item nav-internal" href="../about.html">
    About us
  </a>
</li>

                </ul>
            </li>
            
  </ul>
</nav></div>
      
    </div>
    
    
    <div class="navbar-header-items__end">
      
        <div class="navbar-item navbar-persistent--container">
          

<button class="btn btn-sm pst-navbar-icon search-button search-button__button pst-js-only" title="Search" aria-label="Search" data-bs-placement="bottom" data-bs-toggle="tooltip">
    <i class="fa-solid fa-magnifying-glass fa-lg"></i>
</button>
        </div>
      
      
        <div class="navbar-item">

<button class="btn btn-sm nav-link pst-navbar-icon theme-switch-button pst-js-only" aria-label="Color mode" data-bs-title="Color mode"  data-bs-placement="bottom" data-bs-toggle="tooltip">
  <i class="theme-switch fa-solid fa-sun                fa-lg" data-mode="light" title="Light"></i>
  <i class="theme-switch fa-solid fa-moon               fa-lg" data-mode="dark"  title="Dark"></i>
  <i class="theme-switch fa-solid fa-circle-half-stroke fa-lg" data-mode="auto"  title="System Settings"></i>
</button></div>
      
        <div class="navbar-item"><ul class="navbar-icon-links"
    aria-label="Icon Links">
        <li class="nav-item">
          
          
          <a href="https://github.com/scikit-learn/scikit-learn" title="GitHub" class="nav-link pst-navbar-icon" rel="noopener" target="_blank" data-bs-toggle="tooltip" data-bs-placement="bottom"><i class="fa-brands fa-square-github fa-lg" aria-hidden="true"></i>
            <span class="sr-only">GitHub</span></a>
        </li>
</ul></div>
      
        <div class="navbar-item">
<div class="version-switcher__container dropdown pst-js-only">
  <button id="pst-version-switcher-button-2"
    type="button"
    class="version-switcher__button btn btn-sm dropdown-toggle"
    data-bs-toggle="dropdown"
    aria-haspopup="listbox"
    aria-controls="pst-version-switcher-list-2"
    aria-label="Version switcher list"
  >
    Choose version  <!-- this text may get changed later by javascript -->
    <span class="caret"></span>
  </button>
  <div id="pst-version-switcher-list-2"
    class="version-switcher__menu dropdown-menu list-group-flush py-0"
    role="listbox" aria-labelledby="pst-version-switcher-button-2">
    <!-- dropdown will be populated by javascript on page load -->
  </div>
</div></div>
      
    </div>
    
  </div>
  
  
    <div class="navbar-persistent--mobile">

<button class="btn btn-sm pst-navbar-icon search-button search-button__button pst-js-only" title="Search" aria-label="Search" data-bs-placement="bottom" data-bs-toggle="tooltip">
    <i class="fa-solid fa-magnifying-glass fa-lg"></i>
</button>
    </div>
  

    <button class="pst-navbar-icon sidebar-toggle secondary-toggle" aria-label="On this page">
      <span class="fa-solid fa-outdent"></span>
    </button>
  
</div>

    </header>
  

  <div class="bd-container">
    <div class="bd-container__inner bd-page-width">
      
      
      <dialog id="pst-primary-sidebar-modal"></dialog>
      <div id="pst-primary-sidebar" class="bd-sidebar-primary bd-sidebar">
        

  <div class="sidebar-header-items sidebar-primary__section">
    
    
      <div class="sidebar-header-items__center">
        
          
            <div class="navbar-item">
<nav>
  <ul class="bd-navbar-elements navbar-nav">
    
<li class="nav-item ">
  <a class="nav-link nav-internal" href="../install.html">
    Install
  </a>
</li>


<li class="nav-item current active">
  <a class="nav-link nav-internal" href="../user_guide.html">
    User Guide
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../api/index.html">
    API
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../auto_examples/index.html">
    Examples
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-external" href="https://blog.scikit-learn.org/">
    Community
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../getting_started.html">
    Getting Started
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../whats_new.html">
    Release History
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../glossary.html">
    Glossary
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../developers/index.html">
    Development
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../faq.html">
    FAQ
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../support.html">
    Support
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../related_projects.html">
    Related Projects
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../roadmap.html">
    Roadmap
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../governance.html">
    Governance
  </a>
</li>


<li class="nav-item ">
  <a class="nav-link nav-internal" href="../about.html">
    About us
  </a>
</li>

  </ul>
</nav></div>
          
        
      </div>
    
    
      <div class="sidebar-header-items__end">
        
          <div class="navbar-item">

<button class="btn btn-sm nav-link pst-navbar-icon theme-switch-button pst-js-only" aria-label="Color mode" data-bs-title="Color mode"  data-bs-placement="bottom" data-bs-toggle="tooltip">
  <i class="theme-switch fa-solid fa-sun                fa-lg" data-mode="light" title="Light"></i>
  <i class="theme-switch fa-solid fa-moon               fa-lg" data-mode="dark"  title="Dark"></i>
  <i class="theme-switch fa-solid fa-circle-half-stroke fa-lg" data-mode="auto"  title="System Settings"></i>
</button></div>
        
          <div class="navbar-item"><ul class="navbar-icon-links"
    aria-label="Icon Links">
        <li class="nav-item">
          
          
          <a href="https://github.com/scikit-learn/scikit-learn" title="GitHub" class="nav-link pst-navbar-icon" rel="noopener" target="_blank" data-bs-toggle="tooltip" data-bs-placement="bottom"><i class="fa-brands fa-square-github fa-lg" aria-hidden="true"></i>
            <span class="sr-only">GitHub</span></a>
        </li>
</ul></div>
        
          <div class="navbar-item">
<div class="version-switcher__container dropdown pst-js-only">
  <button id="pst-version-switcher-button-3"
    type="button"
    class="version-switcher__button btn btn-sm dropdown-toggle"
    data-bs-toggle="dropdown"
    aria-haspopup="listbox"
    aria-controls="pst-version-switcher-list-3"
    aria-label="Version switcher list"
  >
    Choose version  <!-- this text may get changed later by javascript -->
    <span class="caret"></span>
  </button>
  <div id="pst-version-switcher-list-3"
    class="version-switcher__menu dropdown-menu list-group-flush py-0"
    role="listbox" aria-labelledby="pst-version-switcher-button-3">
    <!-- dropdown will be populated by javascript on page load -->
  </div>
</div></div>
        
      </div>
    
  </div>
  
    <div class="sidebar-primary-items__start sidebar-primary__section">
        <div class="sidebar-primary-item">
<nav class="bd-docs-nav bd-links"
     aria-label="Section Navigation">
  <p class="bd-links__title" role="heading" aria-level="1">Section Navigation</p>
  <div class="bd-toc-item navbar-nav"><ul class="current nav bd-sidenav">
<li class="toctree-l1 has-children"><a class="reference internal" href="../supervised_learning.html">1. Supervised learning</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="linear_model.html">1.1. Linear Models</a></li>
<li class="toctree-l2"><a class="reference internal" href="lda_qda.html">1.2. Linear and Quadratic Discriminant Analysis</a></li>
<li class="toctree-l2"><a class="reference internal" href="kernel_ridge.html">1.3. Kernel ridge regression</a></li>
<li class="toctree-l2"><a class="reference internal" href="svm.html">1.4. Support Vector Machines</a></li>
<li class="toctree-l2"><a class="reference internal" href="sgd.html">1.5. Stochastic Gradient Descent</a></li>
<li class="toctree-l2"><a class="reference internal" href="neighbors.html">1.6. Nearest Neighbors</a></li>
<li class="toctree-l2"><a class="reference internal" href="gaussian_process.html">1.7. Gaussian Processes</a></li>
<li class="toctree-l2"><a class="reference internal" href="cross_decomposition.html">1.8. Cross decomposition</a></li>
<li class="toctree-l2"><a class="reference internal" href="naive_bayes.html">1.9. Naive Bayes</a></li>
<li class="toctree-l2"><a class="reference internal" href="tree.html">1.10. Decision Trees</a></li>
<li class="toctree-l2"><a class="reference internal" href="ensemble.html">1.11. Ensembles: Gradient boosting, random forests, bagging, voting, stacking</a></li>
<li class="toctree-l2"><a class="reference internal" href="multiclass.html">1.12. Multiclass and multioutput algorithms</a></li>
<li class="toctree-l2"><a class="reference internal" href="feature_selection.html">1.13. Feature selection</a></li>
<li class="toctree-l2"><a class="reference internal" href="semi_supervised.html">1.14. Semi-supervised learning</a></li>
<li class="toctree-l2"><a class="reference internal" href="isotonic.html">1.15. Isotonic regression</a></li>
<li class="toctree-l2"><a class="reference internal" href="calibration.html">1.16. Probability calibration</a></li>
<li class="toctree-l2"><a class="reference internal" href="neural_networks_supervised.html">1.17. Neural network models (supervised)</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../unsupervised_learning.html">2. Unsupervised learning</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="mixture.html">2.1. Gaussian mixture models</a></li>
<li class="toctree-l2"><a class="reference internal" href="manifold.html">2.2. Manifold learning</a></li>
<li class="toctree-l2"><a class="reference internal" href="clustering.html">2.3. Clustering</a></li>
<li class="toctree-l2"><a class="reference internal" href="biclustering.html">2.4. Biclustering</a></li>
<li class="toctree-l2"><a class="reference internal" href="decomposition.html">2.5. Decomposing signals in components (matrix factorization problems)</a></li>
<li class="toctree-l2"><a class="reference internal" href="covariance.html">2.6. Covariance estimation</a></li>
<li class="toctree-l2"><a class="reference internal" href="outlier_detection.html">2.7. Novelty and Outlier Detection</a></li>
<li class="toctree-l2"><a class="reference internal" href="density.html">2.8. Density Estimation</a></li>
<li class="toctree-l2"><a class="reference internal" href="neural_networks_unsupervised.html">2.9. Neural network models (unsupervised)</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../model_selection.html">3. Model selection and evaluation</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="cross_validation.html">3.1. Cross-validation: evaluating estimator performance</a></li>
<li class="toctree-l2"><a class="reference internal" href="grid_search.html">3.2. Tuning the hyper-parameters of an estimator</a></li>
<li class="toctree-l2"><a class="reference internal" href="classification_threshold.html">3.3. Tuning the decision threshold for class prediction</a></li>
<li class="toctree-l2"><a class="reference internal" href="model_evaluation.html">3.4. Metrics and scoring: quantifying the quality of predictions</a></li>
<li class="toctree-l2"><a class="reference internal" href="learning_curve.html">3.5. Validation curves: plotting scores to evaluate models</a></li>
</ul>
</details></li>
<li class="toctree-l1"><a class="reference internal" href="../metadata_routing.html">4. Metadata Routing</a></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../inspection.html">5. Inspection</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="partial_dependence.html">5.1. Partial Dependence and Individual Conditional Expectation plots</a></li>
<li class="toctree-l2"><a class="reference internal" href="permutation_importance.html">5.2. Permutation feature importance</a></li>
</ul>
</details></li>
<li class="toctree-l1"><a class="reference internal" href="../visualizations.html">6. Visualizations</a></li>
<li class="toctree-l1 current active has-children"><a class="reference internal" href="../data_transforms.html">7. Dataset transformations</a><details open="open"><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul class="current">
<li class="toctree-l2 current active"><a class="current reference internal" href="#">7.1. Pipelines and composite estimators</a></li>
<li class="toctree-l2"><a class="reference internal" href="feature_extraction.html">7.2. Feature extraction</a></li>
<li class="toctree-l2"><a class="reference internal" href="preprocessing.html">7.3. Preprocessing data</a></li>
<li class="toctree-l2"><a class="reference internal" href="impute.html">7.4. Imputation of missing values</a></li>
<li class="toctree-l2"><a class="reference internal" href="unsupervised_reduction.html">7.5. Unsupervised dimensionality reduction</a></li>
<li class="toctree-l2"><a class="reference internal" href="random_projection.html">7.6. Random Projection</a></li>
<li class="toctree-l2"><a class="reference internal" href="kernel_approximation.html">7.7. Kernel Approximation</a></li>
<li class="toctree-l2"><a class="reference internal" href="metrics.html">7.8. Pairwise metrics, Affinities and Kernels</a></li>
<li class="toctree-l2"><a class="reference internal" href="preprocessing_targets.html">7.9. Transforming the prediction target (<code class="docutils literal notranslate"><span class="pre">y</span></code>)</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../datasets.html">8. Dataset loading utilities</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="../datasets/toy_dataset.html">8.1. Toy datasets</a></li>
<li class="toctree-l2"><a class="reference internal" href="../datasets/real_world.html">8.2. Real world datasets</a></li>
<li class="toctree-l2"><a class="reference internal" href="../datasets/sample_generators.html">8.3. Generated datasets</a></li>
<li class="toctree-l2"><a class="reference internal" href="../datasets/loading_other_datasets.html">8.4. Loading other datasets</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../computing.html">9. Computing with scikit-learn</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="../computing/scaling_strategies.html">9.1. Strategies to scale computationally: bigger data</a></li>
<li class="toctree-l2"><a class="reference internal" href="../computing/computational_performance.html">9.2. Computational Performance</a></li>
<li class="toctree-l2"><a class="reference internal" href="../computing/parallelism.html">9.3. Parallelism, resource management, and configuration</a></li>
</ul>
</details></li>
<li class="toctree-l1"><a class="reference internal" href="../model_persistence.html">10. Model persistence</a></li>
<li class="toctree-l1"><a class="reference internal" href="../common_pitfalls.html">11. Common pitfalls and recommended practices</a></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../dispatching.html">12. Dispatching</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="array_api.html">12.1. Array API support (experimental)</a></li>
</ul>
</details></li>
<li class="toctree-l1"><a class="reference internal" href="../machine_learning_map.html">13. Choosing the right estimator</a></li>
<li class="toctree-l1"><a class="reference internal" href="../presentations.html">14. External Resources, Videos and Talks</a></li>
</ul>
</div>
</nav></div>
    </div>
  
  
  <div class="sidebar-primary-items__end sidebar-primary__section">
  </div>


      </div>
      
      <main id="main-content" class="bd-main" role="main">
        
        
          <div class="bd-content">
            <div class="bd-article-container">
              
              <div class="bd-header-article d-print-none">
<div class="header-article-items header-article__inner">
  
    <div class="header-article-items__start">
      
        <div class="header-article-item">

<nav aria-label="Breadcrumb" class="d-print-none">
  <ul class="bd-breadcrumbs">
    
    <li class="breadcrumb-item breadcrumb-home">
      <a href="../index.html" class="nav-link" aria-label="Home">
        <i class="fa-solid fa-home"></i>
      </a>
    </li>
    
    <li class="breadcrumb-item"><a href="../user_guide.html" class="nav-link">User Guide</a></li>
    
    
    <li class="breadcrumb-item"><a href="../data_transforms.html" class="nav-link"><span class="section-number">7. </span>Dataset transformations</a></li>
    
    <li class="breadcrumb-item active" aria-current="page"><span class="ellipsis"><span class="section-number">7.1. </span>Pipelines and composite estimators</span></li>
  </ul>
</nav>
</div>
      
    </div>
  
  
</div>
</div>
              
              
<div id="searchbox"></div>
                <article class="bd-article">
                  
  <section id="pipelines-and-composite-estimators">
<span id="combining-estimators"></span><h1><span class="section-number">7.1. </span>Pipelines and composite estimators<a class="headerlink" href="#pipelines-and-composite-estimators" title="Link to this heading">#</a></h1>
<p>To build a composite estimator, transformers are usually combined with other
transformers or with <a class="reference internal" href="../glossary.html#term-predictors"><span class="xref std std-term">predictors</span></a> (such as classifiers or regressors).
The most common tool used for composing estimators is a <a class="reference internal" href="#pipeline"><span class="std std-ref">Pipeline</span></a>. Pipelines require all steps except the last to be a
<a class="reference internal" href="../glossary.html#term-transformer"><span class="xref std std-term">transformer</span></a>. The last step can be anything, a transformer, a
<a class="reference internal" href="../glossary.html#term-predictor"><span class="xref std std-term">predictor</span></a>, or a clustering estimator which might have or not have a
<code class="docutils literal notranslate"><span class="pre">.predict(...)</span></code> method. A pipeline exposes all methods provided by the last
estimator: if the last step provides a <code class="docutils literal notranslate"><span class="pre">transform</span></code> method, then the pipeline
would have a <code class="docutils literal notranslate"><span class="pre">transform</span></code> method and behave like a transformer. If the last step
provides a <code class="docutils literal notranslate"><span class="pre">predict</span></code> method, then the pipeline would expose that method, and
given a data <a class="reference internal" href="../glossary.html#term-X"><span class="xref std std-term">X</span></a>, use all steps except the last to transform the data,
and then give that transformed data to the <code class="docutils literal notranslate"><span class="pre">predict</span></code> method of the last step of
the pipeline. The class <code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code> is often used in combination with
<a class="reference internal" href="#column-transformer"><span class="std std-ref">ColumnTransformer</span></a> or
<a class="reference internal" href="#feature-union"><span class="std std-ref">FeatureUnion</span></a> which concatenate the output of transformers
into a composite feature space.
<a class="reference internal" href="#transformed-target-regressor"><span class="std std-ref">TransformedTargetRegressor</span></a>
deals with transforming the <a class="reference internal" href="../glossary.html#term-target"><span class="xref std std-term">target</span></a> (i.e. log-transform <a class="reference internal" href="../glossary.html#term-y"><span class="xref std std-term">y</span></a>).</p>
<section id="pipeline-chaining-estimators">
<span id="pipeline"></span><h2><span class="section-number">7.1.1. </span>Pipeline: chaining estimators<a class="headerlink" href="#pipeline-chaining-estimators" title="Link to this heading">#</a></h2>
<p><a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> can be used to chain multiple estimators
into one. This is useful as there is often a fixed sequence
of steps in processing the data, for example feature selection, normalization
and classification. <a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> serves multiple purposes here:</p>
<dl class="simple">
<dt>Convenience and encapsulation</dt><dd><p>You only have to call <a class="reference internal" href="../glossary.html#term-fit"><span class="xref std std-term">fit</span></a> and <a class="reference internal" href="../glossary.html#term-predict"><span class="xref std std-term">predict</span></a> once on your
data to fit a whole sequence of estimators.</p>
</dd>
<dt>Joint parameter selection</dt><dd><p>You can <a class="reference internal" href="grid_search.html#grid-search"><span class="std std-ref">grid search</span></a>
over parameters of all estimators in the pipeline at once.</p>
</dd>
<dt>Safety</dt><dd><p>Pipelines help avoid leaking statistics from your test data into the
trained model in cross-validation, by ensuring that the same samples are
used to train the transformers and predictors.</p>
</dd>
</dl>
<p>All estimators in a pipeline, except the last one, must be transformers
(i.e. must have a <a class="reference internal" href="../glossary.html#term-transform"><span class="xref std std-term">transform</span></a> method).
The last estimator may be any type (transformer, classifier, etc.).</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Calling <code class="docutils literal notranslate"><span class="pre">fit</span></code> on the pipeline is the same as calling <code class="docutils literal notranslate"><span class="pre">fit</span></code> on
each estimator in turn, <code class="docutils literal notranslate"><span class="pre">transform</span></code> the input and pass it on to the next step.
The pipeline has all the methods that the last estimator in the pipeline has,
i.e. if the last estimator is a classifier, the <a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> can be used
as a classifier. If the last estimator is a transformer, again, so is the
pipeline.</p>
</div>
<section id="usage">
<h3><span class="section-number">7.1.1.1. </span>Usage<a class="headerlink" href="#usage" title="Link to this heading">#</a></h3>
<section id="build-a-pipeline">
<h4><span class="section-number">7.1.1.1.1. </span>Build a pipeline<a class="headerlink" href="#build-a-pipeline" title="Link to this heading">#</a></h4>
<p>The <a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> is built using a list of <code class="docutils literal notranslate"><span class="pre">(key,</span> <span class="pre">value)</span></code> pairs, where
the <code class="docutils literal notranslate"><span class="pre">key</span></code> is a string containing the name you want to give this step and <code class="docutils literal notranslate"><span class="pre">value</span></code>
is an estimator object:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.pipeline</span><span class="w"> </span><span class="kn">import</span> <span class="n">Pipeline</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.svm</span><span class="w"> </span><span class="kn">import</span> <span class="n">SVC</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.decomposition</span><span class="w"> </span><span class="kn">import</span> <span class="n">PCA</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">estimators</span> <span class="o">=</span> <span class="p">[(</span><span class="s1">&#39;reduce_dim&#39;</span><span class="p">,</span> <span class="n">PCA</span><span class="p">()),</span> <span class="p">(</span><span class="s1">&#39;clf&#39;</span><span class="p">,</span> <span class="n">SVC</span><span class="p">())]</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span> <span class="o">=</span> <span class="n">Pipeline</span><span class="p">(</span><span class="n">estimators</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span>
<span class="go">Pipeline(steps=[(&#39;reduce_dim&#39;, PCA()), (&#39;clf&#39;, SVC())])</span>
</pre></div>
</div>
<details class="sd-sphinx-override sd-dropdown sd-card sd-mb-3" id="shorthand-version-using-make_pipeline">
<summary class="sd-summary-title sd-card-header">
<span class="sd-summary-text">Shorthand version using <a class="reference internal" href="generated/sklearn.pipeline.make_pipeline.html#sklearn.pipeline.make_pipeline" title="sklearn.pipeline.make_pipeline"><code class="xref py py-func docutils literal notranslate"><span class="pre">make_pipeline</span></code></a><a class="headerlink" href="#shorthand-version-using-make_pipeline" title="Link to this dropdown">#</a></span><span class="sd-summary-state-marker sd-summary-chevron-right"><svg version="1.1" width="1.5em" height="1.5em" class="sd-octicon sd-octicon-chevron-right" viewBox="0 0 24 24" aria-hidden="true"><path d="M8.72 18.78a.75.75 0 0 1 0-1.06L14.44 12 8.72 6.28a.751.751 0 0 1 .018-1.042.751.751 0 0 1 1.042-.018l6.25 6.25a.75.75 0 0 1 0 1.06l-6.25 6.25a.75.75 0 0 1-1.06 0Z"></path></svg></span></summary><div class="sd-summary-content sd-card-body docutils">
<p class="sd-card-text">The utility function <a class="reference internal" href="generated/sklearn.pipeline.make_pipeline.html#sklearn.pipeline.make_pipeline" title="sklearn.pipeline.make_pipeline"><code class="xref py py-func docutils literal notranslate"><span class="pre">make_pipeline</span></code></a> is a shorthand
for constructing pipelines;
it takes a variable number of estimators and returns a pipeline,
filling in the names automatically:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.pipeline</span><span class="w"> </span><span class="kn">import</span> <span class="n">make_pipeline</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">make_pipeline</span><span class="p">(</span><span class="n">PCA</span><span class="p">(),</span> <span class="n">SVC</span><span class="p">())</span>
<span class="go">Pipeline(steps=[(&#39;pca&#39;, PCA()), (&#39;svc&#39;, SVC())])</span>
</pre></div>
</div>
</div>
</details></section>
<section id="access-pipeline-steps">
<h4><span class="section-number">7.1.1.1.2. </span>Access pipeline steps<a class="headerlink" href="#access-pipeline-steps" title="Link to this heading">#</a></h4>
<p>The estimators of a pipeline are stored as a list in the <code class="docutils literal notranslate"><span class="pre">steps</span></code> attribute.
A sub-pipeline can be extracted using the slicing notation commonly used
for Python Sequences such as lists or strings (although only a step of 1 is
permitted). This is convenient for performing only some of the transformations
(or their inverse):</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="p">[:</span><span class="mi">1</span><span class="p">]</span>
<span class="go">Pipeline(steps=[(&#39;reduce_dim&#39;, PCA())])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="p">[</span><span class="o">-</span><span class="mi">1</span><span class="p">:]</span>
<span class="go">Pipeline(steps=[(&#39;clf&#39;, SVC())])</span>
</pre></div>
</div>
<details class="sd-sphinx-override sd-dropdown sd-card sd-mb-3" id="accessing-a-step-by-name-or-position">
<summary class="sd-summary-title sd-card-header">
<span class="sd-summary-text">Accessing a step by name or position<a class="headerlink" href="#accessing-a-step-by-name-or-position" title="Link to this dropdown">#</a></span><span class="sd-summary-state-marker sd-summary-chevron-right"><svg version="1.1" width="1.5em" height="1.5em" class="sd-octicon sd-octicon-chevron-right" viewBox="0 0 24 24" aria-hidden="true"><path d="M8.72 18.78a.75.75 0 0 1 0-1.06L14.44 12 8.72 6.28a.751.751 0 0 1 .018-1.042.751.751 0 0 1 1.042-.018l6.25 6.25a.75.75 0 0 1 0 1.06l-6.25 6.25a.75.75 0 0 1-1.06 0Z"></path></svg></span></summary><div class="sd-summary-content sd-card-body docutils">
<p class="sd-card-text">A specific step can also be accessed by index or name by indexing (with <code class="docutils literal notranslate"><span class="pre">[idx]</span></code>) the
pipeline:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="o">.</span><span class="n">steps</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span>
<span class="go">(&#39;reduce_dim&#39;, PCA())</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span>
<span class="go">PCA()</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="p">[</span><span class="s1">&#39;reduce_dim&#39;</span><span class="p">]</span>
<span class="go">PCA()</span>
</pre></div>
</div>
<p class="sd-card-text"><code class="docutils literal notranslate"><span class="pre">Pipeline</span></code>’s <code class="docutils literal notranslate"><span class="pre">named_steps</span></code> attribute allows accessing steps by name with tab
completion in interactive environments:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="o">.</span><span class="n">named_steps</span><span class="o">.</span><span class="n">reduce_dim</span> <span class="ow">is</span> <span class="n">pipe</span><span class="p">[</span><span class="s1">&#39;reduce_dim&#39;</span><span class="p">]</span>
<span class="go">True</span>
</pre></div>
</div>
</div>
</details></section>
<section id="tracking-feature-names-in-a-pipeline">
<h4><span class="section-number">7.1.1.1.3. </span>Tracking feature names in a pipeline<a class="headerlink" href="#tracking-feature-names-in-a-pipeline" title="Link to this heading">#</a></h4>
<p>To enable model inspection, <a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> has a
<code class="docutils literal notranslate"><span class="pre">get_feature_names_out()</span></code> method, just like all transformers. You can use
pipeline slicing to get the feature names going into each step:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.datasets</span><span class="w"> </span><span class="kn">import</span> <span class="n">load_iris</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.linear_model</span><span class="w"> </span><span class="kn">import</span> <span class="n">LogisticRegression</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.feature_selection</span><span class="w"> </span><span class="kn">import</span> <span class="n">SelectKBest</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">iris</span> <span class="o">=</span> <span class="n">load_iris</span><span class="p">()</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span> <span class="o">=</span> <span class="n">Pipeline</span><span class="p">(</span><span class="n">steps</span><span class="o">=</span><span class="p">[</span>
<span class="gp">... </span>   <span class="p">(</span><span class="s1">&#39;select&#39;</span><span class="p">,</span> <span class="n">SelectKBest</span><span class="p">(</span><span class="n">k</span><span class="o">=</span><span class="mi">2</span><span class="p">)),</span>
<span class="gp">... </span>   <span class="p">(</span><span class="s1">&#39;clf&#39;</span><span class="p">,</span> <span class="n">LogisticRegression</span><span class="p">())])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">iris</span><span class="o">.</span><span class="n">data</span><span class="p">,</span> <span class="n">iris</span><span class="o">.</span><span class="n">target</span><span class="p">)</span>
<span class="go">Pipeline(steps=[(&#39;select&#39;, SelectKBest(...)), (&#39;clf&#39;, LogisticRegression(...))])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="p">[:</span><span class="o">-</span><span class="mi">1</span><span class="p">]</span><span class="o">.</span><span class="n">get_feature_names_out</span><span class="p">()</span>
<span class="go">array([&#39;x2&#39;, &#39;x3&#39;], ...)</span>
</pre></div>
</div>
<details class="sd-sphinx-override sd-dropdown sd-card sd-mb-3" id="customize-feature-names">
<summary class="sd-summary-title sd-card-header">
<span class="sd-summary-text">Customize feature names<a class="headerlink" href="#customize-feature-names" title="Link to this dropdown">#</a></span><span class="sd-summary-state-marker sd-summary-chevron-right"><svg version="1.1" width="1.5em" height="1.5em" class="sd-octicon sd-octicon-chevron-right" viewBox="0 0 24 24" aria-hidden="true"><path d="M8.72 18.78a.75.75 0 0 1 0-1.06L14.44 12 8.72 6.28a.751.751 0 0 1 .018-1.042.751.751 0 0 1 1.042-.018l6.25 6.25a.75.75 0 0 1 0 1.06l-6.25 6.25a.75.75 0 0 1-1.06 0Z"></path></svg></span></summary><div class="sd-summary-content sd-card-body docutils">
<p class="sd-card-text">You can also provide custom feature names for the input data using
<code class="docutils literal notranslate"><span class="pre">get_feature_names_out</span></code>:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="p">[:</span><span class="o">-</span><span class="mi">1</span><span class="p">]</span><span class="o">.</span><span class="n">get_feature_names_out</span><span class="p">(</span><span class="n">iris</span><span class="o">.</span><span class="n">feature_names</span><span class="p">)</span>
<span class="go">array([&#39;petal length (cm)&#39;, &#39;petal width (cm)&#39;], ...)</span>
</pre></div>
</div>
</div>
</details></section>
<section id="access-to-nested-parameters">
<span id="pipeline-nested-parameters"></span><h4><span class="section-number">7.1.1.1.4. </span>Access to nested parameters<a class="headerlink" href="#access-to-nested-parameters" title="Link to this heading">#</a></h4>
<p>It is common to adjust the parameters of an estimator within a pipeline. This parameter
is therefore nested because it belongs to a particular sub-step. Parameters of the
estimators in the pipeline are accessible using the <code class="docutils literal notranslate"><span class="pre">&lt;estimator&gt;__&lt;parameter&gt;</span></code>
syntax:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span> <span class="o">=</span> <span class="n">Pipeline</span><span class="p">(</span><span class="n">steps</span><span class="o">=</span><span class="p">[(</span><span class="s2">&quot;reduce_dim&quot;</span><span class="p">,</span> <span class="n">PCA</span><span class="p">()),</span> <span class="p">(</span><span class="s2">&quot;clf&quot;</span><span class="p">,</span> <span class="n">SVC</span><span class="p">())])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="o">.</span><span class="n">set_params</span><span class="p">(</span><span class="n">clf__C</span><span class="o">=</span><span class="mi">10</span><span class="p">)</span>
<span class="go">Pipeline(steps=[(&#39;reduce_dim&#39;, PCA()), (&#39;clf&#39;, SVC(C=10))])</span>
</pre></div>
</div>
<details class="sd-sphinx-override sd-dropdown sd-card sd-mb-3" id="when-does-it-matter?">
<summary class="sd-summary-title sd-card-header">
<span class="sd-summary-text">When does it matter?<a class="headerlink" href="#when-does-it-matter?" title="Link to this dropdown">#</a></span><span class="sd-summary-state-marker sd-summary-chevron-right"><svg version="1.1" width="1.5em" height="1.5em" class="sd-octicon sd-octicon-chevron-right" viewBox="0 0 24 24" aria-hidden="true"><path d="M8.72 18.78a.75.75 0 0 1 0-1.06L14.44 12 8.72 6.28a.751.751 0 0 1 .018-1.042.751.751 0 0 1 1.042-.018l6.25 6.25a.75.75 0 0 1 0 1.06l-6.25 6.25a.75.75 0 0 1-1.06 0Z"></path></svg></span></summary><div class="sd-summary-content sd-card-body docutils">
<p class="sd-card-text">This is particularly important for doing grid searches:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.model_selection</span><span class="w"> </span><span class="kn">import</span> <span class="n">GridSearchCV</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">param_grid</span> <span class="o">=</span> <span class="nb">dict</span><span class="p">(</span><span class="n">reduce_dim__n_components</span><span class="o">=</span><span class="p">[</span><span class="mi">2</span><span class="p">,</span> <span class="mi">5</span><span class="p">,</span> <span class="mi">10</span><span class="p">],</span>
<span class="gp">... </span>                  <span class="n">clf__C</span><span class="o">=</span><span class="p">[</span><span class="mf">0.1</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="mi">100</span><span class="p">])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">grid_search</span> <span class="o">=</span> <span class="n">GridSearchCV</span><span class="p">(</span><span class="n">pipe</span><span class="p">,</span> <span class="n">param_grid</span><span class="o">=</span><span class="n">param_grid</span><span class="p">)</span>
</pre></div>
</div>
<p class="sd-card-text">Individual steps may also be replaced as parameters, and non-final steps may be
ignored by setting them to <code class="docutils literal notranslate"><span class="pre">'passthrough'</span></code>:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">param_grid</span> <span class="o">=</span> <span class="nb">dict</span><span class="p">(</span><span class="n">reduce_dim</span><span class="o">=</span><span class="p">[</span><span class="s1">&#39;passthrough&#39;</span><span class="p">,</span> <span class="n">PCA</span><span class="p">(</span><span class="mi">5</span><span class="p">),</span> <span class="n">PCA</span><span class="p">(</span><span class="mi">10</span><span class="p">)],</span>
<span class="gp">... </span>                  <span class="n">clf</span><span class="o">=</span><span class="p">[</span><span class="n">SVC</span><span class="p">(),</span> <span class="n">LogisticRegression</span><span class="p">()],</span>
<span class="gp">... </span>                  <span class="n">clf__C</span><span class="o">=</span><span class="p">[</span><span class="mf">0.1</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="mi">100</span><span class="p">])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">grid_search</span> <span class="o">=</span> <span class="n">GridSearchCV</span><span class="p">(</span><span class="n">pipe</span><span class="p">,</span> <span class="n">param_grid</span><span class="o">=</span><span class="n">param_grid</span><span class="p">)</span>
</pre></div>
</div>
<div class="admonition seealso">
<p class="admonition-title">See also</p>
<ul class="simple">
<li><p class="sd-card-text"><a class="reference internal" href="grid_search.html#composite-grid-search"><span class="std std-ref">Composite estimators and parameter spaces</span></a></p></li>
</ul>
</div>
</div>
</details><p class="rubric">Examples</p>
<ul class="simple">
<li><p><a class="reference internal" href="../auto_examples/feature_selection/plot_feature_selection_pipeline.html#sphx-glr-auto-examples-feature-selection-plot-feature-selection-pipeline-py"><span class="std std-ref">Pipeline ANOVA SVM</span></a></p></li>
<li><p><a class="reference internal" href="../auto_examples/model_selection/plot_grid_search_text_feature_extraction.html#sphx-glr-auto-examples-model-selection-plot-grid-search-text-feature-extraction-py"><span class="std std-ref">Sample pipeline for text feature extraction and evaluation</span></a></p></li>
<li><p><a class="reference internal" href="../auto_examples/compose/plot_digits_pipe.html#sphx-glr-auto-examples-compose-plot-digits-pipe-py"><span class="std std-ref">Pipelining: chaining a PCA and a logistic regression</span></a></p></li>
<li><p><a class="reference internal" href="../auto_examples/miscellaneous/plot_kernel_approximation.html#sphx-glr-auto-examples-miscellaneous-plot-kernel-approximation-py"><span class="std std-ref">Explicit feature map approximation for RBF kernels</span></a></p></li>
<li><p><a class="reference internal" href="../auto_examples/svm/plot_svm_anova.html#sphx-glr-auto-examples-svm-plot-svm-anova-py"><span class="std std-ref">SVM-Anova: SVM with univariate feature selection</span></a></p></li>
<li><p><a class="reference internal" href="../auto_examples/compose/plot_compare_reduction.html#sphx-glr-auto-examples-compose-plot-compare-reduction-py"><span class="std std-ref">Selecting dimensionality reduction with Pipeline and GridSearchCV</span></a></p></li>
<li><p><a class="reference internal" href="../auto_examples/miscellaneous/plot_pipeline_display.html#sphx-glr-auto-examples-miscellaneous-plot-pipeline-display-py"><span class="std std-ref">Displaying Pipelines</span></a></p></li>
</ul>
</section>
</section>
<section id="caching-transformers-avoid-repeated-computation">
<span id="pipeline-cache"></span><h3><span class="section-number">7.1.1.2. </span>Caching transformers: avoid repeated computation<a class="headerlink" href="#caching-transformers-avoid-repeated-computation" title="Link to this heading">#</a></h3>
<p>Fitting transformers may be computationally expensive. With its
<code class="docutils literal notranslate"><span class="pre">memory</span></code> parameter set, <a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> will cache each transformer
after calling <code class="docutils literal notranslate"><span class="pre">fit</span></code>.
This feature is used to avoid computing the fit transformers within a pipeline
if the parameters and input data are identical. A typical example is the case of
a grid search in which the transformers can be fitted only once and reused for
each configuration. The last step will never be cached, even if it is a transformer.</p>
<p>The parameter <code class="docutils literal notranslate"><span class="pre">memory</span></code> is needed in order to cache the transformers.
<code class="docutils literal notranslate"><span class="pre">memory</span></code> can be either a string containing the directory where to cache the
transformers or a <a class="reference external" href="https://joblib.readthedocs.io/en/latest/memory.html">joblib.Memory</a>
object:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">tempfile</span><span class="w"> </span><span class="kn">import</span> <span class="n">mkdtemp</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">shutil</span><span class="w"> </span><span class="kn">import</span> <span class="n">rmtree</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.decomposition</span><span class="w"> </span><span class="kn">import</span> <span class="n">PCA</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.svm</span><span class="w"> </span><span class="kn">import</span> <span class="n">SVC</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.pipeline</span><span class="w"> </span><span class="kn">import</span> <span class="n">Pipeline</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">estimators</span> <span class="o">=</span> <span class="p">[(</span><span class="s1">&#39;reduce_dim&#39;</span><span class="p">,</span> <span class="n">PCA</span><span class="p">()),</span> <span class="p">(</span><span class="s1">&#39;clf&#39;</span><span class="p">,</span> <span class="n">SVC</span><span class="p">())]</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">cachedir</span> <span class="o">=</span> <span class="n">mkdtemp</span><span class="p">()</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span> <span class="o">=</span> <span class="n">Pipeline</span><span class="p">(</span><span class="n">estimators</span><span class="p">,</span> <span class="n">memory</span><span class="o">=</span><span class="n">cachedir</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span>
<span class="go">Pipeline(memory=...,</span>
<span class="go">         steps=[(&#39;reduce_dim&#39;, PCA()), (&#39;clf&#39;, SVC())])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="c1"># Clear the cache directory when you don&#39;t need it anymore</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">rmtree</span><span class="p">(</span><span class="n">cachedir</span><span class="p">)</span>
</pre></div>
</div>
<details class="sd-sphinx-override sd-dropdown sd-card sd-mb-3" id="side-effect-of-caching-transformers">
<summary class="sd-summary-title sd-card-header sd-bg-warning sd-bg-text-warning">
<span class="sd-summary-text">Side effect of caching transformers<a class="headerlink" href="#side-effect-of-caching-transformers" title="Link to this dropdown">#</a></span><span class="sd-summary-state-marker sd-summary-chevron-right"><svg version="1.1" width="1.5em" height="1.5em" class="sd-octicon sd-octicon-chevron-right" viewBox="0 0 24 24" aria-hidden="true"><path d="M8.72 18.78a.75.75 0 0 1 0-1.06L14.44 12 8.72 6.28a.751.751 0 0 1 .018-1.042.751.751 0 0 1 1.042-.018l6.25 6.25a.75.75 0 0 1 0 1.06l-6.25 6.25a.75.75 0 0 1-1.06 0Z"></path></svg></span></summary><div class="sd-summary-content sd-card-body docutils">
<p class="sd-card-text">Using a <a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> without cache enabled, it is possible to
inspect the original instance such as:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.datasets</span><span class="w"> </span><span class="kn">import</span> <span class="n">load_digits</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">X_digits</span><span class="p">,</span> <span class="n">y_digits</span> <span class="o">=</span> <span class="n">load_digits</span><span class="p">(</span><span class="n">return_X_y</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pca1</span> <span class="o">=</span> <span class="n">PCA</span><span class="p">(</span><span class="n">n_components</span><span class="o">=</span><span class="mi">10</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">svm1</span> <span class="o">=</span> <span class="n">SVC</span><span class="p">()</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span> <span class="o">=</span> <span class="n">Pipeline</span><span class="p">([(</span><span class="s1">&#39;reduce_dim&#39;</span><span class="p">,</span> <span class="n">pca1</span><span class="p">),</span> <span class="p">(</span><span class="s1">&#39;clf&#39;</span><span class="p">,</span> <span class="n">svm1</span><span class="p">)])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pipe</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X_digits</span><span class="p">,</span> <span class="n">y_digits</span><span class="p">)</span>
<span class="go">Pipeline(steps=[(&#39;reduce_dim&#39;, PCA(n_components=10)), (&#39;clf&#39;, SVC())])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="c1"># The pca instance can be inspected directly</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pca1</span><span class="o">.</span><span class="n">components_</span><span class="o">.</span><span class="n">shape</span>
<span class="go">(10, 64)</span>
</pre></div>
</div>
<p class="sd-card-text">Enabling caching triggers a clone of the transformers before fitting.
Therefore, the transformer instance given to the pipeline cannot be
inspected directly.
In the following example, accessing the <a class="reference internal" href="generated/sklearn.decomposition.PCA.html#sklearn.decomposition.PCA" title="sklearn.decomposition.PCA"><code class="xref py py-class docutils literal notranslate"><span class="pre">PCA</span></code></a>
instance <code class="docutils literal notranslate"><span class="pre">pca2</span></code> will raise an <code class="docutils literal notranslate"><span class="pre">AttributeError</span></code> since <code class="docutils literal notranslate"><span class="pre">pca2</span></code> will be an
unfitted transformer.
Instead, use the attribute <code class="docutils literal notranslate"><span class="pre">named_steps</span></code> to inspect estimators within
the pipeline:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">cachedir</span> <span class="o">=</span> <span class="n">mkdtemp</span><span class="p">()</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">pca2</span> <span class="o">=</span> <span class="n">PCA</span><span class="p">(</span><span class="n">n_components</span><span class="o">=</span><span class="mi">10</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">svm2</span> <span class="o">=</span> <span class="n">SVC</span><span class="p">()</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">cached_pipe</span> <span class="o">=</span> <span class="n">Pipeline</span><span class="p">([(</span><span class="s1">&#39;reduce_dim&#39;</span><span class="p">,</span> <span class="n">pca2</span><span class="p">),</span> <span class="p">(</span><span class="s1">&#39;clf&#39;</span><span class="p">,</span> <span class="n">svm2</span><span class="p">)],</span>
<span class="gp">... </span>                       <span class="n">memory</span><span class="o">=</span><span class="n">cachedir</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">cached_pipe</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X_digits</span><span class="p">,</span> <span class="n">y_digits</span><span class="p">)</span>
<span class="go">Pipeline(memory=...,</span>
<span class="go">         steps=[(&#39;reduce_dim&#39;, PCA(n_components=10)), (&#39;clf&#39;, SVC())])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">cached_pipe</span><span class="o">.</span><span class="n">named_steps</span><span class="p">[</span><span class="s1">&#39;reduce_dim&#39;</span><span class="p">]</span><span class="o">.</span><span class="n">components_</span><span class="o">.</span><span class="n">shape</span>
<span class="go">(10, 64)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="c1"># Remove the cache directory</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">rmtree</span><span class="p">(</span><span class="n">cachedir</span><span class="p">)</span>
</pre></div>
</div>
</div>
</details><p class="rubric">Examples</p>
<ul class="simple">
<li><p><a class="reference internal" href="../auto_examples/compose/plot_compare_reduction.html#sphx-glr-auto-examples-compose-plot-compare-reduction-py"><span class="std std-ref">Selecting dimensionality reduction with Pipeline and GridSearchCV</span></a></p></li>
</ul>
</section>
</section>
<section id="transforming-target-in-regression">
<span id="transformed-target-regressor"></span><h2><span class="section-number">7.1.2. </span>Transforming target in regression<a class="headerlink" href="#transforming-target-in-regression" title="Link to this heading">#</a></h2>
<p><a class="reference internal" href="generated/sklearn.compose.TransformedTargetRegressor.html#sklearn.compose.TransformedTargetRegressor" title="sklearn.compose.TransformedTargetRegressor"><code class="xref py py-class docutils literal notranslate"><span class="pre">TransformedTargetRegressor</span></code></a> transforms the
targets <code class="docutils literal notranslate"><span class="pre">y</span></code> before fitting a regression model. The predictions are mapped
back to the original space via an inverse transform. It takes as an argument
the regressor that will be used for prediction, and the transformer that will
be applied to the target variable:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">import</span><span class="w"> </span><span class="nn">numpy</span><span class="w"> </span><span class="k">as</span><span class="w"> </span><span class="nn">np</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.datasets</span><span class="w"> </span><span class="kn">import</span> <span class="n">fetch_california_housing</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.compose</span><span class="w"> </span><span class="kn">import</span> <span class="n">TransformedTargetRegressor</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.preprocessing</span><span class="w"> </span><span class="kn">import</span> <span class="n">QuantileTransformer</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.linear_model</span><span class="w"> </span><span class="kn">import</span> <span class="n">LinearRegression</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.model_selection</span><span class="w"> </span><span class="kn">import</span> <span class="n">train_test_split</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">X</span><span class="p">,</span> <span class="n">y</span> <span class="o">=</span> <span class="n">fetch_california_housing</span><span class="p">(</span><span class="n">return_X_y</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">X</span><span class="p">,</span> <span class="n">y</span> <span class="o">=</span> <span class="n">X</span><span class="p">[:</span><span class="mi">2000</span><span class="p">,</span> <span class="p">:],</span> <span class="n">y</span><span class="p">[:</span><span class="mi">2000</span><span class="p">]</span>  <span class="c1"># select a subset of data</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">transformer</span> <span class="o">=</span> <span class="n">QuantileTransformer</span><span class="p">(</span><span class="n">output_distribution</span><span class="o">=</span><span class="s1">&#39;normal&#39;</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">regressor</span> <span class="o">=</span> <span class="n">LinearRegression</span><span class="p">()</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">regr</span> <span class="o">=</span> <span class="n">TransformedTargetRegressor</span><span class="p">(</span><span class="n">regressor</span><span class="o">=</span><span class="n">regressor</span><span class="p">,</span>
<span class="gp">... </span>                                  <span class="n">transformer</span><span class="o">=</span><span class="n">transformer</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">X_train</span><span class="p">,</span> <span class="n">X_test</span><span class="p">,</span> <span class="n">y_train</span><span class="p">,</span> <span class="n">y_test</span> <span class="o">=</span> <span class="n">train_test_split</span><span class="p">(</span><span class="n">X</span><span class="p">,</span> <span class="n">y</span><span class="p">,</span> <span class="n">random_state</span><span class="o">=</span><span class="mi">0</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">regr</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X_train</span><span class="p">,</span> <span class="n">y_train</span><span class="p">)</span>
<span class="go">TransformedTargetRegressor(...)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="nb">print</span><span class="p">(</span><span class="s1">&#39;R2 score: </span><span class="si">{0:.2f}</span><span class="s1">&#39;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">regr</span><span class="o">.</span><span class="n">score</span><span class="p">(</span><span class="n">X_test</span><span class="p">,</span> <span class="n">y_test</span><span class="p">)))</span>
<span class="go">R2 score: 0.61</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">raw_target_regr</span> <span class="o">=</span> <span class="n">LinearRegression</span><span class="p">()</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X_train</span><span class="p">,</span> <span class="n">y_train</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="nb">print</span><span class="p">(</span><span class="s1">&#39;R2 score: </span><span class="si">{0:.2f}</span><span class="s1">&#39;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">raw_target_regr</span><span class="o">.</span><span class="n">score</span><span class="p">(</span><span class="n">X_test</span><span class="p">,</span> <span class="n">y_test</span><span class="p">)))</span>
<span class="go">R2 score: 0.59</span>
</pre></div>
</div>
<p>For simple transformations, instead of a Transformer object, a pair of
functions can be passed, defining the transformation and its inverse mapping:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="k">def</span><span class="w"> </span><span class="nf">func</span><span class="p">(</span><span class="n">x</span><span class="p">):</span>
<span class="gp">... </span>    <span class="k">return</span> <span class="n">np</span><span class="o">.</span><span class="n">log</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="k">def</span><span class="w"> </span><span class="nf">inverse_func</span><span class="p">(</span><span class="n">x</span><span class="p">):</span>
<span class="gp">... </span>    <span class="k">return</span> <span class="n">np</span><span class="o">.</span><span class="n">exp</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
</pre></div>
</div>
<p>Subsequently, the object is created as:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">regr</span> <span class="o">=</span> <span class="n">TransformedTargetRegressor</span><span class="p">(</span><span class="n">regressor</span><span class="o">=</span><span class="n">regressor</span><span class="p">,</span>
<span class="gp">... </span>                                  <span class="n">func</span><span class="o">=</span><span class="n">func</span><span class="p">,</span>
<span class="gp">... </span>                                  <span class="n">inverse_func</span><span class="o">=</span><span class="n">inverse_func</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">regr</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X_train</span><span class="p">,</span> <span class="n">y_train</span><span class="p">)</span>
<span class="go">TransformedTargetRegressor(...)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="nb">print</span><span class="p">(</span><span class="s1">&#39;R2 score: </span><span class="si">{0:.2f}</span><span class="s1">&#39;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">regr</span><span class="o">.</span><span class="n">score</span><span class="p">(</span><span class="n">X_test</span><span class="p">,</span> <span class="n">y_test</span><span class="p">)))</span>
<span class="go">R2 score: 0.51</span>
</pre></div>
</div>
<p>By default, the provided functions are checked at each fit to be the inverse of
each other. However, it is possible to bypass this checking by setting
<code class="docutils literal notranslate"><span class="pre">check_inverse</span></code> to <code class="docutils literal notranslate"><span class="pre">False</span></code>:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="k">def</span><span class="w"> </span><span class="nf">inverse_func</span><span class="p">(</span><span class="n">x</span><span class="p">):</span>
<span class="gp">... </span>    <span class="k">return</span> <span class="n">x</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">regr</span> <span class="o">=</span> <span class="n">TransformedTargetRegressor</span><span class="p">(</span><span class="n">regressor</span><span class="o">=</span><span class="n">regressor</span><span class="p">,</span>
<span class="gp">... </span>                                  <span class="n">func</span><span class="o">=</span><span class="n">func</span><span class="p">,</span>
<span class="gp">... </span>                                  <span class="n">inverse_func</span><span class="o">=</span><span class="n">inverse_func</span><span class="p">,</span>
<span class="gp">... </span>                                  <span class="n">check_inverse</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">regr</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X_train</span><span class="p">,</span> <span class="n">y_train</span><span class="p">)</span>
<span class="go">TransformedTargetRegressor(...)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="nb">print</span><span class="p">(</span><span class="s1">&#39;R2 score: </span><span class="si">{0:.2f}</span><span class="s1">&#39;</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="n">regr</span><span class="o">.</span><span class="n">score</span><span class="p">(</span><span class="n">X_test</span><span class="p">,</span> <span class="n">y_test</span><span class="p">)))</span>
<span class="go">R2 score: -1.57</span>
</pre></div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The transformation can be triggered by setting either <code class="docutils literal notranslate"><span class="pre">transformer</span></code> or the
pair of functions <code class="docutils literal notranslate"><span class="pre">func</span></code> and <code class="docutils literal notranslate"><span class="pre">inverse_func</span></code>. However, setting both
options will raise an error.</p>
</div>
<p class="rubric">Examples</p>
<ul class="simple">
<li><p><a class="reference internal" href="../auto_examples/compose/plot_transformed_target.html#sphx-glr-auto-examples-compose-plot-transformed-target-py"><span class="std std-ref">Effect of transforming the targets in regression model</span></a></p></li>
</ul>
</section>
<section id="featureunion-composite-feature-spaces">
<span id="feature-union"></span><h2><span class="section-number">7.1.3. </span>FeatureUnion: composite feature spaces<a class="headerlink" href="#featureunion-composite-feature-spaces" title="Link to this heading">#</a></h2>
<p><a class="reference internal" href="generated/sklearn.pipeline.FeatureUnion.html#sklearn.pipeline.FeatureUnion" title="sklearn.pipeline.FeatureUnion"><code class="xref py py-class docutils literal notranslate"><span class="pre">FeatureUnion</span></code></a> combines several transformer objects into a new
transformer that combines their output. A <a class="reference internal" href="generated/sklearn.pipeline.FeatureUnion.html#sklearn.pipeline.FeatureUnion" title="sklearn.pipeline.FeatureUnion"><code class="xref py py-class docutils literal notranslate"><span class="pre">FeatureUnion</span></code></a> takes
a list of transformer objects. During fitting, each of these
is fit to the data independently. The transformers are applied in parallel,
and the feature matrices they output are concatenated side-by-side into a
larger matrix.</p>
<p>When you want to apply different transformations to each field of the data,
see the related class <a class="reference internal" href="generated/sklearn.compose.ColumnTransformer.html#sklearn.compose.ColumnTransformer" title="sklearn.compose.ColumnTransformer"><code class="xref py py-class docutils literal notranslate"><span class="pre">ColumnTransformer</span></code></a>
(see <a class="reference internal" href="#column-transformer"><span class="std std-ref">user guide</span></a>).</p>
<p><a class="reference internal" href="generated/sklearn.pipeline.FeatureUnion.html#sklearn.pipeline.FeatureUnion" title="sklearn.pipeline.FeatureUnion"><code class="xref py py-class docutils literal notranslate"><span class="pre">FeatureUnion</span></code></a> serves the same purposes as <a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> -
convenience and joint parameter estimation and validation.</p>
<p><a class="reference internal" href="generated/sklearn.pipeline.FeatureUnion.html#sklearn.pipeline.FeatureUnion" title="sklearn.pipeline.FeatureUnion"><code class="xref py py-class docutils literal notranslate"><span class="pre">FeatureUnion</span></code></a> and <a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> can be combined to
create complex models.</p>
<p>(A <a class="reference internal" href="generated/sklearn.pipeline.FeatureUnion.html#sklearn.pipeline.FeatureUnion" title="sklearn.pipeline.FeatureUnion"><code class="xref py py-class docutils literal notranslate"><span class="pre">FeatureUnion</span></code></a> has no way of checking whether two transformers
might produce identical features. It only produces a union when the
feature sets are disjoint, and making sure they are is the caller’s
responsibility.)</p>
<section id="id1">
<h3><span class="section-number">7.1.3.1. </span>Usage<a class="headerlink" href="#id1" title="Link to this heading">#</a></h3>
<p>A <a class="reference internal" href="generated/sklearn.pipeline.FeatureUnion.html#sklearn.pipeline.FeatureUnion" title="sklearn.pipeline.FeatureUnion"><code class="xref py py-class docutils literal notranslate"><span class="pre">FeatureUnion</span></code></a> is built using a list of <code class="docutils literal notranslate"><span class="pre">(key,</span> <span class="pre">value)</span></code> pairs,
where the <code class="docutils literal notranslate"><span class="pre">key</span></code> is the name you want to give to a given transformation
(an arbitrary string; it only serves as an identifier)
and <code class="docutils literal notranslate"><span class="pre">value</span></code> is an estimator object:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.pipeline</span><span class="w"> </span><span class="kn">import</span> <span class="n">FeatureUnion</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.decomposition</span><span class="w"> </span><span class="kn">import</span> <span class="n">PCA</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.decomposition</span><span class="w"> </span><span class="kn">import</span> <span class="n">KernelPCA</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">estimators</span> <span class="o">=</span> <span class="p">[(</span><span class="s1">&#39;linear_pca&#39;</span><span class="p">,</span> <span class="n">PCA</span><span class="p">()),</span> <span class="p">(</span><span class="s1">&#39;kernel_pca&#39;</span><span class="p">,</span> <span class="n">KernelPCA</span><span class="p">())]</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">combined</span> <span class="o">=</span> <span class="n">FeatureUnion</span><span class="p">(</span><span class="n">estimators</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">combined</span>
<span class="go">FeatureUnion(transformer_list=[(&#39;linear_pca&#39;, PCA()),</span>
<span class="go">                               (&#39;kernel_pca&#39;, KernelPCA())])</span>
</pre></div>
</div>
<p>Like pipelines, feature unions have a shorthand constructor called
<a class="reference internal" href="generated/sklearn.pipeline.make_union.html#sklearn.pipeline.make_union" title="sklearn.pipeline.make_union"><code class="xref py py-func docutils literal notranslate"><span class="pre">make_union</span></code></a> that does not require explicit naming of the components.</p>
<p>Like <code class="docutils literal notranslate"><span class="pre">Pipeline</span></code>, individual steps may be replaced using <code class="docutils literal notranslate"><span class="pre">set_params</span></code>,
and ignored by setting to <code class="docutils literal notranslate"><span class="pre">'drop'</span></code>:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">combined</span><span class="o">.</span><span class="n">set_params</span><span class="p">(</span><span class="n">kernel_pca</span><span class="o">=</span><span class="s1">&#39;drop&#39;</span><span class="p">)</span>
<span class="go">FeatureUnion(transformer_list=[(&#39;linear_pca&#39;, PCA()),</span>
<span class="go">                               (&#39;kernel_pca&#39;, &#39;drop&#39;)])</span>
</pre></div>
</div>
<p class="rubric">Examples</p>
<ul class="simple">
<li><p><a class="reference internal" href="../auto_examples/compose/plot_feature_union.html#sphx-glr-auto-examples-compose-plot-feature-union-py"><span class="std std-ref">Concatenating multiple feature extraction methods</span></a></p></li>
</ul>
</section>
</section>
<section id="columntransformer-for-heterogeneous-data">
<span id="column-transformer"></span><h2><span class="section-number">7.1.4. </span>ColumnTransformer for heterogeneous data<a class="headerlink" href="#columntransformer-for-heterogeneous-data" title="Link to this heading">#</a></h2>
<p>Many datasets contain features of different types, say text, floats, and dates,
where each type of feature requires separate preprocessing or feature
extraction steps.  Often it is easiest to preprocess data before applying
scikit-learn methods, for example using <a class="reference external" href="https://pandas.pydata.org/">pandas</a>.
Processing your data before passing it to scikit-learn might be problematic for
one of the following reasons:</p>
<ol class="arabic simple">
<li><p>Incorporating statistics from test data into the preprocessors makes
cross-validation scores unreliable (known as <em>data leakage</em>),
for example in the case of scalers or imputing missing values.</p></li>
<li><p>You may want to include the parameters of the preprocessors in a
<a class="reference internal" href="grid_search.html#grid-search"><span class="std std-ref">parameter search</span></a>.</p></li>
</ol>
<p>The <a class="reference internal" href="generated/sklearn.compose.ColumnTransformer.html#sklearn.compose.ColumnTransformer" title="sklearn.compose.ColumnTransformer"><code class="xref py py-class docutils literal notranslate"><span class="pre">ColumnTransformer</span></code></a> helps performing different
transformations for different columns of the data, within a
<a class="reference internal" href="generated/sklearn.pipeline.Pipeline.html#sklearn.pipeline.Pipeline" title="sklearn.pipeline.Pipeline"><code class="xref py py-class docutils literal notranslate"><span class="pre">Pipeline</span></code></a> that is safe from data leakage and that can
be parametrized. <a class="reference internal" href="generated/sklearn.compose.ColumnTransformer.html#sklearn.compose.ColumnTransformer" title="sklearn.compose.ColumnTransformer"><code class="xref py py-class docutils literal notranslate"><span class="pre">ColumnTransformer</span></code></a> works on
arrays, sparse matrices, and
<a class="reference external" href="https://pandas.pydata.org/pandas-docs/stable/">pandas DataFrames</a>.</p>
<p>To each column, a different transformation can be applied, such as
preprocessing or a specific feature extraction method:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">import</span><span class="w"> </span><span class="nn">pandas</span><span class="w"> </span><span class="k">as</span><span class="w"> </span><span class="nn">pd</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">X</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">(</span>
<span class="gp">... </span>    <span class="p">{</span><span class="s1">&#39;city&#39;</span><span class="p">:</span> <span class="p">[</span><span class="s1">&#39;London&#39;</span><span class="p">,</span> <span class="s1">&#39;London&#39;</span><span class="p">,</span> <span class="s1">&#39;Paris&#39;</span><span class="p">,</span> <span class="s1">&#39;Sallisaw&#39;</span><span class="p">],</span>
<span class="gp">... </span>     <span class="s1">&#39;title&#39;</span><span class="p">:</span> <span class="p">[</span><span class="s2">&quot;His Last Bow&quot;</span><span class="p">,</span> <span class="s2">&quot;How Watson Learned the Trick&quot;</span><span class="p">,</span>
<span class="gp">... </span>               <span class="s2">&quot;A Moveable Feast&quot;</span><span class="p">,</span> <span class="s2">&quot;The Grapes of Wrath&quot;</span><span class="p">],</span>
<span class="gp">... </span>     <span class="s1">&#39;expert_rating&#39;</span><span class="p">:</span> <span class="p">[</span><span class="mi">5</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">5</span><span class="p">],</span>
<span class="gp">... </span>     <span class="s1">&#39;user_rating&#39;</span><span class="p">:</span> <span class="p">[</span><span class="mi">4</span><span class="p">,</span> <span class="mi">5</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">3</span><span class="p">]})</span>
</pre></div>
</div>
<p>For this data, we might want to encode the <code class="docutils literal notranslate"><span class="pre">'city'</span></code> column as a categorical
variable using <a class="reference internal" href="generated/sklearn.preprocessing.OneHotEncoder.html#sklearn.preprocessing.OneHotEncoder" title="sklearn.preprocessing.OneHotEncoder"><code class="xref py py-class docutils literal notranslate"><span class="pre">OneHotEncoder</span></code></a> but apply a
<a class="reference internal" href="generated/sklearn.feature_extraction.text.CountVectorizer.html#sklearn.feature_extraction.text.CountVectorizer" title="sklearn.feature_extraction.text.CountVectorizer"><code class="xref py py-class docutils literal notranslate"><span class="pre">CountVectorizer</span></code></a> to the <code class="docutils literal notranslate"><span class="pre">'title'</span></code> column.
As we might use multiple feature extraction methods on the same column, we give
each transformer a unique name, say <code class="docutils literal notranslate"><span class="pre">'city_category'</span></code> and <code class="docutils literal notranslate"><span class="pre">'title_bow'</span></code>.
By default, the remaining rating columns are ignored (<code class="docutils literal notranslate"><span class="pre">remainder='drop'</span></code>):</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.compose</span><span class="w"> </span><span class="kn">import</span> <span class="n">ColumnTransformer</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.feature_extraction.text</span><span class="w"> </span><span class="kn">import</span> <span class="n">CountVectorizer</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.preprocessing</span><span class="w"> </span><span class="kn">import</span> <span class="n">OneHotEncoder</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span> <span class="o">=</span> <span class="n">ColumnTransformer</span><span class="p">(</span>
<span class="gp">... </span>    <span class="p">[(</span><span class="s1">&#39;categories&#39;</span><span class="p">,</span> <span class="n">OneHotEncoder</span><span class="p">(</span><span class="n">dtype</span><span class="o">=</span><span class="s1">&#39;int&#39;</span><span class="p">),</span> <span class="p">[</span><span class="s1">&#39;city&#39;</span><span class="p">]),</span>
<span class="gp">... </span>     <span class="p">(</span><span class="s1">&#39;title_bow&#39;</span><span class="p">,</span> <span class="n">CountVectorizer</span><span class="p">(),</span> <span class="s1">&#39;title&#39;</span><span class="p">)],</span>
<span class="gp">... </span>    <span class="n">remainder</span><span class="o">=</span><span class="s1">&#39;drop&#39;</span><span class="p">,</span> <span class="n">verbose_feature_names_out</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span>

<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X</span><span class="p">)</span>
<span class="go">ColumnTransformer(transformers=[(&#39;categories&#39;, OneHotEncoder(dtype=&#39;int&#39;),</span>
<span class="go">                                 [&#39;city&#39;]),</span>
<span class="go">                                (&#39;title_bow&#39;, CountVectorizer(), &#39;title&#39;)],</span>
<span class="go">                  verbose_feature_names_out=False)</span>

<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span><span class="o">.</span><span class="n">get_feature_names_out</span><span class="p">()</span>
<span class="go">array([&#39;city_London&#39;, &#39;city_Paris&#39;, &#39;city_Sallisaw&#39;, &#39;bow&#39;, &#39;feast&#39;,</span>
<span class="go">&#39;grapes&#39;, &#39;his&#39;, &#39;how&#39;, &#39;last&#39;, &#39;learned&#39;, &#39;moveable&#39;, &#39;of&#39;, &#39;the&#39;,</span>
<span class="go"> &#39;trick&#39;, &#39;watson&#39;, &#39;wrath&#39;], ...)</span>

<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span><span class="o">.</span><span class="n">transform</span><span class="p">(</span><span class="n">X</span><span class="p">)</span><span class="o">.</span><span class="n">toarray</span><span class="p">()</span>
<span class="go">array([[1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0],</span>
<span class="go">       [1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0],</span>
<span class="go">       [0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0],</span>
<span class="go">       [0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1]]...)</span>
</pre></div>
</div>
<p>In the above example, the
<a class="reference internal" href="generated/sklearn.feature_extraction.text.CountVectorizer.html#sklearn.feature_extraction.text.CountVectorizer" title="sklearn.feature_extraction.text.CountVectorizer"><code class="xref py py-class docutils literal notranslate"><span class="pre">CountVectorizer</span></code></a> expects a 1D array as
input and therefore the columns were specified as a string (<code class="docutils literal notranslate"><span class="pre">'title'</span></code>).
However, <a class="reference internal" href="generated/sklearn.preprocessing.OneHotEncoder.html#sklearn.preprocessing.OneHotEncoder" title="sklearn.preprocessing.OneHotEncoder"><code class="xref py py-class docutils literal notranslate"><span class="pre">OneHotEncoder</span></code></a>
as most of other transformers expects 2D data, therefore in that case you need
to specify the column as a list of strings (<code class="docutils literal notranslate"><span class="pre">['city']</span></code>).</p>
<p>Apart from a scalar or a single item list, the column selection can be specified
as a list of multiple items, an integer array, a slice, a boolean mask, or
with a <a class="reference internal" href="generated/sklearn.compose.make_column_selector.html#sklearn.compose.make_column_selector" title="sklearn.compose.make_column_selector"><code class="xref py py-func docutils literal notranslate"><span class="pre">make_column_selector</span></code></a>. The
<a class="reference internal" href="generated/sklearn.compose.make_column_selector.html#sklearn.compose.make_column_selector" title="sklearn.compose.make_column_selector"><code class="xref py py-func docutils literal notranslate"><span class="pre">make_column_selector</span></code></a> is used to select columns based
on data type or column name:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.preprocessing</span><span class="w"> </span><span class="kn">import</span> <span class="n">StandardScaler</span>
<span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.compose</span><span class="w"> </span><span class="kn">import</span> <span class="n">make_column_selector</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">ct</span> <span class="o">=</span> <span class="n">ColumnTransformer</span><span class="p">([</span>
<span class="gp">... </span>      <span class="p">(</span><span class="s1">&#39;scale&#39;</span><span class="p">,</span> <span class="n">StandardScaler</span><span class="p">(),</span>
<span class="gp">... </span>      <span class="n">make_column_selector</span><span class="p">(</span><span class="n">dtype_include</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">number</span><span class="p">)),</span>
<span class="gp">... </span>      <span class="p">(</span><span class="s1">&#39;onehot&#39;</span><span class="p">,</span>
<span class="gp">... </span>      <span class="n">OneHotEncoder</span><span class="p">(),</span>
<span class="gp">... </span>      <span class="n">make_column_selector</span><span class="p">(</span><span class="n">pattern</span><span class="o">=</span><span class="s1">&#39;city&#39;</span><span class="p">,</span> <span class="n">dtype_include</span><span class="o">=</span><span class="nb">object</span><span class="p">))])</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">ct</span><span class="o">.</span><span class="n">fit_transform</span><span class="p">(</span><span class="n">X</span><span class="p">)</span>
<span class="go">array([[ 0.904...,  0.      ,  1. ,  0. ,  0. ],</span>
<span class="go">       [-1.507...,  1.414...,  1. ,  0. ,  0. ],</span>
<span class="go">       [-0.301...,  0.      ,  0. ,  1. ,  0. ],</span>
<span class="go">       [ 0.904..., -1.414...,  0. ,  0. ,  1. ]])</span>
</pre></div>
</div>
<p>Strings can reference columns if the input is a DataFrame, integers are always
interpreted as the positional columns.</p>
<p>We can keep the remaining rating columns by setting
<code class="docutils literal notranslate"><span class="pre">remainder='passthrough'</span></code>. The values are appended to the end of the
transformation:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span> <span class="o">=</span> <span class="n">ColumnTransformer</span><span class="p">(</span>
<span class="gp">... </span>    <span class="p">[(</span><span class="s1">&#39;city_category&#39;</span><span class="p">,</span> <span class="n">OneHotEncoder</span><span class="p">(</span><span class="n">dtype</span><span class="o">=</span><span class="s1">&#39;int&#39;</span><span class="p">),[</span><span class="s1">&#39;city&#39;</span><span class="p">]),</span>
<span class="gp">... </span>     <span class="p">(</span><span class="s1">&#39;title_bow&#39;</span><span class="p">,</span> <span class="n">CountVectorizer</span><span class="p">(),</span> <span class="s1">&#39;title&#39;</span><span class="p">)],</span>
<span class="gp">... </span>    <span class="n">remainder</span><span class="o">=</span><span class="s1">&#39;passthrough&#39;</span><span class="p">)</span>

<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span><span class="o">.</span><span class="n">fit_transform</span><span class="p">(</span><span class="n">X</span><span class="p">)</span>
<span class="go">array([[1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 5, 4],</span>
<span class="go">       [1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 3, 5],</span>
<span class="go">       [0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 4, 4],</span>
<span class="go">       [0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 5, 3]]...)</span>
</pre></div>
</div>
<p>The <code class="docutils literal notranslate"><span class="pre">remainder</span></code> parameter can be set to an estimator to transform the
remaining rating columns. The transformed values are appended to the end of
the transformation:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.preprocessing</span><span class="w"> </span><span class="kn">import</span> <span class="n">MinMaxScaler</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span> <span class="o">=</span> <span class="n">ColumnTransformer</span><span class="p">(</span>
<span class="gp">... </span>    <span class="p">[(</span><span class="s1">&#39;city_category&#39;</span><span class="p">,</span> <span class="n">OneHotEncoder</span><span class="p">(),</span> <span class="p">[</span><span class="s1">&#39;city&#39;</span><span class="p">]),</span>
<span class="gp">... </span>     <span class="p">(</span><span class="s1">&#39;title_bow&#39;</span><span class="p">,</span> <span class="n">CountVectorizer</span><span class="p">(),</span> <span class="s1">&#39;title&#39;</span><span class="p">)],</span>
<span class="gp">... </span>    <span class="n">remainder</span><span class="o">=</span><span class="n">MinMaxScaler</span><span class="p">())</span>

<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span><span class="o">.</span><span class="n">fit_transform</span><span class="p">(</span><span class="n">X</span><span class="p">)[:,</span> <span class="o">-</span><span class="mi">2</span><span class="p">:]</span>
<span class="go">array([[1. , 0.5],</span>
<span class="go">       [0. , 1. ],</span>
<span class="go">       [0.5, 0.5],</span>
<span class="go">       [1. , 0. ]])</span>
</pre></div>
</div>
<p id="make-column-transformer">The <a class="reference internal" href="generated/sklearn.compose.make_column_transformer.html#sklearn.compose.make_column_transformer" title="sklearn.compose.make_column_transformer"><code class="xref py py-func docutils literal notranslate"><span class="pre">make_column_transformer</span></code></a> function is available
to more easily create a <a class="reference internal" href="generated/sklearn.compose.ColumnTransformer.html#sklearn.compose.ColumnTransformer" title="sklearn.compose.ColumnTransformer"><code class="xref py py-class docutils literal notranslate"><span class="pre">ColumnTransformer</span></code></a> object.
Specifically, the names will be given automatically. The equivalent for the
above example would be:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.compose</span><span class="w"> </span><span class="kn">import</span> <span class="n">make_column_transformer</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span> <span class="o">=</span> <span class="n">make_column_transformer</span><span class="p">(</span>
<span class="gp">... </span>    <span class="p">(</span><span class="n">OneHotEncoder</span><span class="p">(),</span> <span class="p">[</span><span class="s1">&#39;city&#39;</span><span class="p">]),</span>
<span class="gp">... </span>    <span class="p">(</span><span class="n">CountVectorizer</span><span class="p">(),</span> <span class="s1">&#39;title&#39;</span><span class="p">),</span>
<span class="gp">... </span>    <span class="n">remainder</span><span class="o">=</span><span class="n">MinMaxScaler</span><span class="p">())</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span>
<span class="go">ColumnTransformer(remainder=MinMaxScaler(),</span>
<span class="go">                  transformers=[(&#39;onehotencoder&#39;, OneHotEncoder(), [&#39;city&#39;]),</span>
<span class="go">                                (&#39;countvectorizer&#39;, CountVectorizer(),</span>
<span class="go">                                 &#39;title&#39;)])</span>
</pre></div>
</div>
<p>If <a class="reference internal" href="generated/sklearn.compose.ColumnTransformer.html#sklearn.compose.ColumnTransformer" title="sklearn.compose.ColumnTransformer"><code class="xref py py-class docutils literal notranslate"><span class="pre">ColumnTransformer</span></code></a> is fitted with a dataframe
and the dataframe only has string column names, then transforming a dataframe
will use the column names to select the columns:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">ct</span> <span class="o">=</span> <span class="n">ColumnTransformer</span><span class="p">(</span>
<span class="gp">... </span>         <span class="p">[(</span><span class="s2">&quot;scale&quot;</span><span class="p">,</span> <span class="n">StandardScaler</span><span class="p">(),</span> <span class="p">[</span><span class="s2">&quot;expert_rating&quot;</span><span class="p">])])</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X</span><span class="p">)</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">X_new</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s2">&quot;expert_rating&quot;</span><span class="p">:</span> <span class="p">[</span><span class="mi">5</span><span class="p">,</span> <span class="mi">6</span><span class="p">,</span> <span class="mi">1</span><span class="p">],</span>
<span class="gp">... </span>                      <span class="s2">&quot;ignored_new_col&quot;</span><span class="p">:</span> <span class="p">[</span><span class="mf">1.2</span><span class="p">,</span> <span class="mf">0.3</span><span class="p">,</span> <span class="o">-</span><span class="mf">0.1</span><span class="p">]})</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">ct</span><span class="o">.</span><span class="n">transform</span><span class="p">(</span><span class="n">X_new</span><span class="p">)</span>
<span class="go">array([[ 0.9...],</span>
<span class="go">       [ 2.1...],</span>
<span class="go">       [-3.9...]])</span>
</pre></div>
</div>
</section>
<section id="visualizing-composite-estimators">
<span id="id2"></span><h2><span class="section-number">7.1.5. </span>Visualizing Composite Estimators<a class="headerlink" href="#visualizing-composite-estimators" title="Link to this heading">#</a></h2>
<p>Estimators are displayed with an HTML representation when shown in a
jupyter notebook. This is useful to diagnose or visualize a Pipeline with
many estimators. This visualization is activated by default:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span>  
</pre></div>
</div>
<p>It can be deactivated by setting the <code class="docutils literal notranslate"><span class="pre">display</span></code> option in <a class="reference internal" href="generated/sklearn.set_config.html#sklearn.set_config" title="sklearn.set_config"><code class="xref py py-func docutils literal notranslate"><span class="pre">set_config</span></code></a>
to ‘text’:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn</span><span class="w"> </span><span class="kn">import</span> <span class="n">set_config</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">set_config</span><span class="p">(</span><span class="n">display</span><span class="o">=</span><span class="s1">&#39;text&#39;</span><span class="p">)</span>  
<span class="gp">&gt;&gt;&gt; </span><span class="c1"># displays text representation in a jupyter context</span>
<span class="gp">&gt;&gt;&gt; </span><span class="n">column_trans</span>  
</pre></div>
</div>
<p>An example of the HTML output can be seen in the
<strong>HTML representation of Pipeline</strong> section of
<a class="reference internal" href="../auto_examples/compose/plot_column_transformer_mixed_types.html#sphx-glr-auto-examples-compose-plot-column-transformer-mixed-types-py"><span class="std std-ref">Column Transformer with Mixed Types</span></a>.
As an alternative, the HTML can be written to a file using
<a class="reference internal" href="generated/sklearn.utils.estimator_html_repr.html#sklearn.utils.estimator_html_repr" title="sklearn.utils.estimator_html_repr"><code class="xref py py-func docutils literal notranslate"><span class="pre">estimator_html_repr</span></code></a>:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">&gt;&gt;&gt; </span><span class="kn">from</span><span class="w"> </span><span class="nn">sklearn.utils</span><span class="w"> </span><span class="kn">import</span> <span class="n">estimator_html_repr</span>
<span class="gp">&gt;&gt;&gt; </span><span class="k">with</span> <span class="nb">open</span><span class="p">(</span><span class="s1">&#39;my_estimator.html&#39;</span><span class="p">,</span> <span class="s1">&#39;w&#39;</span><span class="p">)</span> <span class="k">as</span> <span class="n">f</span><span class="p">:</span>  
<span class="gp">... </span>    <span class="n">f</span><span class="o">.</span><span class="n">write</span><span class="p">(</span><span class="n">estimator_html_repr</span><span class="p">(</span><span class="n">clf</span><span class="p">))</span>
</pre></div>
</div>
<p class="rubric">Examples</p>
<ul class="simple">
<li><p><a class="reference internal" href="../auto_examples/compose/plot_column_transformer.html#sphx-glr-auto-examples-compose-plot-column-transformer-py"><span class="std std-ref">Column Transformer with Heterogeneous Data Sources</span></a></p></li>
<li><p><a class="reference internal" href="../auto_examples/compose/plot_column_transformer_mixed_types.html#sphx-glr-auto-examples-compose-plot-column-transformer-mixed-types-py"><span class="std std-ref">Column Transformer with Mixed Types</span></a></p></li>
</ul>
</section>
</section>


                </article>
              
              
                <footer class="bd-footer-article">
                  <div class="footer-article-items footer-article__inner">
  
    <div class="footer-article-item">
<div class="prev-next-area">
    <a class="left-prev"
       href="../data_transforms.html"
       title="previous page">
      <i class="fa-solid fa-angle-left"></i>
      <div class="prev-next-info">
        <p class="prev-next-subtitle">previous</p>
        <p class="prev-next-title"><span class="section-number">7. </span>Dataset transformations</p>
      </div>
    </a>
    <a class="right-next"
       href="feature_extraction.html"
       title="next page">
      <div class="prev-next-info">
        <p class="prev-next-subtitle">next</p>
        <p class="prev-next-title"><span class="section-number">7.2. </span>Feature extraction</p>
      </div>
      <i class="fa-solid fa-angle-right"></i>
    </a>
</div></div>
  
</div>
                </footer>
              
              
            </div>
            
            
                <dialog id="pst-secondary-sidebar-modal"></dialog>
                <div id="pst-secondary-sidebar" class="bd-sidebar-secondary bd-toc"><div class="sidebar-secondary-items sidebar-secondary__inner">


  <div class="sidebar-secondary-item">
<div
    id="pst-page-navigation-heading-2"
    class="page-toc tocsection onthispage">
    <i class="fa-solid fa-list"></i> On this page
  </div>
  <nav class="bd-toc-nav page-toc" aria-labelledby="pst-page-navigation-heading-2">
    <ul class="visible nav section-nav flex-column">
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#pipeline-chaining-estimators">7.1.1. Pipeline: chaining estimators</a><ul class="nav section-nav flex-column">
<li class="toc-h3 nav-item toc-entry"><a class="reference internal nav-link" href="#usage">7.1.1.1. Usage</a><ul class="nav section-nav flex-column">
<li class="toc-h4 nav-item toc-entry"><a class="reference internal nav-link" href="#build-a-pipeline">7.1.1.1.1. Build a pipeline</a></li>
<li class="toc-h4 nav-item toc-entry"><a class="reference internal nav-link" href="#access-pipeline-steps">7.1.1.1.2. Access pipeline steps</a></li>
<li class="toc-h4 nav-item toc-entry"><a class="reference internal nav-link" href="#tracking-feature-names-in-a-pipeline">7.1.1.1.3. Tracking feature names in a pipeline</a></li>
<li class="toc-h4 nav-item toc-entry"><a class="reference internal nav-link" href="#access-to-nested-parameters">7.1.1.1.4. Access to nested parameters</a></li>
</ul>
</li>
<li class="toc-h3 nav-item toc-entry"><a class="reference internal nav-link" href="#caching-transformers-avoid-repeated-computation">7.1.1.2. Caching transformers: avoid repeated computation</a></li>
</ul>
</li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#transforming-target-in-regression">7.1.2. Transforming target in regression</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#featureunion-composite-feature-spaces">7.1.3. FeatureUnion: composite feature spaces</a><ul class="nav section-nav flex-column">
<li class="toc-h3 nav-item toc-entry"><a class="reference internal nav-link" href="#id1">7.1.3.1. Usage</a></li>
</ul>
</li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#columntransformer-for-heterogeneous-data">7.1.4. ColumnTransformer for heterogeneous data</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#visualizing-composite-estimators">7.1.5. Visualizing Composite Estimators</a></li>
</ul>
  </nav></div>

  <div class="sidebar-secondary-item">
  <div role="note" aria-label="source link">
    <h3>This Page</h3>
    <ul class="this-page-menu">
      <li><a href="../_sources/modules/compose.rst.txt"
            rel="nofollow">Show Source</a></li>
    </ul>
   </div></div>

</div></div>
              
            
          </div>
          <footer class="bd-footer-content">
            
          </footer>
        
      </main>
    </div>
  </div>
  
  <!-- Scripts loaded after <body> so the DOM is not blocked -->
  <script defer src="../_static/scripts/bootstrap.js?digest=8878045cc6db502f8baf"></script>
<script defer src="../_static/scripts/pydata-sphinx-theme.js?digest=8878045cc6db502f8baf"></script>

  <footer class="bd-footer">
<div class="bd-footer__inner bd-page-width">
  
    <div class="footer-items__start">
      
        <div class="footer-item">

  <p class="copyright">
    
      © Copyright 2007 - 2025, scikit-learn developers (BSD License).
      <br/>
    
  </p>
</div>
      
    </div>
  
  
</div>

  </footer>
  </body>
</html>