{"id":233851,"date":"2024-03-21T09:27:07","date_gmt":"2024-03-21T09:27:07","guid":{"rendered":"https:\/\/datascientest.com\/?p=233851"},"modified":"2024-03-21T09:27:08","modified_gmt":"2024-03-21T09:27:08","slug":"category-encoders-tout-savoir","status":"publish","type":"post","link":"https:\/\/datascientest.com\/category-encoders-tout-savoir","title":{"rendered":"Category encoders : le couteau suisse de l\u2019encodage"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"233851\" class=\"elementor elementor-233851\" data-elementor-post-type=\"post\">\n\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-80ee517 e-flex e-con-boxed e-con\" data-id=\"80ee517\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-a9e8f35 elementor-widget elementor-widget-heading\" data-id=\"a9e8f35\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-heading-title{padding:0;margin:0;line-height:1}.elementor-widget-heading .elementor-heading-title[class*=elementor-size-]>a{color:inherit;font-size:inherit;line-height:inherit}.elementor-widget-heading .elementor-heading-title.elementor-size-small{font-size:15px}.elementor-widget-heading .elementor-heading-title.elementor-size-medium{font-size:19px}.elementor-widget-heading .elementor-heading-title.elementor-size-large{font-size:29px}.elementor-widget-heading .elementor-heading-title.elementor-size-xl{font-size:39px}.elementor-widget-heading .elementor-heading-title.elementor-size-xxl{font-size:59px}<\/style><h2 class=\"elementor-heading-title elementor-size-default\">Dans un jeu de donn\u00e9es, deux types de variables pr\u00e9dominent : les variables quantitatives, qui mesurent des quantit\u00e9s (une taille, un prix\u2026), et les variables qualitatives ou cat\u00e9gorielles,  qui d\u00e9terminent plut\u00f4t des cat\u00e9gories ou des modalit\u00e9s (un sexe, une couleur\u2026). Or comme le Machine Learning s\u2019appuie sur diverses op\u00e9rations math\u00e9matiques, il est crucial de convertir, ou plut\u00f4t d\u2019encoder, nos variables cat\u00e9gorielles en valeurs num\u00e9riques afin qu\u2019elles puissent \u00eatre trait\u00e9es par nos mod\u00e8les. Cet encodage intervient lors de la phase de pr\u00e9traitement des donn\u00e9es, aussi appel\u00e9e \"Preprocessing\".<\/h2>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-13ee334 elementor-widget elementor-widget-text-editor\" data-id=\"13ee334\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-text-editor.elementor-drop-cap-view-stacked .elementor-drop-cap{background-color:#69727d;color:#fff}.elementor-widget-text-editor.elementor-drop-cap-view-framed .elementor-drop-cap{color:#69727d;border:3px solid;background-color:transparent}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap{margin-top:8px}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap-letter{width:1em;height:1em}.elementor-widget-text-editor .elementor-drop-cap{float:left;text-align:center;line-height:1;font-size:50px}.elementor-widget-text-editor .elementor-drop-cap-letter{display:inline-block}<\/style>\t\t\t\t<p><span style=\"font-weight: 400;\">La librairie Python <\/span><i><span style=\"font-weight: 400;\">category encoders<\/span><\/i><span style=\"font-weight: 400;\">, comme son nom l\u2019indique, offre alors une <\/span><b>vari\u00e9t\u00e9 de techniques d\u2019encodage<\/b><span style=\"font-weight: 400;\"> pour nos variables. Comme toute librairie, elle s\u2019installe avec la commande <\/span><i><span style=\"font-weight: 400;\">pip install category_encoders <\/span><\/i><span style=\"font-weight: 400;\">et s\u2019importe avec <\/span><i><span style=\"font-weight: 400;\">import category_encoders as ce<\/span><\/i><span style=\"font-weight: 400;\">.<\/span><\/p><p><span style=\"font-weight: 400;\">\u00c0 travers cet article, voyons ensemble quelques-unes des nombreuses techniques d\u2019encodage propos\u00e9es par cette librairie. Nous consid\u00e9rons \u00e9galement, \u00e0 titre d\u2019exemple, le <a href=\"https:\/\/datascientest.com\/quest-ce-quun-dataframe\">DataFrame<\/a> suivant. Ce DataFrame contient cinq colonnes : trois variables cat\u00e9gorielles (nom, sexe, et taille), et deux quantitatives (\u00e2ge et situation). La variable cible que nous souhaitons pr\u00e9dire est la situation professionnelle, inscrite dans la colonne \u00e9ponyme : elle prend la valeur 1 si la personne \u00e9tudi\u00e9e est employ\u00e9e, 0 si elle est au ch\u00f4mage.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e443274 elementor-widget elementor-widget-image\" data-id=\"e443274\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-image{text-align:center}.elementor-widget-image a{display:inline-block}.elementor-widget-image a img[src$=\".svg\"]{width:48px}.elementor-widget-image img{vertical-align:middle;display:inline-block}<\/style>\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"269\" height=\"100\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image3.png\" class=\"attachment-full size-full wp-image-233855\" alt=\"\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-a3070c2 e-flex e-con-boxed e-con\" data-id=\"a3070c2\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-16a0342 elementor-widget elementor-widget-heading\" data-id=\"16a0342\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">L\u2019encodage One-Hot<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1feb6a2 elementor-widget elementor-widget-text-editor\" data-id=\"1feb6a2\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Commen\u00e7ons par la m\u00e9thode d&rsquo;encodage qui est sans doute la plus courante. Sur chaque colonne \u00e0 encoder, <\/span><b>l\u2019encodage one-hot (ou <\/b><b><i>one-hot encoding<\/i><\/b><b>) la scinde en X colonnes<\/b><span style=\"font-weight: 400;\">, X \u00e9tant le <\/span><b>nombre de valeurs distinctes<\/b><span style=\"font-weight: 400;\">. Appliquons un one-hot encoding sur la colonne \u201csexe\u201d de notre DataFrame :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-08a8f8d elementor-widget elementor-widget-image\" data-id=\"08a8f8d\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"338\" height=\"108\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image1.png\" class=\"attachment-full size-full wp-image-233856\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image1.png 338w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image1-300x96.png 300w\" sizes=\"(max-width: 338px) 100vw, 338px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-68375f3 elementor-widget elementor-widget-text-editor\" data-id=\"68375f3\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Comme nous pouvons le remarquer, <\/span><b>la colonne \u00ab\u00a0sexe\u00a0\u00bb s\u2019est scind\u00e9e en deux colonnes<\/b><span style=\"font-weight: 400;\"> Femme et Homme. De m\u00eame les valeurs contenues dans cette colonne sont maintenant <\/span><b>binaires <\/b><span style=\"font-weight: 400;\">: 1 si le sexe de la colonne correspond \u00e0 la personne \u00e9tudi\u00e9e, 0 sinon.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-c2b3f97 elementor-widget elementor-widget-image\" data-id=\"c2b3f97\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"1000\" height=\"563\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders1.jpg\" class=\"attachment-full size-full wp-image-233867\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders1.jpg 1000w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders1-300x169.jpg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders1-768x432.jpg 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7b4e8c0 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"7b4e8c0\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formation-data-analyst\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Tout apprendre de l'encodage de donn\u00e9es<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-a505acd e-flex e-con-boxed e-con\" data-id=\"a505acd\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-037db59 elementor-widget elementor-widget-heading\" data-id=\"037db59\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">L\u2019encodage binaire<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9ac18be elementor-widget elementor-widget-text-editor\" data-id=\"9ac18be\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><b>L\u2019encodage binaire <\/b><span style=\"font-weight: 400;\">fonctionne quasiment comme le one-hot : il scinde la colonne \u00e0 encoder en X colonnes distinctes dont les valeurs sont binaires. La diff\u00e9rence ici sera au niveau des valeurs : si l\u2019encodage one-hot applique 0 ou 1 aux valeurs sans ordre particulier, <\/span><b>l\u2019encodage binaire lui comptera ces valeurs en base 2<\/b><span style=\"font-weight: 400;\">. Voyons cela avec le m\u00eame DataFrame :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-a26b57d elementor-widget elementor-widget-image\" data-id=\"a26b57d\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"335\" height=\"107\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image4.png\" class=\"attachment-full size-full wp-image-233857\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image4.png 335w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image4-300x96.png 300w\" sizes=\"(max-width: 335px) 100vw, 335px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1013fd8 elementor-widget elementor-widget-text-editor\" data-id=\"1013fd8\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Si on obtient comme tout \u00e0 l\u2019heure deux colonnes distinctes au niveau du sexe, <\/span><b>les valeurs, quant \u00e0 elles, n\u2019ont pas tout \u00e0 fait le m\u00eame sens<\/b><span style=\"font-weight: 400;\">. L\u2019encodage binaire a en effet d\u00e9nombr\u00e9 la valeur \u00ab\u00a0F\u00a0\u00bb, qui appara\u00eet en 1er, comme le chiffre 1 (<\/span><i><span style=\"font-weight: 400;\">01<\/span><\/i><span style=\"font-weight: 400;\"> en base 2), et la valeur \u00ab\u00a0M\u00a0\u00bb comme le chiffre 2 (<\/span><i><span style=\"font-weight: 400;\">10 <\/span><\/i><span style=\"font-weight: 400;\">en base 2).<\/span><\/p><p><span style=\"font-weight: 400;\">En r\u00e9alit\u00e9, les encodages one-hot et binaire peuvent \u00eatre consid\u00e9r\u00e9s comme un sous-ensemble d\u2019un autre type d\u2019encodage pris en charge par <\/span><i><span style=\"font-weight: 400;\">category_encoders<\/span><\/i><span style=\"font-weight: 400;\"> : <\/span><b>l\u2019encodage en base-N<\/b><span style=\"font-weight: 400;\">. Cet encodage permet de convertir nos variables cat\u00e9gorielles, sous forme de tableau, en leur <\/span><b>\u00e9quivalent en base-N<\/b><span style=\"font-weight: 400;\">. Ainsi, si l\u2019encodage binaire correspond \u00e0 un encodage base-2, le one-hot peut \u00eatre consid\u00e9r\u00e9 comme un encodage en base-1.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-88d9c43 e-flex e-con-boxed e-con\" data-id=\"88d9c43\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-216722b elementor-widget elementor-widget-heading\" data-id=\"216722b\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">L\u2019encodage ordinal<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-dc4ef27 elementor-widget elementor-widget-text-editor\" data-id=\"dc4ef27\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Il existe une sous-cat\u00e9gorisation au sein des variables cat\u00e9gorielles dont nous n\u2019avons pas parl\u00e9 jusqu\u2019\u00e0 pr\u00e9sent. G\u00e9n\u00e9ralement, les variables cat\u00e9gorielles sont dites <\/span><b>\u00ab\u00a0nominales\u00a0\u00bb<\/b><span style=\"font-weight: 400;\">, dans le sens o\u00f9 elles admettent un ensemble fini de valeurs qui n\u2019admettent <\/span><b>aucune relation entre elles<\/b><span style=\"font-weight: 400;\">. Dans notre DataFrame par exemple, les variables \u00ab\u00a0nom\u00a0\u00bb et \u00ab\u00a0\u00e2ge\u00a0\u00bb sont nominales, car elles contiennent respectivement 3 et 2 valeurs sans lien entre elles.<\/span><\/p><p><span style=\"font-weight: 400;\">Pourtant, lorsque ces variables cat\u00e9gorielles admettent une <\/span><b>relation d\u2019ordre<\/b><span style=\"font-weight: 400;\"> entre leurs valeurs, elles sont alors consid\u00e9r\u00e9es comme <\/span><b>\u00ab\u00a0ordinales\u00a0\u00bb<\/b><span style=\"font-weight: 400;\">. Cela peut \u00eatre une mention \u00e0 un examen, un classement \u00e0 une comp\u00e9tition, une tranche d\u2019\u00e2ge, etc. Dans notre exemple, la variable \u00ab\u00a0taille\u00a0\u00bb est une variable ordinale, puisqu\u2019on peut y \u00e9tablir un <\/span><b>classement entre ses valeurs<\/b><span style=\"font-weight: 400;\"> : petite &gt; moyenne &gt; grande.<\/span><\/p><p><span style=\"font-weight: 400;\">Ainsi pour traiter des variables ordinales, l\u2019<\/span><b>encodage ordinal <\/b><span style=\"font-weight: 400;\">sera plus ad\u00e9quat, puisqu\u2019il <\/span><b>consid\u00e8rera et maintiendra le classement<\/b><span style=\"font-weight: 400;\"> \u00e9tabli lors de l\u2019encodage des valeurs. Appliquons un encodage ordinal \u00e0 notre DataFrame sur la colonne \u00ab\u00a0taille\u00a0\u00bb :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-2de8fe0 elementor-widget elementor-widget-image\" data-id=\"2de8fe0\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"258\" height=\"117\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image6.png\" class=\"attachment-full size-full wp-image-233858\" alt=\"\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-cdc71e0 elementor-widget elementor-widget-text-editor\" data-id=\"cdc71e0\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Comme nous pouvons le constater, en plus d\u2019avoir converti num\u00e9riquement les valeurs de la colonne, <\/span><b>la hi\u00e9rarchie de nos tailles a bien \u00e9t\u00e9 maintenue<\/b><span style=\"font-weight: 400;\"> !<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-852b591 elementor-widget elementor-widget-image\" data-id=\"852b591\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"1000\" height=\"605\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders2.jpg\" class=\"attachment-full size-full wp-image-233868\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders2.jpg 1000w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders2-300x182.jpg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders2-768x465.jpg 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-59cd5ab elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"59cd5ab\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formation-data-analyst\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Ma\u00eetriser l'encodage de donn\u00e9es<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-bab9d82 e-flex e-con-boxed e-con\" data-id=\"bab9d82\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-d1a8be2 elementor-widget elementor-widget-heading\" data-id=\"d1a8be2\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">L\u2019encodage cible<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-370efef elementor-widget elementor-widget-text-editor\" data-id=\"370efef\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Dans tout mod\u00e8le de Machine Learning, il y a une variable cible que nous voulons pr\u00e9dire via ce mod\u00e8le. Dans notre exemple, il s\u2019agit de la situation professionnelle. Ainsi il peut \u00eatre judicieux d\u2019appliquer un encodage sur nos variables cat\u00e9gorielles qui, au lieu de leur d\u00e9finir une certaine valeur num\u00e9rique, la <\/span><b>calcule en fonction de la variable cible<\/b><span style=\"font-weight: 400;\">. C\u2019est ici que s\u2019applique l\u2019<\/span><b>encodage cible<\/b><span style=\"font-weight: 400;\">, ou <\/span><b><i>\u00ab\u00a0target encoder\u00a0\u00bb <\/i><\/b><span style=\"font-weight: 400;\">!<\/span><\/p><p><span style=\"font-weight: 400;\">L\u2019encodage cible traduit les variables cat\u00e9gorielles en calculant la <\/span><b>moyenne<\/b> <b>de la variable cible correspondante<\/b><span style=\"font-weight: 400;\"> pour chaque cat\u00e9gorie distincte de la variable cat\u00e9gorielle que nous souhaitons encoder. Appliquons cet encodage sur les colonnes <\/span><i><span style=\"font-weight: 400;\">sexe<\/span><\/i><span style=\"font-weight: 400;\"> et <\/span><i><span style=\"font-weight: 400;\">taille <\/span><\/i><span style=\"font-weight: 400;\">de notre DataFrame d\u2019exemple, en sp\u00e9cifiant bien que notre variable cible est la colonne <\/span><i><span style=\"font-weight: 400;\">situation<\/span><\/i><span style=\"font-weight: 400;\"> :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-b08b5f2 elementor-widget elementor-widget-image\" data-id=\"b08b5f2\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"294\" height=\"110\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image5.png\" class=\"attachment-full size-full wp-image-233859\" alt=\"\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9b86343 elementor-widget elementor-widget-text-editor\" data-id=\"9b86343\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Ici, chaque valeur des colonnes <\/span><i><span style=\"font-weight: 400;\">sexe <\/span><\/i><span style=\"font-weight: 400;\">et <\/span><i><span style=\"font-weight: 400;\">taille <\/span><\/i><span style=\"font-weight: 400;\">prend bien une valeur num\u00e9rique calcul\u00e9e comme la moyenne de la variable cible <\/span><i><span style=\"font-weight: 400;\">situation<\/span><\/i><span style=\"font-weight: 400;\">. Le <\/span><b>calcul g\u00e9n\u00e9ral<\/b><span style=\"font-weight: 400;\"> pour chaque cat\u00e9gorie unique dans les colonnes cat\u00e9gorielles est effectu\u00e9 comme suit :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-23b8848 elementor-widget elementor-widget-image\" data-id=\"23b8848\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"572\" height=\"68\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image2.png\" class=\"attachment-full size-full wp-image-233860\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image2.png 572w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/image2-300x36.png 300w\" sizes=\"(max-width: 572px) 100vw, 572px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-433955d elementor-widget elementor-widget-text-editor\" data-id=\"433955d\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Ce calcul permet d&rsquo;obtenir une <\/span><b>repr\u00e9sentation num\u00e9rique<\/b><span style=\"font-weight: 400;\"> de chaque cat\u00e9gorie cat\u00e9gorielle bas\u00e9e sur la moyenne de la variable cible associ\u00e9e \u00e0 cette cat\u00e9gorie. Ces valeurs encod\u00e9es peuvent ensuite \u00eatre utilis\u00e9es comme <\/span><b>caract\u00e9ristiques <\/b><span style=\"font-weight: 400;\">dans un mod\u00e8le de Machine Learning.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-84d413f e-flex e-con-boxed e-con\" data-id=\"84d413f\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-bddbf51 elementor-widget elementor-widget-heading\" data-id=\"bddbf51\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Conclusion<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-6c6f39d elementor-widget elementor-widget-text-editor\" data-id=\"6c6f39d\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<span style=\"font-weight: 400;\">La librairie <i>category_encoders<\/i> offre un large \u00e9ventail d\u2019outils pour encoder les variables cat\u00e9gorielles en valeurs num\u00e9riques. Si on a pr\u00e9sent\u00e9 ici quatre m\u00e9thodes, <i>category_encoders<\/i> prend en charge autour d\u2019une quinzaine de techniques d\u2019encodage, qu\u2019il est possible d\u2019utiliser selon diverses situations !<\/span>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-41186e2 elementor-widget elementor-widget-image\" data-id=\"41186e2\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"1000\" height=\"634\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders3.jpg\" class=\"attachment-full size-full wp-image-233869\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders3.jpg 1000w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders3-300x190.jpg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/category_encoders3-768x487.jpg 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0301283 elementor-widget elementor-widget-text-editor\" data-id=\"0301283\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>Pour <strong>devenir un expert en encodage des donn\u00e9es<\/strong>, n\u2019h\u00e9sitez pas \u00e0 suivre les formations <a href=\"https:\/\/datascientest.com\/formation-data-scientist\">Data Scientist<\/a> &amp; <a href=\"https:\/\/datascientest.com\/formation-data-analyst\">Data Analyst<\/a> propos\u00e9es par <strong>DataScientest<\/strong> !<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d4732c8 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"d4732c8\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formation-data-analyst\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Se former sur l'encodage de donn\u00e9es<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9367605 elementor-widget elementor-widget-html\" data-id=\"9367605\" data-element_type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<script type=\"application\/ld+json\">\r\n{\r\n  \"@context\": \"https:\/\/schema.org\",\r\n  \"@type\": \"FAQPage\",\r\n  \"mainEntity\": [{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Qu\u2019est-ce que la librairie Python category encoders ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"La librairie Python category encoders, comme son nom l\u2019indique, offre alors une vari\u00e9t\u00e9 de techniques d\u2019encodage pour nos variables. Comme toute librairie, elle s\u2019installe avec la commande pip install category_encoders et s\u2019importe avec import category_encoders as ce.\"\r\n    }\r\n  },{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Quelles sont les diff\u00e9rentes techniques d\u2019encodage ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Il existe 4 techniques d\u2019encodage : l'encodage One-Hot, l'encodage binaire, l'encodage ordinal, l'encodage cible.\"\r\n    }\r\n  },{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Comment devenir expert en encodage de donn\u00e9es ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Pour devenir un expert en encodage des donn\u00e9es, n\u2019h\u00e9sitez pas \u00e0 suivre les formations Data Scientist & Data Analyst propos\u00e9es par DataScientest !\"\r\n    }\r\n  }]\r\n}\r\n<\/script>\r\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>Dans un jeu de donn\u00e9es, deux types de variables pr\u00e9dominent : les variables quantitatives, qui mesurent des quantit\u00e9s (une taille, un prix\u2026), et les variables qualitatives ou cat\u00e9gorielles, qui d\u00e9terminent plut\u00f4t des cat\u00e9gories ou des modalit\u00e9s (un sexe, une couleur\u2026). Or comme le Machine Learning s\u2019appuie sur diverses op\u00e9rations math\u00e9matiques, il est crucial de convertir, [&hellip;]<\/p>\n","protected":false},"author":77,"featured_media":233853,"comment_status":"open","ping_status":"open","sticky":false,"template":"elementor_theme","format":"standard","meta":{"content-type":"","inline_featured_image":false},"categories":[2497],"tags":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v21.7 (Yoast SEO v21.7) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Category encoders : le couteau suisse de l\u2019encodage<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/datascientest.com\/category-encoders-tout-savoir\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Category encoders : le couteau suisse de l\u2019encodage\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/category-encoders-tout-savoir\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2024-03-21T09:27:07+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-03-21T09:27:08+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/Illu_BLOG_category_encoders.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1024\" \/>\n\t<meta property=\"og:image:height\" content=\"562\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Raphael Kassel\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u00c9crit par\" \/>\n\t<meta name=\"twitter:data1\" content=\"Raphael Kassel\" \/>\n\t<meta name=\"twitter:label2\" content=\"Dur\u00e9e de lecture estim\u00e9e\" \/>\n\t<meta name=\"twitter:data2\" content=\"7 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/datascientest.com\/category-encoders-tout-savoir\",\"url\":\"https:\/\/datascientest.com\/category-encoders-tout-savoir\",\"name\":\"Category encoders : le couteau suisse de l\u2019encodage\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2024-03-21T09:27:07+00:00\",\"dateModified\":\"2024-03-21T09:27:08+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\"},\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/category-encoders-tout-savoir\"]}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/datascientest.com\/#website\",\"url\":\"https:\/\/datascientest.com\/\",\"name\":\"Formation Data Science | DataScientest.com\",\"description\":\"Leader de la formation en Data Science\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/datascientest.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"fr-FR\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\",\"name\":\"Raphael Kassel\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg\",\"contentUrl\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg\",\"caption\":\"Raphael Kassel\"},\"description\":\"Raphael Kassel est le Head of Data Programs chez DataScientest. Il supervise une \u00e9quipe Data, impl\u00e9mente de nouveaux parcours comme \\\"Chef de projet IA\\\" et \\\"Data Product Manager\\\", et contribue \u00e0 la modernisation des op\u00e9rations du groupe. En collaboration \u00e9troite avec l'\u00e9quipe de d\u00e9veloppement, il adapte l'accompagnement acad\u00e9mique aux \u00e9volutions de l'\u00e9quipe data et \u00e0 l'internationalisation.\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/raphaelkassel\/\",\"Data Science\"],\"url\":\"https:\/\/datascientest.com\/author\/raphaelkassel\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Category encoders : le couteau suisse de l\u2019encodage","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/datascientest.com\/category-encoders-tout-savoir","og_locale":"fr_FR","og_type":"article","og_title":"Category encoders : le couteau suisse de l\u2019encodage","og_url":"https:\/\/datascientest.com\/category-encoders-tout-savoir","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2024-03-21T09:27:07+00:00","article_modified_time":"2024-03-21T09:27:08+00:00","og_image":[{"width":1024,"height":562,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/Illu_BLOG_category_encoders.png","type":"image\/png"}],"author":"Raphael Kassel","twitter_card":"summary_large_image","twitter_misc":{"\u00c9crit par":"Raphael Kassel","Dur\u00e9e de lecture estim\u00e9e":"7 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/datascientest.com\/category-encoders-tout-savoir","url":"https:\/\/datascientest.com\/category-encoders-tout-savoir","name":"Category encoders : le couteau suisse de l\u2019encodage","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2024-03-21T09:27:07+00:00","dateModified":"2024-03-21T09:27:08+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1"},"inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/category-encoders-tout-savoir"]}]},{"@type":"WebSite","@id":"https:\/\/datascientest.com\/#website","url":"https:\/\/datascientest.com\/","name":"Formation Data Science | DataScientest.com","description":"Leader de la formation en Data Science","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/datascientest.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"fr-FR"},{"@type":"Person","@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1","name":"Raphael Kassel","image":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/datascientest.com\/#\/schema\/person\/image\/","url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg","contentUrl":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg","caption":"Raphael Kassel"},"description":"Raphael Kassel est le Head of Data Programs chez DataScientest. Il supervise une \u00e9quipe Data, impl\u00e9mente de nouveaux parcours comme \"Chef de projet IA\" et \"Data Product Manager\", et contribue \u00e0 la modernisation des op\u00e9rations du groupe. En collaboration \u00e9troite avec l'\u00e9quipe de d\u00e9veloppement, il adapte l'accompagnement acad\u00e9mique aux \u00e9volutions de l'\u00e9quipe data et \u00e0 l'internationalisation.","sameAs":["https:\/\/www.linkedin.com\/in\/raphaelkassel\/","Data Science"],"url":"https:\/\/datascientest.com\/author\/raphaelkassel"}]}},"_links":{"self":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/233851"}],"collection":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/users\/77"}],"replies":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/comments?post=233851"}],"version-history":[{"count":0,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/233851\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/233853"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=233851"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=233851"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=233851"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}