Uygulamada genellikle, \u00f6\u011frenci kal\u0131nt\u0131s\u0131 3’\u00fcn mutlak de\u011ferinden b\u00fcy\u00fck olan bir veri setindeki herhangi bir g\u00f6zlemin<\/a> ayk\u0131r\u0131 de\u011fer oldu\u011funu s\u00f6yleriz.<\/span><\/p>\n A\u015fa\u011f\u0131daki s\u00f6zdizimini kullanan MASS paketindeki studres()<\/strong> i\u015flevini kullanarak R’deki herhangi bir regresyon modelinin \u00f6\u011frencile\u015ftirilmi\u015f art\u0131klar\u0131n\u0131 h\u0131zl\u0131 bir \u015fekilde elde edebiliriz:<\/span><\/p>\n \u00e7iviler(model)<\/strong><\/span><\/p>\n burada model<\/em> herhangi bir do\u011frusal modeli temsil eder.<\/span><\/p>\n Yerle\u015fik mtcars<\/em> veri k\u00fcmesini kullanarak R’de a\u015fa\u011f\u0131daki basit do\u011frusal regresyon<\/a> modelini olu\u015fturdu\u011fumuzu varsayal\u0131m:<\/span><\/p>\n Veri k\u00fcmesindeki her g\u00f6zlem i\u00e7in \u00f6\u011frencile\u015ftirilmi\u015f art\u0131klar\u0131 hesaplamak amac\u0131yla MASS paketindeki studres()<\/strong> fonksiyonunu kullanabiliriz:<\/span><\/p>\n Ayr\u0131ca, yorday\u0131c\u0131 de\u011fi\u015fkenlerin de\u011ferlerinin, \u00fczerinde \u00e7al\u0131\u015f\u0131lan kar\u015f\u0131l\u0131k gelen art\u0131klara kar\u015f\u0131 h\u0131zl\u0131 bir grafi\u011fini de olu\u015fturabiliriz:<\/span> <\/p>\n Grafikten, hi\u00e7bir g\u00f6zlemin mutlak de\u011feri 3’ten b\u00fcy\u00fck olan bir \u00f6\u011frenci kal\u0131nt\u0131s\u0131na sahip olmad\u0131\u011f\u0131n\u0131 g\u00f6rebiliriz, dolay\u0131s\u0131yla veri setinde net ayk\u0131r\u0131 de\u011ferler yoktur.<\/span><\/p>\n \u0130sterseniz orijinal veri setindeki her g\u00f6zlemin \u00f6\u011frencile\u015ftirilmi\u015f kal\u0131nt\u0131lar\u0131n\u0131 da geri ekleyebiliriz:<\/span><\/p>\n Daha sonra hangi g\u00f6zlemlerin ayk\u0131r\u0131 de\u011ferlere en yak\u0131n oldu\u011fu hakk\u0131nda bir fikir edinmek i\u00e7in her g\u00f6zlemi \u00f6\u011frenci kal\u0131nt\u0131s\u0131na g\u00f6re en b\u00fcy\u00fckten en k\u00fc\u00e7\u00fc\u011fe do\u011fru s\u0131ralayabiliriz:<\/span><\/p>\n R’de basit do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a> \u00d6\u011frenci kal\u0131nt\u0131s\u0131, basit\u00e7e, tahmini standart sapmas\u0131na b\u00f6l\u00fcnen bir kal\u0131nt\u0131d\u0131r. Uygulamada genellikle, \u00f6\u011frenci kal\u0131nt\u0131s\u0131 3’\u00fcn mutlak de\u011ferinden b\u00fcy\u00fck olan bir veri setindeki herhangi bir g\u00f6zlemin ayk\u0131r\u0131 de\u011fer oldu\u011funu s\u00f6yleriz. A\u015fa\u011f\u0131daki s\u00f6zdizimini kullanan MASS paketindeki studres() i\u015flevini kullanarak R’deki herhangi bir regresyon modelinin \u00f6\u011frencile\u015ftirilmi\u015f art\u0131klar\u0131n\u0131 h\u0131zl\u0131 bir \u015fekilde elde edebiliriz: \u00e7iviler(model) burada model herhangi bir do\u011frusal modeli […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[11],"tags":[],"class_list":["post-1247","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n \u00d6rnek: R’de \u00f6\u011frencile\u015ftirilmi\u015f art\u0131klar\u0131n hesaplanmas\u0131<\/strong><\/span><\/h3>\n
#build simple linear regression model<\/span>\nmodel <- lm(mpg ~ disp, data=mtcars)\n<\/strong><\/pre>\n library<\/span> (MASS)<\/span>\n\n#calculate studentized residuals<\/span>\nstud_resids <- studres(model)\n\n#view first three studentized residuals<\/span>\nhead(stud_resids, 3)\n\n Mazda RX4 Mazda RX4 Wag Datsun 710 \n -0.6236250 -0.6236250 -0.7405315 \n<\/strong><\/pre>\n #plot predictor variable vs. studentized residuals\n<\/span>plot(mtcars$disp, stud_resids, ylab=' Studentized Residuals<\/span> ', xlab=' Displacement<\/span> ') \n\n#add horizontal line at 0\n<\/span>abline(0, 0)<\/strong> <\/pre>\n![\"R'de](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/etudiant1.png\")
<\/p>\n #add studentized residuals to orignal dataset\n<\/span>final_data <- cbind<\/span> (mtcars[c(' mpg<\/span> ', ' disp<\/span> ')], stud_resids)\n\n#view final dataset\n<\/span>head(final_data)\n\n mpg disp stud_resids\nMazda RX4 21.0 160 -0.6236250\nMazda RX4 Wag 21.0 160 -0.6236250\nDatsun 710 22.8 108 -0.7405315\nHornet 4 Drive 21.4 258 0.7556078\nHornet Sportabout 18.7 360 1.2658336\nValiant 18.1 225 -0.6896297<\/strong><\/pre>\n #sort studentized residuals descending\n<\/span>final_data[ order<\/span> (-stud_resids),]\n\n mpg disp stud_resids\nToyota Corolla 33.9 71.1 2.52397102\nPontiac Firebird 19.2 400.0 2.06825391\nFiat 128 32.4 78.7 2.03684699\nLotus Europa 30.4 95.1 1.53905536\nHonda Civic 30.4 75.7 1.27099586\nHornet Sportabout 18.7 360.0 1.26583364\nChrysler Imperial 14.7 440.0 1.06486066\nHornet 4 Drive 21.4 258.0 0.75560776\nPorsche 914-2 26.0 120.3 0.42424678\nFiat X1-9 27.3 79.0 0.30183728\nMerc 240D 24.4 146.7 0.26235893\nFord Pantera L 15.8 351.0 0.20825609\nCadillac Fleetwood 10.4 472.0 0.08338531\nLincoln Continental 10.4 460.0 -0.07863385\nDuster 360 14.3 360.0 -0.14476167\nMerc 450SL 17.3 275.8 -0.28759769\nDodge Challenger 15.5 318.0 -0.30826585\nMerc 230 22.8 140.8 -0.30945955\nMerc 450SE 16.4 275.8 -0.56742476\nAMC Javelin 15.2 304.0 -0.58138205\nCamaro Z28 13.3 350.0 -0.58848471\nMazda RX4 Wag 21.0 160.0 -0.62362497\nMazda RX4 21.0 160.0 -0.62362497\nMaserati Bora 15.0 301.0 -0.68315010\nValiant 18.1 225.0 -0.68962974\nDatsun 710 22.8 108.0 -0.74053152\nMerc 450SLC 15.2 275.8 -0.94814699\nToyota Corona 21.5 120.1 -0.99751166\nVolvo 142E 21.4 121.0 -1.01790487\nMerc 280 19.2 167.6 -1.09979261\nFerrari Dino 19.7 145.0 -1.24732999\nMerc 280C 17.8 167.6 -1.57258064\n<\/strong><\/pre>\n Ek kaynaklar<\/strong><\/span><\/h3>\n
R’de \u00e7oklu do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a>
R’de art\u0131k arsa nas\u0131l olu\u015fturulur<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"