<\/p>\n
Bu e\u011fitimde, nokta bulutlar\u0131 i\u00e7in R’de titre\u015fim<\/strong> fonksiyonunun ne zaman ve nas\u0131l kullan\u0131laca\u011f\u0131 a\u00e7\u0131klanmaktad\u0131r.<\/span><\/p>\n Da\u011f\u0131l\u0131m grafikleri,<\/strong> iki s\u00fcrekli de\u011fi\u015fken aras\u0131ndaki ili\u015fkiyi g\u00f6rselle\u015ftirmek i\u00e7in m\u00fckemmeldir. \u00d6rne\u011fin, a\u015fa\u011f\u0131daki da\u011f\u0131l\u0131m grafi\u011fi 100 sporcunun boy ve kilo aras\u0131ndaki ili\u015fkiyi g\u00f6rselle\u015ftirmemize yard\u0131mc\u0131 olur:<\/span><\/p>\n Ancak bazen s\u00fcrekli bir de\u011fi\u015fken ile neredeyse<\/em> s\u00fcrekli olan ba\u015fka bir de\u011fi\u015fken aras\u0131ndaki ili\u015fkiyi g\u00f6rselle\u015ftirmek isteyebiliriz.<\/span><\/p>\n \u00d6rne\u011fin, bir basketbolcunun sezonun ilk 10 ma\u00e7\u0131nda ba\u015flad\u0131\u011f\u0131 ma\u00e7 say\u0131s\u0131n\u0131 ve ma\u00e7 ba\u015f\u0131na ald\u0131\u011f\u0131 puan\u0131 g\u00f6steren a\u015fa\u011f\u0131daki veri setine sahip oldu\u011fumuzu varsayal\u0131m:<\/span><\/p>\n Oyun ba\u015f\u0131na puan<\/em> s\u00fcrekli bir de\u011fi\u015fkendir, ancak ba\u015flat\u0131lan oyunlar<\/em> ayr\u0131k bir de\u011fi\u015fkendir. Bu iki de\u011fi\u015fkenin da\u011f\u0131l\u0131m grafi\u011fini olu\u015fturmaya \u00e7al\u0131\u015f\u0131rsak \u015f\u00f6yle g\u00f6r\u00fcnecektir:<\/span><\/p>\n Bu da\u011f\u0131l\u0131m grafi\u011finden, ba\u015flayan oyunlar<\/em> ile oyun ba\u015f\u0131na ortalama puanlar\u0131n<\/em> pozitif bir ili\u015fkiye sahip oldu\u011funu s\u00f6yleyebiliriz, ancak \u00e7o\u011fu \u00f6rt\u00fc\u015ft\u00fc\u011f\u00fc i\u00e7in olay \u00f6rg\u00fcs\u00fcnde bireysel noktalar\u0131 g\u00f6rmek biraz zor.<\/span><\/p>\n Jitter<\/strong> i\u015flevini kullanarak, at\u0131lan X ekseni de\u011fi\u015fken k\u00fcmelerine<\/em> bir miktar “g\u00fcr\u00fclt\u00fc” ekleyebiliriz, b\u00f6ylece \u00e7izimdeki bireysel noktalar\u0131 daha net g\u00f6rebiliriz:<\/span><\/p>\n Verilere daha da fazla g\u00fcr\u00fclt\u00fc eklemek i\u00e7in iste\u011fe ba\u011fl\u0131 olarak titre\u015fime say\u0131sal bir arg\u00fcman ekleyebiliriz:<\/span><\/p>\n Ancak \u00e7ok fazla titre\u015fim eklememeye dikkat etmeliyiz \u00e7\u00fcnk\u00fc bu, orijinal verileri \u00e7ok fazla bozabilir:<\/span><\/p>\n Karars\u0131zl\u0131k, \u00f6zellikle ayr\u0131k de\u011fi\u015fkenin d\u00fczeylerinden birinin di\u011fer d\u00fczeylerden \u00e7ok daha fazla de\u011fere sahip olmas\u0131 durumunda kullan\u0131\u015fl\u0131d\u0131r.<\/span><\/p>\n \u00d6rne\u011fin, a\u015fa\u011f\u0131daki veri setinde sezonun ilk 5 ma\u00e7\u0131ndan 2’sine ba\u015flayan \u00fc\u00e7 y\u00fcz basketbol oyuncusu var, ancak yaln\u0131zca 1, 3, 4 veya 5 ma\u00e7a ba\u015flayan yakla\u015f\u0131k 100 oyuncu var: Oynanan oyun say\u0131s\u0131n\u0131 ma\u00e7 ba\u015f\u0131na ortalama puana g\u00f6re g\u00f6rselle\u015ftirdi\u011fimizde, 2 oyun oynayan oyuncular\u0131n say\u0131s\u0131n\u0131n daha fazla oldu\u011funu s\u00f6yleyebiliriz ancak ka\u00e7 ki\u015finin<\/em> 2 oyun oynad\u0131\u011f\u0131n\u0131 tam olarak s\u00f6ylemek zor:<\/span><\/p>\n Ancak oyun ba\u015flang\u0131c\u0131<\/em> de\u011fi\u015fkenine titre\u015fimi ekledi\u011fimizde, 2 oyuna ka\u00e7 oyuncunun daha ba\u015flad\u0131\u011f\u0131n\u0131 g\u00f6rebiliriz:<\/span><\/p>\n Titre\u015fim miktar\u0131n\u0131 biraz art\u0131rmak bu fark\u0131 daha da ortaya \u00e7\u0131kar\u0131r:<\/span><\/p>\n Yaln\u0131zca g\u00f6rselle\u015ftirmeler i\u00e7in titre\u015fim<\/strong><\/span><\/p>\n<\/h3>\n Daha \u00f6nce de belirtildi\u011fi gibi titre\u015fim, verilere rastgele g\u00fcr\u00fclt\u00fc ekler; bu, verileri bir nokta bulutunda g\u00f6rselle\u015ftirmek istedi\u011fimizde faydal\u0131 olabilir. Titre\u015fim fonksiyonunu kullanarak, bir veri setindeki iki de\u011fi\u015fken aras\u0131ndaki ger\u00e7ek temel ili\u015fkinin daha iyi bir resmini elde edebiliriz.<\/span><\/p>\n Ancak regresyon gibi istatistiksel analizler kullan\u0131ld\u0131\u011f\u0131nda, bir veri setindeki de\u011fi\u015fkenlere rastgele g\u00fcr\u00fclt\u00fc eklemek, analizin sonu\u00e7lar\u0131n\u0131 etkileyece\u011finden mant\u0131kl\u0131 de\u011fildir.<\/span><\/p>\n Dolay\u0131s\u0131yla titre\u015fimin veri analizi i\u00e7in de\u011fil, yaln\u0131zca veri g\u00f6rselle\u015ftirmesi i\u00e7in kullan\u0131lmas\u0131 ama\u00e7lanm\u0131\u015ft\u0131r.<\/span><\/p>\n","protected":false},"excerpt":{"rendered":" Bu e\u011fitimde, nokta bulutlar\u0131 i\u00e7in R’de titre\u015fim fonksiyonunun ne zaman ve nas\u0131l kullan\u0131laca\u011f\u0131 a\u00e7\u0131klanmaktad\u0131r. Jitter ne zaman kullan\u0131l\u0131r? Da\u011f\u0131l\u0131m grafikleri, iki s\u00fcrekli de\u011fi\u015fken aras\u0131ndaki ili\u015fkiyi g\u00f6rselle\u015ftirmek i\u00e7in m\u00fckemmeldir. \u00d6rne\u011fin, a\u015fa\u011f\u0131daki da\u011f\u0131l\u0131m grafi\u011fi 100 sporcunun boy ve kilo aras\u0131ndaki ili\u015fkiyi g\u00f6rselle\u015ftirmemize yard\u0131mc\u0131 olur: #define vectors of heights and weights weights <- runif(100, 160, 240) heights <- […]<\/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-479","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n Jitter ne zaman kullan\u0131l\u0131r?<\/strong><\/span><\/h3>\n
#define vectors of heights and weights\n<\/span>weights <- runif(100, 160, 240) \nheights <- (weights\/3) + rnorm(100)\n\n#create data frame of heights and weights\n<\/span>data <- as.data.frame(cbind(weights, heights))\n\n#view first six rows of data frame\n<\/span>head(data)\n\n# weights heights\n#1 170.8859 57.20745\n#2 183.2481 62.01162\n#3 235.6884 77.93126\n#4 231.9864 77.12520\n#5 200.8562 67.93486\n#6 169.6987 57.54977\n#create scatterplot of heights vs weights\n<\/span>plot(data$weights, data$heights, pch = 16, col = 'steelblue')<\/strong><\/pre>\n #create data frame<\/span>\ngames_started <- sample(1:10, 300, TRUE)\npoints_per_game <- 3*games_started + rnorm(300)\ndata <- as.data.frame(cbind(games_started, points_per_game))\n\n#view first six rows of data frame\n<\/span>head(data)\n\n# games_started points_per_game\n#1 9 25.831554\n#2 9 26.673983\n#3 10 29.850948\n#4 4 12.024353\n#5 4 11.534192\n#6 1 4.383127<\/strong><\/pre>\n #create scatterplot of games started vs average points per game\n<\/span>plot(data$games_started, data$points_per_game, pch = 16, col = 'steelblue')<\/strong><\/pre>\n #add jitter to games started\n<\/em><\/span>plot( jitter<\/span> (data$games_started), data$points_per_game, pch = 16, col = 'steelblue')<\/strong><\/pre>\n #add jitter to games started\n<\/em><\/span>plot( jitter<\/span> (data$games_started, 2<\/span> ), data$points_per_game, pch = 16, col = 'steelblue')<\/strong><\/pre>\n plot( jitter<\/span> (data$games_started, 20<\/span> ), data$points_per_game, pch = 16, col = 'steelblue')<\/strong><\/pre>\n Titre\u015fim, verilerin daha iyi g\u00f6r\u00fclmesini sa\u011flar<\/strong><\/span><\/h3>\n
<\/span><\/p>\n games_started <- sample(1:5, 100, TRUE)\npoints_per_game <- 3*games_started + rnorm(100)\ndata <- as.data.frame(cbind(games_started, points_per_game))\n\ngames_twos <- rep(2, 200)\npoints_twos <- 3*games_twos + rnorm(200)\ndata_twos <- as.data.frame(cbind(games_twos, points_twos))\nnames(data_twos) <- c('games_started', 'points_per_game')\n\nall_data <- rbind(data, data_twos)<\/strong><\/pre>\n plot(all_data$games_started, all_data$points_per_game, pch = 16, col = 'steelblue')<\/strong><\/pre>\n
plot( jitter<\/span> (all_data$games_started), all_data$points_per_game,\n pch=16, col='steelblue')<\/strong><\/pre>\n plot( jitter<\/span> (all_data$games_started, 1.5<\/span> ), all_data$points_per_game,\n pch=16, col='steelblue')<\/strong><\/pre>\n\n