{"id":2482,"date":"2023-07-22T02:13:43","date_gmt":"2023-07-22T02:13:43","guid":{"rendered":"https:\/\/statorials.org\/tr\/rde-parcali-regresyon\/"},"modified":"2023-07-22T02:13:43","modified_gmt":"2023-07-22T02:13:43","slug":"rde-parcali-regresyon","status":"publish","type":"post","link":"https:\/\/statorials.org\/tr\/rde-parcali-regresyon\/","title":{"rendered":"R'de par\u00e7al\u0131 regresyon nas\u0131l ger\u00e7ekle\u015ftirilir (ad\u0131m ad\u0131m)"},"content":{"rendered":"

<\/p>\n

\n

Par\u00e7al\u0131 regresyon<\/strong> , bir veri setinde net “kesme noktalar\u0131” oldu\u011funda s\u0131kl\u0131kla kulland\u0131\u011f\u0131m\u0131z bir regresyon y\u00f6ntemidir.<\/span><\/p>\n

A\u015fa\u011f\u0131daki ad\u0131m ad\u0131m \u00f6rnek, R’de par\u00e7al\u0131 regresyonun nas\u0131l ger\u00e7ekle\u015ftirilece\u011fini g\u00f6sterir.<\/span><\/p>\n

1. Ad\u0131m: Verileri olu\u015fturun<\/strong><\/span><\/h3>\n

\u00d6ncelikle a\u015fa\u011f\u0131daki veri \u00e7er\u00e7evesini olu\u015ftural\u0131m:<\/span><\/p>\n

 #view DataFrame\n<\/span>df <- data. frame<\/span> (x=c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16),\n                 y=c(2, 4, 5, 6, 8, 10, 12, 13, 15, 19, 24, 28, 31, 34, 39, 44))\n\n#view first six rows of data frame<\/span>\nhead(df)\n\n  xy\n1 1 2\n2 2 4\n3 3 5\n4 4 6\n5 5 8\n6 6 10\n<\/strong><\/pre>\n
 2. Ad\u0131m: Verileri g\u00f6rselle\u015ftirin<\/strong><\/span><\/h3>\n
 Daha sonra verileri g\u00f6rselle\u015ftirmek i\u00e7in bir da\u011f\u0131l\u0131m grafi\u011fi olu\u015ftural\u0131m:<\/span> <\/p>\n
 #create scatterplot of x vs. y\n<\/span>plot(df$x, df$y, pch= 16<\/span> , col=' steelblue<\/span> ')\n<\/strong><\/pre>\n
\"\"<\/p>\n
 X ve y aras\u0131ndaki ili\u015fkinin x = 9<\/strong> civar\u0131nda aniden de\u011fi\u015fti\u011fini g\u00f6rebiliriz.<\/span><\/p>\n
 Ad\u0131m 3: Par\u00e7al\u0131 regresyon modelini yerle\u015ftirin<\/strong><\/span><\/h3>\n
 Par\u00e7al\u0131 bir regresyon modelini veri k\u00fcmemize s\u0131\u011fd\u0131rmak i\u00e7in R’deki b\u00f6l\u00fcmlenmi\u015f<\/a> paketteki segmented()<\/strong> i\u015flevini kullanabiliriz:<\/span><\/p>\n
 library<\/span> (segmented)\n\n#fit simple linear regression model<\/span>\nfit <- lm(y ~ x, data=df)\n\n#fit piecewise regression model to original model, estimating a breakpoint at x=9\n<\/span>segmented. fit<\/span> <- segmented(fit, seg.Z = ~x, psi= 9<\/span> )\n\n#view summary of segmented model\n<\/span>summary( segmented.fit<\/span> )\n\nCall: \nsegmented.lm(obj = fit, seg.Z = ~x, psi = 9)\n\nEstimated Break-Point(s):\n         East. St.Err\npsi1.x 8.762 0.26\n\nMeaningful coefficients of the linear terms:\n            Estimate Std. Error t value Pr(>|t|)    \n(Intercept) 0.32143 0.48343 0.665 0.519    \nx 1.59524 0.09573 16.663 1.16e-09 ***\nU1.x 2.40476 0.13539 17.762 NA    \n---\nSignificant. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nResidual standard error: 0.6204 on 12 degrees of freedom\nMultiple R-Squared: 0.9983, Adjusted R-squared: 0.9978 \n\nConvergence achieved in 2 iter. (rel. changes 0)\n<\/strong><\/pre>\n
 segmented()<\/strong> i\u015flevi x = 8,762’de bir kesme noktas\u0131 tespit eder.<\/span><\/p>\n
 Uygun par\u00e7al\u0131 regresyon modeli \u015fu \u015fekildedir:<\/span><\/p>\n
 E\u011fer x \u2264 8,762 ise: y = 0,32143 + 1,59524*(x)<\/span><\/p>\n
 E\u011fer x > 8,762 ise: y = 0,32143 + 1,59524*(8,762) + (1,59524+2,40476)*(x-8,762)<\/span><\/p>\n
 \u00d6rne\u011fin x = 5<\/strong> de\u011ferine sahip oldu\u011fumuzu varsayal\u0131m. Tahmini de\u011fer \u015f\u00f6yle olacakt\u0131r:<\/span><\/p>\n