{"id":2482,"date":"2023-07-22T02:13:43","date_gmt":"2023-07-22T02:13:43","guid":{"rendered":"https:\/\/statorials.org\/id\/regresi-sedikit-demi-sedikit-di-r\/"},"modified":"2023-07-22T02:13:43","modified_gmt":"2023-07-22T02:13:43","slug":"regresi-sedikit-demi-sedikit-di-r","status":"publish","type":"post","link":"https:\/\/statorials.org\/id\/regresi-sedikit-demi-sedikit-di-r\/","title":{"rendered":"Cara melakukan regresi sepotong di r (langkah demi langkah)"},"content":{"rendered":"

<\/p>\n

\n

Regresi sepotong-sepotong<\/strong> adalah metode regresi yang sering kita gunakan ketika terdapat \u201cbreakpoint\u201d yang jelas dalam suatu kumpulan data.<\/span><\/p>\n

Contoh langkah demi langkah berikut menunjukkan cara melakukan regresi sepotong-sepotong di R.<\/span><\/p>\n

Langkah 1: Buat datanya<\/strong><\/span><\/h3>\n

Pertama, mari buat bingkai data berikut:<\/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
 Langkah 2: Visualisasikan datanya<\/strong><\/span><\/h3>\n
 Selanjutnya, mari buat diagram sebar untuk memvisualisasikan data:<\/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
 Kita dapat melihat bahwa hubungan antara x dan y nampaknya berubah secara tiba-tiba di sekitar x = 9<\/strong> .<\/span><\/p>\n
 Langkah 3: Sesuaikan model regresi sepotong-sepotong<\/strong><\/span><\/h3>\n
 Kita dapat menggunakan fungsi segmented()<\/strong> dari paket tersegmentasi<\/a> di R untuk menyesuaikan model regresi sepotong demi sepotong ke kumpulan data kita:<\/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
 Fungsi segmented()<\/strong> mendeteksi breakpoint pada x = 8.762.<\/span><\/p>\n
 Model regresi sepotong-sepotong yang dipasang adalah:<\/span><\/p>\n
 Jika x \u2264 8,762: y = 0,32143 + 1,59524*(x)<\/span><\/p>\n
 Jika x > 8,762: y = 0,32143 + 1,59524*(8,762) + (1,59524+2,40476)*(x-8,762)<\/span><\/p>\n
 Misalnya kita mempunyai nilai x = 5<\/strong> . Perkiraan nilainya adalah:<\/span><\/p>\n