Analisis Regresi Logistik untuk Memprediksi Jawaban Responden Exit Poll pada Pilkada Jawa Tengah 2024
Regresi Logistik
Library
## Loading required package: ggplot2## Loading required package: lattice## Warning: package 'corrplot' was built under R version 4.4.3## corrplot 0.95 loaded## Loading required package: reshape## Loading required package: MASS## Loading required package: viridisLite## Loading required package: NLP##
## Attaching package: 'NLP'## The following object is masked from 'package:ggplot2':
##
## annotate## Loading required package: RColorBrewer## Type 'citation("pROC")' for a citation.##
## Attaching package: 'pROC'## The following objects are masked from 'package:stats':
##
## cov, smooth, var##
## Attaching package: 'dplyr'## The following object is masked from 'package:MASS':
##
## select## The following object is masked from 'package:reshape':
##
## rename## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionData
Peubah Dummy
datum$Pilihan <- as.factor(datum$Pilihan)
datum$Gender <- as.factor(datum$Gender)
datum$Agama <- as.factor(datum$Agama)
datum$Suku <- as.factor(datum$Suku)
datum$Pendidikan <- as.factor(datum$Pendidikan)
datum$Partai <- as.factor(datum$Partai)
head(datum,10)## # A tibble: 10 × 8
## Pilihan Gender Agama Suku Pendidikan Partai Like_1 Like_2
## <fct> <fct> <fct> <fct> <fct> <fct> <dbl> <dbl>
## 1 Ahmad Taj Laki-Laki Islam Jawa Di bawahnya PDIP/Hanura 4 4
## 2 Ahmad Taj Perempuan Islam Jawa SLTA/S1/S2 PDIP/Hanura 4 2
## 3 Ahmad Taj Laki-Laki Kristen Jawa Di bawahnya Lainnya 4 2
## 4 Ahmad Taj Perempuan Islam Jawa SLTA/S1/S2 Lainnya 4 2
## 5 Ahmad Taj Laki-Laki Islam Jawa SLTA/S1/S2 Lainnya 4 2
## 6 Ahmad Taj Laki-Laki Islam Jawa SLTA/S1/S2 Lainnya 4 4
## 7 Andika Hendi Perempuan Islam Jawa Di bawahnya PDIP/Hanura 2 12
## 8 Andika Hendi Perempuan Islam Jawa Di bawahnya PDIP/Hanura 3 6
## 9 Ahmad Taj Laki-Laki Islam Jawa SLTA/S1/S2 Lainnya 4 4
## 10 Andika Hendi Laki-Laki Islam Jawa SLTA/S1/S2 PDIP/Hanura 2 5Splitting Data
Pemodelan
##
## Call:
## glm(formula = Pilihan ~ ., family = binomial(link = "logit"),
## data = trl)
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.84267 0.22569 -3.734 0.000189 ***
## GenderPerempuan -0.45719 0.15701 -2.912 0.003594 **
## AgamaKristen 0.45185 0.57708 0.783 0.433625
## AgamaLainnya -7.30486 535.41133 -0.014 0.989114
## SukuLainnya -5.73418 42.54105 -0.135 0.892777
## SukuSunda 1.13895 0.59418 1.917 0.055255 .
## PendidikanSLTA/S1/S2 0.23510 0.15872 1.481 0.138543
## PartaiPDIP/Hanura 1.16324 0.17831 6.524 6.85e-11 ***
## Like_1 -1.12194 0.08098 -13.855 < 2e-16 ***
## Like_2 1.21751 0.07942 15.330 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 1864.6 on 1386 degrees of freedom
## Residual deviance: 1017.8 on 1377 degrees of freedom
## AIC: 1037.8
##
## Number of Fisher Scoring iterations: 12# Membuat model kosong (hanya intercept)
model_kosong <- glm(Pilihan ~ 1, data = trl, family = binomial(link = "logit"))
# Membandingkan model penuh dan model kosong
anova_g <- anova(model_kosong, logmod, test = "Chisq")
print(anova_g)## Analysis of Deviance Table
##
## Model 1: Pilihan ~ 1
## Model 2: Pilihan ~ Gender + Agama + Suku + Pendidikan + Partai + Like_1 +
## Like_2
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 1386 1864.6
## 2 1377 1017.8 9 846.8 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Menampilkan p-value
p_value <- anova_g$`Pr(>Chi)`[2]
if(p_value < 0.05){
cat("Hasil: Uji G menunjukkan model regresi signifikan secara simultan (p-value =", round(p_value, 4), ").\n")
} else {
cat("Hasil: Model regresi tidak signifikan secara simultan (p-value =", round(p_value, 4), ").\n")
}## Hasil: Uji G menunjukkan model regresi signifikan secara simultan (p-value = 0 ).Uji Simultan
## $Models
##
## Model: "glm, Pilihan ~ ., binomial(link = \"logit\"), trl"
## Null: "glm, Pilihan ~ 1, binomial(link = \"logit\"), trl"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.454137
## Cox and Snell (ML) 0.456936
## Nagelkerke (Cragg and Uhler) 0.618069
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -9 -423.4 846.8 1.7814e-176
##
## $Number.of.observations
##
## Model: 1387
## Null: 1387
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"p-value < 0.05 maka ada peubah yg berpengaruh
Uji Parsial
## Loading required package: carData##
## Attaching package: 'car'## The following object is masked from 'package:dplyr':
##
## recode## Analysis of Deviance Table (Type II tests)
##
## Response: Pilihan
## Df Chisq Pr(>Chisq)
## Gender 1 8.4784 0.003594 **
## Agama 2 0.6133 0.735915
## Suku 2 3.6922 0.157855
## Pendidikan 1 2.1941 0.138543
## Partai 1 42.5608 6.852e-11 ***
## Like_1 1 191.9527 < 2.2e-16 ***
## Like_2 1 235.0165 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Untuk uji simultan didapatkan p-value untuk peubah INCUMBENT, T5, INCUMBENT PE > 0,05. Artinya peubah tersebut tidak berpengaruh signifikan terhadap model. Peubah PCTKURSI berpengaruh signifikan
Prediksi
## 1 2 3 4 5 6
## 9.234056e-01 2.854538e-01 7.520201e-04 9.437585e-01 9.234056e-01 1.911139e-01
## 7 8 9 10 11 12
## 2.409062e-03 1.905308e-03 1.828648e-01 3.868961e-01 9.963771e-01 7.864321e-03
## 13 14 15 16 17 18
## 7.361012e-03 9.781749e-01 7.186874e-01 3.147422e-01 5.279173e-05 6.296291e-08
## 19 20 21 22 23 24
## 5.611110e-01 3.381061e-02 3.868961e-01 9.333538e-01 9.954213e-01 3.868961e-01
## 25 26 27 28 29 30
## 3.868961e-01 1.828648e-01 6.296291e-08 4.204680e-01 2.854538e-01 5.238162e-02
## 31 32 33 34 35 36
## 9.996816e-01 2.854538e-01 4.439171e-01 1.007114e-01 4.439171e-01 8.869438e-01
## 37 38 39 40 41 42
## 6.535717e-02 9.878637e-01 9.965678e-01 1.421105e-02 8.453100e-01 2.450186e-04
## 43 44 45 46 47 48
## 1.007114e-01 2.854538e-01 9.996065e-01 4.439171e-01 5.509151e-01 7.969935e-01
## 49 50 51 52 53 54
## 8.675039e-01 3.868961e-01 9.975877e-01 6.896394e-01 8.093265e-03 6.596102e-01
## 55 56 57 58 59 60
## 3.741047e-01 6.179300e-01 9.878637e-01 4.148064e-01 5.509151e-01 9.860718e-01
## 61 62 63 64 65 66
## 2.226477e-02 1.704679e-01 6.591933e-05 2.854538e-01 4.439171e-01 4.204680e-01
## 67 68 69 70 71 72
## 9.963771e-01 3.741047e-01 9.860718e-01 2.854538e-01 1.007114e-01 4.439171e-01
## 73 74 75 76 77 78
## 9.860718e-01 1.911139e-01 2.481109e-01 2.854538e-01 5.238162e-02 6.272300e-02
## 79 80 81 82 83 84
## 1.768262e-02 2.063259e-01 5.269046e-01 5.827814e-03 2.318187e-02 1.895453e-05
## 85 86 87 88 89 90
## 1.768262e-02 7.295280e-01 3.357106e-01 2.682870e-02 5.238162e-02 6.179300e-01
## 91 92 93 94 95 96
## 9.584285e-01 6.535717e-02 6.252086e-01 6.688219e-01 3.357106e-01 8.186362e-01
## 97 98 99 100 101 102
## 3.357106e-01 2.444310e-02 2.854538e-01 5.827814e-03 5.827814e-03 9.999189e-01
## 103 104 105 106 107 108
## 6.535717e-02 2.063259e-01 5.827814e-03 8.493132e-01 1.573754e-01 2.224159e-01
## 109 110 111 112 113 114
## 7.603710e-01 1.480144e-02 5.238162e-02 1.768262e-02 3.644848e-01 4.439171e-01
## 115 116 117 118 119 120
## 1.327909e-03 1.075437e-06 6.179300e-01 3.357106e-01 1.573754e-01 1.151173e-01
## 121 122 123 124 125 126
## 3.933845e-04 8.721797e-01 4.305644e-01 5.851363e-01 1.165911e-02 3.973725e-01
## 127 128 129 130 131 132
## 5.238162e-02 2.854538e-01 7.857845e-04 1.704679e-01 3.868961e-01 3.644848e-01
## 133 134 135 136 137 138
## 2.745146e-01 2.663658e-01 7.361012e-03 6.081322e-01 9.654996e-01 2.854538e-01
## 139 140 141 142 143 144
## 7.102619e-01 1.767736e-01 1.573754e-01 4.439171e-01 4.425715e-06 1.561401e-02
## 145 146 147 148 149 150
## 1.911139e-01 6.306579e-01 6.535717e-02 3.707572e-02 3.381061e-02 6.081322e-01
## 151 152 153 154 155 156
## 1.126728e-02 6.688219e-01 2.715695e-05 9.486214e-01 7.295280e-01 9.584285e-01
## 157 158 159 160 161 162
## 3.868961e-01 4.239170e-02 6.596102e-01 5.827814e-03 2.095989e-03 8.453100e-01
## 163 164 165 166 167 168
## 9.954213e-01 3.868961e-01 3.868961e-01 8.922724e-01 3.868961e-01 1.151173e-01
## 169 170 171 172 173 174
## 1.301110e-01 9.967062e-01 7.186874e-01 6.807304e-01 4.644565e-02 1.238327e-02
## 175 176 177 178 179 180
## 4.239170e-02 9.912096e-04 1.451126e-01 4.439171e-01 9.358776e-01 6.807304e-01
## 181 182 183 184 185 186
## 2.854538e-01 6.535717e-02 6.596102e-01 1.911139e-01 3.520089e-01 9.878637e-01
## 187 188 189 190 191 192
## 4.672588e-03 9.011206e-01 8.093265e-03 9.999472e-01 5.509151e-01 7.969935e-01
## 193 194 195 196 197 198
## 3.868961e-01 8.323964e-01 4.239170e-02 5.611110e-01 6.688219e-01 2.455721e-07
## 199 200 201 202 203 204
## 1.057341e-01 4.439171e-01 8.522740e-01 1.238327e-02 2.854538e-01 6.591933e-05
## 205 206 207 208 209 210
## 3.381061e-02 7.295280e-01 6.023404e-04 6.473220e-01 3.697267e-03 1.193921e-01
## 211 212 213 214 215 216
## 3.357106e-01 1.841454e-02 9.271568e-01 1.561401e-02 1.238327e-02 1.557333e-02
## 217 218 219 220 221 222
## 5.202452e-03 6.081322e-01 3.381061e-02 9.742613e-01 3.868961e-01 3.381061e-02
## 223 224 225 226 227 228
## 1.942155e-02 1.413157e-01 2.224159e-01 8.922724e-01 2.854538e-01 1.126728e-02
## 229 230 231 232 233 234
## 8.721797e-01 4.439171e-01 9.384652e-01 5.086670e-02 1.451126e-01 2.558961e-01
## 235 236 237 238 239 240
## 5.611110e-01 3.868961e-01 8.120188e-01 9.234056e-01 9.954213e-01 5.570765e-01
## 241 242 243 244 245 246
## 4.439171e-01 9.584285e-01 5.611110e-01 9.584285e-01 9.929438e-01 9.854930e-01
## 247 248 249 250 251 252
## 3.856312e-01 7.144348e-02 1.057341e-01 7.186874e-01 6.807304e-01 9.011206e-01
## 253 254 255 256 257 258
## 3.707572e-02 1.301110e-01 2.224159e-01 5.238162e-02 5.238162e-02 5.611110e-01
## 259 260 261 262 263 264
## 8.975357e-07 7.144348e-02 4.239170e-02 9.953093e-01 6.688219e-01 4.239170e-02
## 265 266 267 268 269 270
## 3.357106e-01 9.274441e-01 4.439171e-01 6.688219e-01 7.361012e-03 2.945029e-01
## 271 272 273 274 275 276
## 3.868961e-01 3.868961e-01 9.703177e-02 1.151173e-01 3.868961e-01 3.933845e-04
## 277 278 279 280 281 282
## 6.272300e-02 3.520089e-01 3.381061e-02 2.245124e-06 7.361012e-03 4.426004e-01
## 283 284 285 286 287 288
## 7.902288e-01 3.933845e-04 2.854538e-01 1.841454e-02 2.854538e-01 1.573754e-01
## 289 290 291 292 293 294
## 1.767736e-01 8.265547e-01 1.911139e-01 1.335418e-04 6.688219e-01 6.378362e-01
## 295 296 297 298 299 300
## 4.672588e-03 4.644565e-02 5.238162e-02 5.238162e-02 4.978687e-01 3.357106e-01
## 301 302 303 304 305 306
## 6.989678e-01 5.238162e-02 6.179300e-01 4.439171e-01 3.357106e-01 2.854538e-01
## 307 308 309 310 311 312
## 5.611110e-01 4.228426e-02 9.873253e-01 3.967382e-01 5.271722e-01 9.703177e-02
## 313 314 315 316 317 318
## 3.357106e-01 5.611110e-01 4.439171e-01 1.734115e-01 7.187854e-07 9.358776e-01
## 319 320 321 322 323 324
## 1.301110e-01 1.837277e-09 7.295280e-01 9.970967e-01 8.827354e-01 5.733372e-02
## 325 326 327 328 329 330
## 3.357106e-01 8.975357e-07 6.596102e-01 4.064289e-02 3.357106e-01 9.954213e-01
## 331 332 333 334 335 336
## 8.522740e-01 5.744282e-01 9.501089e-01 9.999999e-01 3.037566e-03 2.444310e-02
## 337 338 339 340 341 342
## 6.596102e-01 9.298976e-01 6.688219e-01 8.398317e-01 5.744282e-01 9.526194e-01
## 343 344 345 346
## 6.179300e-01 4.672588e-03 9.584285e-01 4.439171e-01## Warning: package 'ResourceSelection' was built under R version 4.4.3## ResourceSelection 0.3-6 2023-06-27# Konversi label aktual ke numerik
testrl$Pilihan_numeric <- as.numeric(testrl$Pilihan) - 1
# Uji Hosmer-Lemeshow
hoslem.test(testrl$Pilihan_numeric, predi, g = 10)##
## Hosmer and Lemeshow goodness of fit (GOF) test
##
## data: testrl$Pilihan_numeric, predi
## X-squared = 143.16, df = 8, p-value < 2.2e-16## 1 2 3 4 5
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Andika Hendi"
## 6 7 8 9 10
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 11 12 13 14 15
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Andika Hendi"
## 16 17 18 19 20
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 21 22 23 24 25
## "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 26 27 28 29 30
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 31 32 33 34 35
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 36 37 38 39 40
## "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Ahmad Taj"
## 41 42 43 44 45
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 46 47 48 49 50
## "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Andika Hendi" "Ahmad Taj"
## 51 52 53 54 55
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 56 57 58 59 60
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Andika Hendi"
## 61 62 63 64 65
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 66 67 68 69 70
## "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 71 72 73 74 75
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 76 77 78 79 80
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 81 82 83 84 85
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 86 87 88 89 90
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 91 92 93 94 95
## "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Ahmad Taj"
## 96 97 98 99 100
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 101 102 103 104 105
## "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 106 107 108 109 110
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 111 112 113 114 115
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 116 117 118 119 120
## "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 121 122 123 124 125
## "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 126 127 128 129 130
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 131 132 133 134 135
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 136 137 138 139 140
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 141 142 143 144 145
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 146 147 148 149 150
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 151 152 153 154 155
## "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Andika Hendi"
## 156 157 158 159 160
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 161 162 163 164 165
## "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 166 167 168 169 170
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 171 172 173 174 175
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 176 177 178 179 180
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Andika Hendi"
## 181 182 183 184 185
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 186 187 188 189 190
## "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi"
## 191 192 193 194 195
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 196 197 198 199 200
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 201 202 203 204 205
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 206 207 208 209 210
## "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 211 212 213 214 215
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 216 217 218 219 220
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi"
## 221 222 223 224 225
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 226 227 228 229 230
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 231 232 233 234 235
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 236 237 238 239 240
## "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Andika Hendi" "Andika Hendi"
## 241 242 243 244 245
## "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Andika Hendi" "Andika Hendi"
## 246 247 248 249 250
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 251 252 253 254 255
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 256 257 258 259 260
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 261 262 263 264 265
## "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 266 267 268 269 270
## "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 271 272 273 274 275
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 276 277 278 279 280
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 281 282 283 284 285
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 286 287 288 289 290
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 291 292 293 294 295
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Ahmad Taj"
## 296 297 298 299 300
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 301 302 303 304 305
## "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 306 307 308 309 310
## "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 311 312 313 314 315
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 316 317 318 319 320
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 321 322 323 324 325
## "Andika Hendi" "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 326 327 328 329 330
## "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 331 332 333 334 335
## "Andika Hendi" "Andika Hendi" "Andika Hendi" "Andika Hendi" "Ahmad Taj"
## 336 337 338 339 340
## "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Andika Hendi" "Andika Hendi"
## 341 342 343 344 345
## "Andika Hendi" "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Andika Hendi"
## 346
## "Ahmad Taj"## Confusion Matrix and Statistics
##
## Reference
## Prediction Ahmad Taj Andika Hendi
## Ahmad Taj 181 42
## Andika Hendi 25 98
##
## Accuracy : 0.8064
## 95% CI : (0.7607, 0.8467)
## No Information Rate : 0.5954
## P-Value [Acc > NIR] : < 2e-16
##
## Kappa : 0.5901
##
## Mcnemar's Test P-Value : 0.05062
##
## Sensitivity : 0.8786
## Specificity : 0.7000
## Pos Pred Value : 0.8117
## Neg Pred Value : 0.7967
## Prevalence : 0.5954
## Detection Rate : 0.5231
## Detection Prevalence : 0.6445
## Balanced Accuracy : 0.7893
##
## 'Positive' Class : Ahmad Taj
## ROC
## Setting levels: control = Ahmad Taj, case = Andika Hendi## Setting direction: controls < casesauc <- auc(roc_curve)
plot(roc_curve, col = "lightblue", main = paste("Kurva ROC (AUC =", round(auc, 2), ")", sep = ""), col.main = "pink", lwd = 2)
legend("bottomright", legend = paste("AUC =", round(auc, 2)), col = "darkred", lty = 1, cex = 0.8)
# Membaca data baru dari sheet ke-3
data_baru <- readxl::read_xlsx("C:/Users/ACER/Documents/Magang/Laporan Individu Magang/Untitled2.xlsx", sheet=3)
# Menyeleksi peubah sesuai dengan model
data_baru <- data_baru %>% select('Gender', 'Agama', 'Suku', 'Pendidikan', 'Partai', 'Like_1', 'Like_2')# Pastikan peubah memiliki tipe data yang sama dengan model
data_baru$Gender <- as.factor(data_baru$Gender)
data_baru$Agama <- as.factor(data_baru$Agama)
data_baru$Suku <- as.factor(data_baru$Suku)
data_baru$Pendidikan <- as.factor(data_baru$Pendidikan)
data_baru$Partai <- as.factor(data_baru$Partai)# Prediksi probabilitas
prediksi_prob <- predict(logmod, newdata = data_baru, type = "response")
# Prediksi Pilihan berdasarkan probabilitas (>0.5 untuk 'Andika Hendi', <=0.5 untuk 'Ahmad Taj')
prediksi_pilihan <- ifelse(prediksi_prob > 0.5, 'Andika Hendi', 'Ahmad Taj')
# Menampilkan hasil prediksi
prediksi_pilihan## 1 2 3 4 5
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi"
## 6 7 8 9 10
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 11 12 13 14 15
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 16 17 18 19 20
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 21 22 23 24 25
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 26 27 28 29 30
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 31 32 33 34 35
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 36 37 38 39 40
## "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Andika Hendi" "Ahmad Taj"
## 41 42 43 44 45
## "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Ahmad Taj"
## 46 47 48 49 50
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Andika Hendi" "Andika Hendi"
## 51 52 53 54 55
## "Andika Hendi" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 56 57 58 59 60
## "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 61 62 63 64 65
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 66 67 68 69 70
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 71 72 73 74 75
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 76 77 78 79 80
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 81 82 83 84 85
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi"
## 86 87 88 89 90
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 91 92 93 94 95
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 96 97 98 99 100
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 101 102 103 104 105
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj"
## 106 107 108 109 110
## "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi"
## 111 112 113 114 115
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi"
## 116 117 118 119 120
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 121 122 123 124 125
## "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj" "Ahmad Taj"
## 126 127 128 129 130
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 131 132 133 134 135
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Andika Hendi"
## 136 137 138 139 140
## "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 141 142 143 144 145
## "Andika Hendi" "Andika Hendi" "Andika Hendi" "Ahmad Taj" "Ahmad Taj"
## 146 147 148 149 150
## "Andika Hendi" "Ahmad Taj" "Ahmad Taj" "Andika Hendi" "Andika Hendi"
## 151
## "Ahmad Taj"## Warning: package 'openxlsx' was built under R version 4.4.3