# Baca data
data <- read.csv("bmkg_cuaca.csv")
head(data)
## Tanggal Tn Tx Tavg RH_avg RR ss ff_x ddd_x ff_avg ddd_car
## 1 01-01-2024 26.0 35.0 29.9 83 3.8 6.5 3 60 1 C
## 2 02-01-2024 26.2 35.2 30.6 78 21.2 6.7 4 160 2 C
## 3 03-01-2024 27.0 35.8 27.8 90 0.0 8.9 4 60 1 C
## 4 04-01-2024 25.8 34.4 29.0 82 1.5 1.9 4 70 2 W
## 5 05-01-2024 25.5 33.4 28.8 83 2.2 6.0 5 110 2 W
## 6 06-01-2024 25.4 31.3 28.0 87 0.0 3.8 3 90 1 C
## Kategori_Cuaca
## 1 Hujan Sedang
## 2 Hujan Lebat
## 3 Cerah Berawan
## 4 Gerimis
## 5 Gerimis
## 6 Berawan
#-------------PREPROCESSING---------
# Cek jumlah NA per kolom
colSums(is.na(data))
## Tanggal Tn Tx Tavg RH_avg
## 0 12 3 1 1
## RR ss ff_x ddd_x ff_avg
## 48 4 0 0 0
## ddd_car Kategori_Cuaca
## 0 0
# Gantilah nilai 8888 menjadi NA, kemudian hapus baris yang mengandung NA
data[data == 8888] <- NA
data <- na.omit(data)
# di kolom ddd_car ternyata ada yang Ganda karena Spasi
data$ddd_car <- trimws(data$ddd_car)
# Identifikasi kolom numerik (selain Tanggal dan Kategori_Cuaca)
numeric_cols <- names(data)[sapply(data, is.numeric) & !(names(data) %in% c("Tanggal"))]
# Hapus baris jika semua kolom numeriknya NA
data <- data[!apply(data[numeric_cols], 1, function(x) all(is.na(x))), ]
# Fungsi untuk mengisi NA dengan interpolasi linear antar baris
interpolate_na <- function(x) {
approx(1:length(x), x, 1:length(x), rule = 2)$y
}
# Terapkan interpolasi untuk setiap kolom numerik
data[numeric_cols] <- lapply(data[numeric_cols], interpolate_na)
# Tampilkan data hasil imputasi
head(data)
## Tanggal Tn Tx Tavg RH_avg RR ss ff_x ddd_x ff_avg ddd_car
## 1 01-01-2024 26.0 35.0 29.9 83 3.8 6.5 3 60 1 C
## 2 02-01-2024 26.2 35.2 30.6 78 21.2 6.7 4 160 2 C
## 3 03-01-2024 27.0 35.8 27.8 90 0.0 8.9 4 60 1 C
## 4 04-01-2024 25.8 34.4 29.0 82 1.5 1.9 4 70 2 W
## 5 05-01-2024 25.5 33.4 28.8 83 2.2 6.0 5 110 2 W
## 6 06-01-2024 25.4 31.3 28.0 87 0.0 3.8 3 90 1 C
## Kategori_Cuaca
## 1 Hujan Sedang
## 2 Hujan Lebat
## 3 Cerah Berawan
## 4 Gerimis
## 5 Gerimis
## 6 Berawan
# Cek jumlah NA per kolom
colSums(is.na(data))
## Tanggal Tn Tx Tavg RH_avg
## 0 0 0 0 0
## RR ss ff_x ddd_x ff_avg
## 0 0 0 0 0
## ddd_car Kategori_Cuaca
## 0 0
# Hitung matriks korelasi untuk kolom numerik saja
cor_matrix <- cor(data[numeric_cols], use = "complete.obs")
# Tampilkan matriks korelasi (bisa di-plot juga nanti)
print(round(cor_matrix, 2))
## Tn Tx Tavg RH_avg RR ss ff_x ddd_x ff_avg
## Tn 1.00 0.32 0.54 0.16 -0.14 -0.01 -0.15 0.10 -0.13
## Tx 0.32 1.00 0.74 -0.45 -0.05 0.25 0.11 -0.07 0.06
## Tavg 0.54 0.74 1.00 -0.51 -0.17 0.28 0.01 -0.23 0.12
## RH_avg 0.16 -0.45 -0.51 1.00 0.37 -0.50 -0.28 0.39 -0.46
## RR -0.14 -0.05 -0.17 0.37 1.00 -0.33 0.02 0.18 -0.24
## ss -0.01 0.25 0.28 -0.50 -0.33 1.00 0.13 -0.27 0.22
## ff_x -0.15 0.11 0.01 -0.28 0.02 0.13 1.00 0.19 0.62
## ddd_x 0.10 -0.07 -0.23 0.39 0.18 -0.27 0.19 1.00 -0.03
## ff_avg -0.13 0.06 0.12 -0.46 -0.24 0.22 0.62 -0.03 1.00
# Visualisasi korelasi menggunakan heatmap (opsional, butuh package corrplot)
if(!require(corrplot)) install.packages("corrplot")
## Loading required package: corrplot
## corrplot 0.95 loaded
library(corrplot)
corrplot(cor_matrix, method = "color", type = "upper", tl.cex=0.7, tl.col="black", diag=FALSE)
#hapus Tn, Tx karena sudah diwakili Tavg. Hapus ff_x karena sudah diwakili ff_avg (redudant)
columns_to_drop <- c("Tn", "Tx", "ff_x", "Tanggal")
data <- data[, !(names(data) %in% columns_to_drop)]
head(data)
## Tavg RH_avg RR ss ddd_x ff_avg ddd_car Kategori_Cuaca
## 1 29.9 83 3.8 6.5 60 1 C Hujan Sedang
## 2 30.6 78 21.2 6.7 160 2 C Hujan Lebat
## 3 27.8 90 0.0 8.9 60 1 C Cerah Berawan
## 4 29.0 82 1.5 1.9 70 2 W Gerimis
## 5 28.8 83 2.2 6.0 110 2 W Gerimis
## 6 28.0 87 0.0 3.8 90 1 C Berawan
library(tidyverse)
## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.1 ✔ tibble 3.2.1
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.0.4
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
# Ubah nilai ddd_car ke factor dengan label deskriptif
data <- data %>%
mutate(
ddd_car = factor(ddd_car, levels = c("C", "E", "NE", "NW", "S", "SW", "W"),
labels = c("Tenang (Calm)", "Timur", "Timur Laut", "Barat Laut",
"Selatan", "Barat Daya", "Barat"))
)
write.csv(data, "data_bmkg_preprocessed.csv", row.names = FALSE)
data <- read.csv("data_bmkg_preprocessed.csv")
#EDA
# Pastikan packages terinstal dan diload
packages <- c("tidyverse", "GGally", "corrplot", "gridExtra")
new.packages <- packages[!(packages %in% installed.packages())]
if(length(new.packages)) install.packages(new.packages)
lapply(packages, library, character.only = TRUE)
## Registered S3 method overwritten by 'GGally':
## method from
## +.gg ggplot2
##
## Attaching package: 'gridExtra'
##
## The following object is masked from 'package:dplyr':
##
## combine
## [[1]]
## [1] "lubridate" "forcats" "stringr" "dplyr" "purrr" "readr"
## [7] "tidyr" "tibble" "ggplot2" "tidyverse" "corrplot" "stats"
## [13] "graphics" "grDevices" "utils" "datasets" "methods" "base"
##
## [[2]]
## [1] "GGally" "lubridate" "forcats" "stringr" "dplyr" "purrr"
## [7] "readr" "tidyr" "tibble" "ggplot2" "tidyverse" "corrplot"
## [13] "stats" "graphics" "grDevices" "utils" "datasets" "methods"
## [19] "base"
##
## [[3]]
## [1] "GGally" "lubridate" "forcats" "stringr" "dplyr" "purrr"
## [7] "readr" "tidyr" "tibble" "ggplot2" "tidyverse" "corrplot"
## [13] "stats" "graphics" "grDevices" "utils" "datasets" "methods"
## [19] "base"
##
## [[4]]
## [1] "gridExtra" "GGally" "lubridate" "forcats" "stringr" "dplyr"
## [7] "purrr" "readr" "tidyr" "tibble" "ggplot2" "tidyverse"
## [13] "corrplot" "stats" "graphics" "grDevices" "utils" "datasets"
## [19] "methods" "base"
# Gunakan data hasil preprocessing
data_clean <- data # data hasil preprocessing dan imputasi NA
# Cek struktur data
str(data_clean)
## 'data.frame': 217 obs. of 8 variables:
## $ Tavg : num 29.9 30.6 27.8 29 28.8 28 27.7 27.7 28.7 28.5 ...
## $ RH_avg : int 83 78 90 82 83 87 88 89 86 83 ...
## $ RR : num 3.8 21.2 0 1.5 2.2 0 2.1 78.1 26.5 3.3 ...
## $ ss : num 6.5 6.7 8.9 1.9 6 3.8 2 0.6 4.5 5 ...
## $ ddd_x : int 60 160 60 70 110 90 340 200 230 190 ...
## $ ff_avg : int 1 2 1 2 2 1 1 2 2 1 ...
## $ ddd_car : chr "Tenang (Calm)" "Tenang (Calm)" "Tenang (Calm)" "Barat" ...
## $ Kategori_Cuaca: chr "Hujan Sedang" "Hujan Lebat" "Cerah Berawan" "Gerimis" ...
# Ringkasan statistik deskriptif
summary(data_clean)
## Tavg RH_avg RR ss
## Min. :25.90 Min. :53.00 Min. : 0.000 Min. : 0.000
## 1st Qu.:28.00 1st Qu.:72.00 1st Qu.: 0.000 1st Qu.: 5.200
## Median :28.90 Median :78.00 Median : 0.000 Median : 7.900
## Mean :28.84 Mean :77.02 Mean : 6.572 Mean : 6.469
## 3rd Qu.:29.70 3rd Qu.:82.00 3rd Qu.: 4.500 3rd Qu.: 8.000
## Max. :31.00 Max. :96.00 Max. :84.500 Max. :10.000
## ddd_x ff_avg ddd_car Kategori_Cuaca
## Min. : 30.0 Min. :1.000 Length:217 Length:217
## 1st Qu.: 80.0 1st Qu.:2.000 Class :character Class :character
## Median :100.0 Median :2.000 Mode :character Mode :character
## Mean :128.5 Mean :2.442
## 3rd Qu.:150.0 3rd Qu.:3.000
## Max. :340.0 Max. :4.000
# Jumlah observasi per kategori cuaca
table(data_clean$Kategori_Cuaca)
##
## Berawan Cerah Cerah Berawan Gerimis Hujan Lebat
## 6 60 60 28 24
## Hujan Sedang
## 39
# Visualisasi distribusi kategori cuaca
ggplot(data_clean, aes(x = Kategori_Cuaca, fill = Kategori_Cuaca)) +
geom_bar() +
labs(title = "Distribusi Kategori Cuaca", x = "Kategori", y = "Jumlah") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
# Boxplot suhu rata-rata per kategori
ggplot(data_clean, aes(x = Kategori_Cuaca, y = Tavg, fill = Kategori_Cuaca)) +
geom_boxplot() +
labs(title = "Boxplot Suhu Rata-rata per Kategori Cuaca") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
# Korelasi antar variabel numerik
numeric_only <- data_clean[, sapply(data_clean, is.numeric)]
cor_matrix <- cor(numeric_only, use = "complete.obs")
corrplot(cor_matrix, method = "color", type = "upper",
tl.cex = 0.7, tl.col = "black", diag = FALSE)
#-------------UJI ASUMSI---------
#Uji linearitas dengan scatterplott matrix
# Filter hanya kolom numerik
numeric_data <- data[sapply(data, is.numeric)]
# Plot scatterplot matrix
pairs(numeric_data, main = "Scatterplot Matrix untuk Uji Linearitas", pch = 21, bg = "lightblue")
#Berdasarkan hasil scatterplot matrix, diketahui bahwa beberapa pasangan variabel seperti Tavg dengan RH_avg atau ss memperlihatkan pola sebaran titik yang cukup linear. Namun, sebagian besar pasangan variabel lainnya menunjukkan hubungan non-linear atau tidak membentuk pola linear yang jelas. Hal ini menunjukkan bahwa asumsi linearitas hanya terpenuhi sebagian. Oleh karena itu, jika model linier akan digunakan, perlu dilakukan transformasi data atau pemilihan fitur yang tepat.
#Uji Outlier (dengan Mahalanobis Distance)
# Pilih hanya kolom numerik untuk analisis Mahalanobis
numeric_data <- data[sapply(data, is.numeric)]
# Hitung Mahalanobis distance
mahal_dist <- mahalanobis(numeric_data, colMeans(numeric_data), cov(numeric_data))
# Hitung batas cutoff outlier (dengan derajat bebas = jumlah kolom)
cutoff <- qchisq(0.975, df = ncol(numeric_data))
# Identifikasi baris yang merupakan outlier
outliers <- which(mahal_dist > cutoff)
# Tampilkan jumlah dan indeks outlier
length(outliers)
## [1] 10
print(outliers)
## [1] 8 24 31 37 47 54 58 76 94 162
# Visualisasi
plot(mahal_dist, type = "h", main = "Mahalanobis Distance", ylab = "Distance")
abline(h = cutoff, col = "red", lty = 2)
#output menunjukkan terdapat 12 observasi (baris data) yang terindikasi sebagai outlier multivariat yang nantinya akan dihapus karena pada multinomial regression outliers sangat mempengaruhi model
data_clean <- data[-outliers, ]
# Cek hasilnya
nrow(data) # jumlah baris sebelum
## [1] 217
nrow(data_clean) # jumlah baris sesudah
## [1] 207
head(data_clean)
## Tavg RH_avg RR ss ddd_x ff_avg ddd_car Kategori_Cuaca
## 1 29.9 83 3.8 6.5 60 1 Tenang (Calm) Hujan Sedang
## 2 30.6 78 21.2 6.7 160 2 Tenang (Calm) Hujan Lebat
## 3 27.8 90 0.0 8.9 60 1 Tenang (Calm) Cerah Berawan
## 4 29.0 82 1.5 1.9 70 2 Barat Gerimis
## 5 28.8 83 2.2 6.0 110 2 Barat Gerimis
## 6 28.0 87 0.0 3.8 90 1 Tenang (Calm) Berawan
# Uji independensi (autokorelasi)
# Pastikan package ini terinstal
if(!require(lmtest)) install.packages("lmtest")
## Loading required package: lmtest
## Loading required package: zoo
##
## Attaching package: 'zoo'
##
## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numeric
library(lmtest)
# Buat salinan data agar aman
data_indep <- data_clean
# Ubah Kategori_Cuaca jadi numerik (sementara)
data_indep$Kategori_Cuaca_Num <- as.numeric(as.factor(data_indep$Kategori_Cuaca))
# Buat model regresi linier (dummy)
model_dw <- lm(Kategori_Cuaca_Num ~ ., data = data_indep[, sapply(data_indep, is.numeric)])
# Uji Durbin-Watson
dw_result <- dwtest(model_dw)
print(dw_result)
##
## Durbin-Watson test
##
## data: model_dw
## DW = 1.946, p-value = 0.281
## alternative hypothesis: true autocorrelation is greater than 0
#Nilai DW mendekati 2 → Tidak ada autokorelasi (✅ memenuhi asumsi independensi).
# Uji Multikolinearitas – dengan VIF (Variance Inflation Factor)
# Pastikan package car sudah terinstal
if (!require(car)) install.packages("car")
## Loading required package: car
## Loading required package: carData
##
## Attaching package: 'car'
##
## The following object is masked from 'package:dplyr':
##
## recode
##
## The following object is masked from 'package:purrr':
##
## some
library(car)
# Buat model regresi linier dengan target dummy (misalnya Tavg atau variabel target sesungguhnya)
# Bisa ganti "Tavg" dengan variabel target kamu yang sebenarnya
model_vif <- lm(Kategori_Cuaca_Num ~ ., data = data_indep[, sapply(data_indep, is.numeric)])
# Hitung nilai VIF
vif_values <- vif(model_vif)
print(vif_values)
## Tavg RH_avg RR ss ddd_x ff_avg
## 1.341802 2.298771 1.223165 1.312329 1.221417 1.398698
# VIF > 5 menunjukkan kemungkinan multikolinearitas. VIF > 10 menunjukkan multikolinearitas serius
# Asumsi tidak ada multikolinearitas terpenuhi.
#Uji Normalitas Multivariat – dengan Mardia Test
# Pastikan package MVN sudah terinstal
if (!require(MVN)) install.packages("MVN")
## Loading required package: MVN
library(MVN)
# Ambil hanya kolom numerik
numeric_data <- data[sapply(data_clean, is.numeric)]
# Lakukan Mardia Test
mardia_result <- mvn(data = numeric_data, mvnTest = "mardia")
print(mardia_result$multivariateNormality)
## Test Statistic p value Result
## 1 Mardia Skewness 639.62833983284 5.49409057142149e-100 NO
## 2 Mardia Kurtosis 7.3856030405196 1.51656465163796e-13 NO
## 3 MVN <NA> <NA> NO
#diperoleh p-value yang jauh lebih kecil dari 0.05. Hal ini menunjukkan bahwa: Data tidak memenuhi asumsi normalitas multivariat.
#Uji normalitas dengan kolmogorov spirnov
numeric_data <- data[sapply(data, is.numeric)]
# Lakukan uji Kolmogorov–Smirnov pada setiap kolom
ks_results <- lapply(numeric_data, function(col) {
# Normalisasi kolom jika perlu
col <- scale(col)
ks.test(col, "pnorm")
})
## Warning in ks.test.default(col, "pnorm"): ties should not be present for the
## one-sample Kolmogorov-Smirnov test
## Warning in ks.test.default(col, "pnorm"): ties should not be present for the
## one-sample Kolmogorov-Smirnov test
## Warning in ks.test.default(col, "pnorm"): ties should not be present for the
## one-sample Kolmogorov-Smirnov test
## Warning in ks.test.default(col, "pnorm"): ties should not be present for the
## one-sample Kolmogorov-Smirnov test
## Warning in ks.test.default(col, "pnorm"): ties should not be present for the
## one-sample Kolmogorov-Smirnov test
## Warning in ks.test.default(col, "pnorm"): ties should not be present for the
## one-sample Kolmogorov-Smirnov test
# Tampilkan hasil p-value tiap kolom
for (i in seq_along(ks_results)) {
cat("Kolom:", names(ks_results)[i], "\n")
print(ks_results[[i]]$p.value)
cat("\n")
}
## Kolom: Tavg
## [1] 0.3006132
##
## Kolom: RH_avg
## [1] 0.1609301
##
## Kolom: RR
## [1] 7.764448e-21
##
## Kolom: ss
## [1] 2.969106e-10
##
## Kolom: ddd_x
## [1] 3.667445e-16
##
## Kolom: ff_avg
## [1] 7.298252e-16
#uji homogenitas varians dengan barlet test
# Misalnya kita ingin menguji homogenitas varians dari setiap variabel numerik terhadap kategori cuaca
for (col in names(data_clean)[sapply(data_clean, is.numeric)]) {
cat("Bartlett Test untuk variabel:", col, "\n")
print(bartlett.test(data_clean[[col]] ~ data_clean$Kategori_Cuaca))
cat("\n")
}
## Bartlett Test untuk variabel: Tavg
##
## Bartlett test of homogeneity of variances
##
## data: data_clean[[col]] by data_clean$Kategori_Cuaca
## Bartlett's K-squared = 7.1928, df = 5, p-value = 0.2067
##
##
## Bartlett Test untuk variabel: RH_avg
##
## Bartlett test of homogeneity of variances
##
## data: data_clean[[col]] by data_clean$Kategori_Cuaca
## Bartlett's K-squared = 4.9165, df = 5, p-value = 0.4262
##
##
## Bartlett Test untuk variabel: RR
##
## Bartlett test of homogeneity of variances
##
## data: data_clean[[col]] by data_clean$Kategori_Cuaca
## Bartlett's K-squared = Inf, df = 5, p-value < 2.2e-16
##
##
## Bartlett Test untuk variabel: ss
##
## Bartlett test of homogeneity of variances
##
## data: data_clean[[col]] by data_clean$Kategori_Cuaca
## Bartlett's K-squared = 227.32, df = 5, p-value < 2.2e-16
##
##
## Bartlett Test untuk variabel: ddd_x
##
## Bartlett test of homogeneity of variances
##
## data: data_clean[[col]] by data_clean$Kategori_Cuaca
## Bartlett's K-squared = 107.24, df = 5, p-value < 2.2e-16
##
##
## Bartlett Test untuk variabel: ff_avg
##
## Bartlett test of homogeneity of variances
##
## data: data_clean[[col]] by data_clean$Kategori_Cuaca
## Bartlett's K-squared = 6.2578, df = 5, p-value = 0.2819
#----------------MLR---------------
library(nnet)
library(dplyr) # untuk %>% dan data manipulation
library(broom) # untuk tidy()
library(kableExtra) # untuk kable_styling()
##
## Attaching package: 'kableExtra'
##
## The following object is masked from 'package:dplyr':
##
## group_rows
library(knitr) # untuk kable()
fit_full <- multinom(Kategori_Cuaca ~ RR+Tavg+RH_avg+ss+ddd_x+ff_avg+ddd_car, data_clean)
## # weights: 84 (65 variable)
## initial value 370.894210
## iter 10 value 234.290778
## iter 20 value 194.790984
## iter 30 value 128.942975
## iter 40 value 80.217288
## iter 50 value 66.761095
## iter 60 value 56.287683
## iter 70 value 48.129583
## iter 80 value 45.881164
## iter 90 value 44.716892
## iter 100 value 42.517399
## final value 42.517399
## stopped after 100 iterations
#Interpret for the relative log odds
summary(fit_full)
## Warning in sqrt(diag(vc)): NaNs produced
## Call:
## multinom(formula = Kategori_Cuaca ~ RR + Tavg + RH_avg + ss +
## ddd_x + ff_avg + ddd_car, data = data_clean)
##
## Coefficients:
## (Intercept) RR Tavg RH_avg ss ddd_x
## Cerah 114.124932 16.813176 -6.2656644 -1.5409423 10.667590 0.07896792
## Cerah Berawan 153.639540 8.282869 -5.4308207 -0.7390146 7.718777 0.06919883
## Gerimis -315.377722 32.277232 3.8578633 1.6782064 5.361079 0.09871998
## Hujan Lebat 4.506905 48.048530 2.8020820 -3.7625872 -1.638738 0.18739507
## Hujan Sedang -118.267389 39.014629 -0.1559959 0.7540256 5.034844 0.09464518
## ff_avg ddd_carBarat Daya ddd_carBarat Laut ddd_carSelatan
## Cerah 7.1307930 125.99075 32.739433 -149.19582
## Cerah Berawan 7.1810570 15.66337 45.892377 104.53938
## Gerimis 10.1337177 30.16470 1.856387 -44.28981
## Hujan Lebat -11.8834790 53.45077 28.432923 -4.54002
## Hujan Sedang 0.4463558 34.98387 -117.649474 90.04106
## ddd_carTenang (Calm) ddd_carTimur ddd_carTimur Laut
## Cerah 96.026124 105.45896 150.61396
## Cerah Berawan -2.600540 7.99057 52.65552
## Gerimis 5.495606 14.12862 60.14670
## Hujan Lebat 1.041417 -25.85826 118.42329
## Hujan Sedang -4.208387 15.22438 63.03140
##
## Std. Errors:
## (Intercept) RR Tavg RH_avg ss ddd_x
## Cerah 6.94147906 3.724792 1.211786 0.6885924 2.324110 0.10420114
## Cerah Berawan 6.82656632 3.897549 1.214907 0.6710433 2.197607 0.10240229
## Gerimis 0.49424248 2.289549 1.249003 0.6015901 2.044282 0.09274115
## Hujan Lebat 0.02965934 2.306960 2.568362 1.0364578 1.371542 0.19998336
## Hujan Sedang 0.28399505 2.543429 1.615863 0.6986036 2.144737 0.09567693
## ff_avg ddd_carBarat Daya ddd_carBarat Laut ddd_carSelatan
## Cerah 3.3253663 4.834473e+00 1.926981e-11 8.843567e-12
## Cerah Berawan 3.2921813 2.570777e-06 NaN 8.125216e-13
## Gerimis 3.2338605 4.088302e+00 NaN 1.540735e-15
## Hujan Lebat 0.1527412 4.679369e-03 1.905820e-26 NaN
## Hujan Sedang 4.8699014 6.676690e+00 4.928367e-72 1.284103e-25
## ddd_carTenang (Calm) ddd_carTimur ddd_carTimur Laut
## Cerah 4.7653406 4.2289338 2.104957e+00
## Cerah Berawan 4.1520844 4.9296845 2.560951e+00
## Gerimis 5.6522086 5.0664194 2.970325e+00
## Hujan Lebat 0.7200695 0.1516557 6.515452e-51
## Hujan Sedang 8.6659637 6.1156506 5.530772e-01
##
## Residual Deviance: 85.0348
## AIC: 215.0348
#disini ada warning karena harusnya antara RR, ss, dan ddd_car salah satu didrop karena multikoliniaritas tingggi
#or
tidy(fit_full, conf.int = TRUE) %>%
kable() %>%
kable_styling("basic", full_width = FALSE)
## Warning in sqrt(diag(vc)): NaNs produced
## Warning in sqrt(diag(vcov(object))): NaNs produced
|
y.level
|
term
|
estimate
|
std.error
|
statistic
|
p.value
|
conf.low
|
conf.high
|
|
Cerah
|
(Intercept)
|
114.1249322
|
6.9414791
|
1.644101e+01
|
0.0000000
|
100.5198833
|
127.7299812
|
|
Cerah
|
RR
|
16.8131757
|
3.7247920
|
4.513856e+00
|
0.0000064
|
9.5127176
|
24.1136338
|
|
Cerah
|
Tavg
|
-6.2656644
|
1.2117857
|
-5.170604e+00
|
0.0000002
|
-8.6407208
|
-3.8906080
|
|
Cerah
|
RH_avg
|
-1.5409423
|
0.6885924
|
-2.237815e+00
|
0.0252331
|
-2.8905587
|
-0.1913260
|
|
Cerah
|
ss
|
10.6675904
|
2.3241100
|
4.589968e+00
|
0.0000044
|
6.1124185
|
15.2227622
|
|
Cerah
|
ddd_x
|
0.0789679
|
0.1042011
|
7.578412e-01
|
0.4485460
|
-0.1252626
|
0.2831984
|
|
Cerah
|
ff_avg
|
7.1307930
|
3.3253663
|
2.144363e+00
|
0.0320038
|
0.6131949
|
13.6483912
|
|
Cerah
|
ddd_carBarat Daya
|
125.9907463
|
4.8344728
|
2.606091e+01
|
0.0000000
|
116.5153537
|
135.4661389
|
|
Cerah
|
ddd_carBarat Laut
|
32.7394329
|
0.0000000
|
1.699001e+12
|
0.0000000
|
32.7394329
|
32.7394329
|
|
Cerah
|
ddd_carSelatan
|
-149.1958217
|
0.0000000
|
-1.687055e+13
|
0.0000000
|
-149.1958217
|
-149.1958217
|
|
Cerah
|
ddd_carTenang (Calm)
|
96.0261239
|
4.7653406
|
2.015095e+01
|
0.0000000
|
86.6862279
|
105.3660199
|
|
Cerah
|
ddd_carTimur
|
105.4589572
|
4.2289338
|
2.493748e+01
|
0.0000000
|
97.1703992
|
113.7475152
|
|
Cerah
|
ddd_carTimur Laut
|
150.6139627
|
2.1049568
|
7.155204e+01
|
0.0000000
|
146.4883232
|
154.7396022
|
|
Cerah Berawan
|
(Intercept)
|
153.6395403
|
6.8265663
|
2.250612e+01
|
0.0000000
|
140.2597161
|
167.0193644
|
|
Cerah Berawan
|
RR
|
8.2828691
|
3.8975489
|
2.125148e+00
|
0.0335742
|
0.6438138
|
15.9219245
|
|
Cerah Berawan
|
Tavg
|
-5.4308207
|
1.2149067
|
-4.470154e+00
|
0.0000078
|
-7.8119941
|
-3.0496473
|
|
Cerah Berawan
|
RH_avg
|
-0.7390146
|
0.6710433
|
-1.101292e+00
|
0.2707696
|
-2.0542353
|
0.5762061
|
|
Cerah Berawan
|
ss
|
7.7187768
|
2.1976067
|
3.512356e+00
|
0.0004442
|
3.4115468
|
12.0260069
|
|
Cerah Berawan
|
ddd_x
|
0.0691988
|
0.1024023
|
6.757547e-01
|
0.4991964
|
-0.1315060
|
0.2699036
|
|
Cerah Berawan
|
ff_avg
|
7.1810570
|
3.2921813
|
2.181246e+00
|
0.0291652
|
0.7285002
|
13.6336138
|
|
Cerah Berawan
|
ddd_carBarat Daya
|
15.6633684
|
0.0000026
|
6.092853e+06
|
0.0000000
|
15.6633634
|
15.6633734
|
|
Cerah Berawan
|
ddd_carBarat Laut
|
45.8923770
|
NaN
|
NaN
|
NaN
|
NaN
|
NaN
|
|
Cerah Berawan
|
ddd_carSelatan
|
104.5393839
|
0.0000000
|
1.286604e+14
|
0.0000000
|
104.5393839
|
104.5393839
|
|
Cerah Berawan
|
ddd_carTenang (Calm)
|
-2.6005395
|
4.1520844
|
-6.263215e-01
|
0.5311041
|
-10.7384754
|
5.5373963
|
|
Cerah Berawan
|
ddd_carTimur
|
7.9905700
|
4.9296845
|
1.620909e+00
|
0.1050372
|
-1.6714341
|
17.6525741
|
|
Cerah Berawan
|
ddd_carTimur Laut
|
52.6555244
|
2.5609508
|
2.056093e+01
|
0.0000000
|
47.6361531
|
57.6748957
|
|
Gerimis
|
(Intercept)
|
-315.3777219
|
0.4942425
|
-6.381032e+02
|
0.0000000
|
-316.3464194
|
-314.4090244
|
|
Gerimis
|
RR
|
32.2772319
|
2.2895492
|
1.409764e+01
|
0.0000000
|
27.7897978
|
36.7646659
|
|
Gerimis
|
Tavg
|
3.8578633
|
1.2490025
|
3.088755e+00
|
0.0020100
|
1.4098633
|
6.3058633
|
|
Gerimis
|
RH_avg
|
1.6782064
|
0.6015901
|
2.789618e+00
|
0.0052770
|
0.4991115
|
2.8573013
|
|
Gerimis
|
ss
|
5.3610786
|
2.0442823
|
2.622475e+00
|
0.0087294
|
1.3543590
|
9.3677983
|
|
Gerimis
|
ddd_x
|
0.0987200
|
0.0927411
|
1.064468e+00
|
0.2871168
|
-0.0830493
|
0.2804893
|
|
Gerimis
|
ff_avg
|
10.1337177
|
3.2338605
|
3.133629e+00
|
0.0017266
|
3.7954675
|
16.4719679
|
|
Gerimis
|
ddd_carBarat Daya
|
30.1647036
|
4.0883021
|
7.378296e+00
|
0.0000000
|
22.1517786
|
38.1776285
|
|
Gerimis
|
ddd_carBarat Laut
|
1.8563866
|
NaN
|
NaN
|
NaN
|
NaN
|
NaN
|
|
Gerimis
|
ddd_carSelatan
|
-44.2898146
|
0.0000000
|
-2.874590e+16
|
0.0000000
|
-44.2898146
|
-44.2898146
|
|
Gerimis
|
ddd_carTenang (Calm)
|
5.4956064
|
5.6522086
|
9.722936e-01
|
0.3309045
|
-5.5825188
|
16.5737316
|
|
Gerimis
|
ddd_carTimur
|
14.1286223
|
5.0664194
|
2.788680e+00
|
0.0052923
|
4.1986227
|
24.0586219
|
|
Gerimis
|
ddd_carTimur Laut
|
60.1467000
|
2.9703254
|
2.024920e+01
|
0.0000000
|
54.3249692
|
65.9684308
|
|
Hujan Lebat
|
(Intercept)
|
4.5069046
|
0.0296593
|
1.519556e+02
|
0.0000000
|
4.4487734
|
4.5650359
|
|
Hujan Lebat
|
RR
|
48.0485304
|
2.3069598
|
2.082764e+01
|
0.0000000
|
43.5269722
|
52.5700886
|
|
Hujan Lebat
|
Tavg
|
2.8020820
|
2.5683620
|
1.091000e+00
|
0.2752730
|
-2.2318149
|
7.8359790
|
|
Hujan Lebat
|
RH_avg
|
-3.7625872
|
1.0364578
|
-3.630237e+00
|
0.0002832
|
-5.7940072
|
-1.7311672
|
|
Hujan Lebat
|
ss
|
-1.6387379
|
1.3715424
|
-1.194814e+00
|
0.2321597
|
-4.3269115
|
1.0494357
|
|
Hujan Lebat
|
ddd_x
|
0.1873951
|
0.1999834
|
9.370533e-01
|
0.3487311
|
-0.2045651
|
0.5793552
|
|
Hujan Lebat
|
ff_avg
|
-11.8834790
|
0.1527412
|
-7.780139e+01
|
0.0000000
|
-12.1828463
|
-11.5841117
|
|
Hujan Lebat
|
ddd_carBarat Daya
|
53.4507688
|
0.0046794
|
1.142264e+04
|
0.0000000
|
53.4415974
|
53.4599402
|
|
Hujan Lebat
|
ddd_carBarat Laut
|
28.4329233
|
0.0000000
|
1.491899e+27
|
0.0000000
|
28.4329233
|
28.4329233
|
|
Hujan Lebat
|
ddd_carSelatan
|
-4.5400196
|
NaN
|
NaN
|
NaN
|
NaN
|
NaN
|
|
Hujan Lebat
|
ddd_carTenang (Calm)
|
1.0414167
|
0.7200695
|
1.446272e+00
|
0.1481008
|
-0.3698937
|
2.4527270
|
|
Hujan Lebat
|
ddd_carTimur
|
-25.8582632
|
0.1516557
|
-1.705064e+02
|
0.0000000
|
-26.1555029
|
-25.5610235
|
|
Hujan Lebat
|
ddd_carTimur Laut
|
118.4232904
|
0.0000000
|
1.817576e+52
|
0.0000000
|
118.4232904
|
118.4232904
|
|
Hujan Sedang
|
(Intercept)
|
-118.2673890
|
0.2839951
|
-4.164417e+02
|
0.0000000
|
-118.8240091
|
-117.7107690
|
|
Hujan Sedang
|
RR
|
39.0146291
|
2.5434295
|
1.533938e+01
|
0.0000000
|
34.0295989
|
43.9996592
|
|
Hujan Sedang
|
Tavg
|
-0.1559959
|
1.6158634
|
-9.654030e-02
|
0.9230915
|
-3.3230298
|
3.0110381
|
|
Hujan Sedang
|
RH_avg
|
0.7540256
|
0.6986036
|
1.079333e+00
|
0.2804395
|
-0.6152123
|
2.1232636
|
|
Hujan Sedang
|
ss
|
5.0348442
|
2.1447370
|
2.347535e+00
|
0.0188981
|
0.8312369
|
9.2384515
|
|
Hujan Sedang
|
ddd_x
|
0.0946452
|
0.0956769
|
9.892163e-01
|
0.3225573
|
-0.0928782
|
0.2821685
|
|
Hujan Sedang
|
ff_avg
|
0.4463558
|
4.8699014
|
9.165600e-02
|
0.9269713
|
-9.0984754
|
9.9911871
|
|
Hujan Sedang
|
ddd_carBarat Daya
|
34.9838733
|
6.6766896
|
5.239703e+00
|
0.0000002
|
21.8978022
|
48.0699444
|
|
Hujan Sedang
|
ddd_carBarat Laut
|
-117.6494737
|
0.0000000
|
-2.387190e+73
|
0.0000000
|
-117.6494737
|
-117.6494737
|
|
Hujan Sedang
|
ddd_carSelatan
|
90.0410587
|
0.0000000
|
7.011979e+26
|
0.0000000
|
90.0410587
|
90.0410587
|
|
Hujan Sedang
|
ddd_carTenang (Calm)
|
-4.2083873
|
8.6659637
|
-4.856225e-01
|
0.6272348
|
-21.1933641
|
12.7765894
|
|
Hujan Sedang
|
ddd_carTimur
|
15.2243801
|
6.1156506
|
2.489413e+00
|
0.0127954
|
3.2379251
|
27.2108351
|
|
Hujan Sedang
|
ddd_carTimur Laut
|
63.0314024
|
0.5530772
|
1.139649e+02
|
0.0000000
|
61.9473910
|
64.1154137
|
# Simpan hasil tidy model ke dalam CSV
output_tidy <- tidy(fit_full, conf.int = TRUE)
## Warning in sqrt(diag(vc)): NaNs produced
## Warning in sqrt(diag(vc)): NaNs produced
write.csv(output_tidy, "Interpret for the relative log odds.csv", row.names = FALSE)
#Interpret for the relative risk ratios----
exp(coef(fit_full))
## (Intercept) RR Tavg RH_avg ss
## Cerah 3.662927e+49 2.003870e+07 0.00190045 0.21417918 42941.343935
## Cerah Berawan 5.306456e+66 3.955527e+03 0.00437950 0.47758430 2250.205525
## Gerimis 1.079433e-137 1.041894e+14 47.36404000 5.35594073 212.954524
## Hujan Lebat 9.064081e+01 7.365659e+20 16.47892087 0.02322358 0.194225
## Hujan Sedang 4.336362e-52 8.786949e+16 0.85556272 2.12553947 153.675646
## ddd_x ff_avg ddd_carBarat Daya ddd_carBarat Laut
## Cerah 1.082170 1.249868e+03 5.212978e+54 1.654074e+14
## Cerah Berawan 1.071649 1.314297e+03 6.346210e+06 8.527193e+19
## Gerimis 1.103757 2.517780e+04 1.259982e+13 6.400567e+00
## Hujan Lebat 1.206104 6.903521e-06 1.634459e+23 2.229778e+12
## Hujan Sedang 1.099269 1.562607e+00 1.560641e+15 8.044199e-52
## ddd_carSelatan ddd_carTenang (Calm) ddd_carTimur
## Cerah 1.603533e-65 5.053773e+41 6.313107e+45
## Cerah Berawan 2.516967e+45 7.423352e-02 2.952980e+03
## Gerimis 5.823417e-20 2.436192e+02 1.367674e+06
## Hujan Lebat 1.067320e-02 2.833228e+00 5.887067e-12
## Hujan Sedang 1.271554e+39 1.487033e-02 4.091328e+06
## ddd_carTimur Laut
## Cerah 2.575211e+65
## Cerah Berawan 7.379105e+22
## Gerimis 1.322452e+26
## Hujan Lebat 2.695141e+51
## Hujan Sedang 2.366956e+27
#or
tidy(fit_full, conf.int = TRUE, exponentiate = TRUE) %>%
kable() %>%
kable_styling("basic", full_width = FALSE)
## Warning in sqrt(diag(vc)): NaNs produced
## Warning in sqrt(diag(vc)): NaNs produced
|
y.level
|
term
|
estimate
|
std.error
|
statistic
|
p.value
|
conf.low
|
conf.high
|
|
Cerah
|
(Intercept)
|
3.662927e+49
|
6.9414791
|
1.644101e+01
|
0.0000000
|
4.520960e+43
|
2.967741e+55
|
|
Cerah
|
RR
|
2.003870e+07
|
3.7247920
|
4.513856e+00
|
0.0000064
|
1.353072e+04
|
2.967687e+10
|
|
Cerah
|
Tavg
|
1.900500e-03
|
1.2117857
|
-5.170604e+00
|
0.0000002
|
1.768000e-04
|
2.043290e-02
|
|
Cerah
|
RH_avg
|
2.141792e-01
|
0.6885924
|
-2.237815e+00
|
0.0252331
|
5.554520e-02
|
8.258633e-01
|
|
Cerah
|
ss
|
4.294134e+04
|
2.3241100
|
4.589968e+00
|
0.0000044
|
4.514292e+02
|
4.084714e+06
|
|
Cerah
|
ddd_x
|
1.082170e+00
|
0.1042011
|
7.578412e-01
|
0.4485460
|
8.822652e-01
|
1.327368e+00
|
|
Cerah
|
ff_avg
|
1.249868e+03
|
3.3253663
|
2.144363e+00
|
0.0320038
|
1.846321e+00
|
8.460986e+05
|
|
Cerah
|
ddd_carBarat Daya
|
5.212978e+54
|
4.8344728
|
2.606091e+01
|
0.0000000
|
3.999219e+50
|
6.795112e+58
|
|
Cerah
|
ddd_carBarat Laut
|
1.654074e+14
|
0.0000000
|
1.699001e+12
|
0.0000000
|
1.654074e+14
|
1.654074e+14
|
|
Cerah
|
ddd_carSelatan
|
0.000000e+00
|
0.0000000
|
-1.687055e+13
|
0.0000000
|
0.000000e+00
|
0.000000e+00
|
|
Cerah
|
ddd_carTenang (Calm)
|
5.053773e+41
|
4.7653406
|
2.015095e+01
|
0.0000000
|
4.439667e+37
|
5.752823e+45
|
|
Cerah
|
ddd_carTimur
|
6.313107e+45
|
4.2289338
|
2.493748e+01
|
0.0000000
|
1.586968e+42
|
2.511413e+49
|
|
Cerah
|
ddd_carTimur Laut
|
2.575211e+65
|
2.1049568
|
7.155204e+01
|
0.0000000
|
4.159781e+63
|
1.594246e+67
|
|
Cerah Berawan
|
(Intercept)
|
5.306456e+66
|
6.8265663
|
2.250612e+01
|
0.0000000
|
8.203908e+60
|
3.432324e+72
|
|
Cerah Berawan
|
RR
|
3.955527e+03
|
3.8975489
|
2.125148e+00
|
0.0335742
|
1.903727e+00
|
8.218716e+06
|
|
Cerah Berawan
|
Tavg
|
4.379500e-03
|
1.2149067
|
-4.470154e+00
|
0.0000078
|
4.048000e-04
|
4.737560e-02
|
|
Cerah Berawan
|
RH_avg
|
4.775843e-01
|
0.6710433
|
-1.101292e+00
|
0.2707696
|
1.281908e-01
|
1.779275e+00
|
|
Cerah Berawan
|
ss
|
2.250206e+03
|
2.1976067
|
3.512356e+00
|
0.0004442
|
3.031209e+01
|
1.670431e+05
|
|
Cerah Berawan
|
ddd_x
|
1.071649e+00
|
0.1024023
|
6.757547e-01
|
0.4991964
|
8.767740e-01
|
1.309838e+00
|
|
Cerah Berawan
|
ff_avg
|
1.314297e+03
|
3.2921813
|
2.181246e+00
|
0.0291652
|
2.071971e+00
|
8.336874e+05
|
|
Cerah Berawan
|
ddd_carBarat Daya
|
6.346210e+06
|
0.0000026
|
6.092853e+06
|
0.0000000
|
6.346178e+06
|
6.346242e+06
|
|
Cerah Berawan
|
ddd_carBarat Laut
|
8.527193e+19
|
NaN
|
NaN
|
NaN
|
NaN
|
NaN
|
|
Cerah Berawan
|
ddd_carSelatan
|
2.516967e+45
|
0.0000000
|
1.286604e+14
|
0.0000000
|
2.516967e+45
|
2.516967e+45
|
|
Cerah Berawan
|
ddd_carTenang (Calm)
|
7.423350e-02
|
4.1520844
|
-6.263215e-01
|
0.5311041
|
2.170000e-05
|
2.540158e+02
|
|
Cerah Berawan
|
ddd_carTimur
|
2.952980e+03
|
4.9296845
|
1.620909e+00
|
0.1050372
|
1.879773e-01
|
4.638905e+07
|
|
Cerah Berawan
|
ddd_carTimur Laut
|
7.379105e+22
|
2.5609508
|
2.056093e+01
|
0.0000000
|
4.876615e+20
|
1.116578e+25
|
|
Gerimis
|
(Intercept)
|
0.000000e+00
|
0.4942425
|
-6.381032e+02
|
0.0000000
|
0.000000e+00
|
0.000000e+00
|
|
Gerimis
|
RR
|
1.041894e+14
|
2.2895492
|
1.409764e+01
|
0.0000000
|
1.172076e+12
|
9.261717e+15
|
|
Gerimis
|
Tavg
|
4.736404e+01
|
1.2490025
|
3.088755e+00
|
0.0020100
|
4.095396e+00
|
5.477743e+02
|
|
Gerimis
|
RH_avg
|
5.355941e+00
|
0.6015901
|
2.789618e+00
|
0.0052770
|
1.647257e+00
|
1.741447e+01
|
|
Gerimis
|
ss
|
2.129545e+02
|
2.0442823
|
2.622475e+00
|
0.0087294
|
3.874277e+00
|
1.170532e+04
|
|
Gerimis
|
ddd_x
|
1.103757e+00
|
0.0927411
|
1.064468e+00
|
0.2871168
|
9.203057e-01
|
1.323777e+00
|
|
Gerimis
|
ff_avg
|
2.517780e+04
|
3.2338605
|
3.133629e+00
|
0.0017266
|
4.449904e+01
|
1.424573e+07
|
|
Gerimis
|
ddd_carBarat Daya
|
1.259982e+13
|
4.0883021
|
7.378296e+00
|
0.0000000
|
4.172489e+09
|
3.804814e+16
|
|
Gerimis
|
ddd_carBarat Laut
|
6.400567e+00
|
NaN
|
NaN
|
NaN
|
NaN
|
NaN
|
|
Gerimis
|
ddd_carSelatan
|
0.000000e+00
|
0.0000000
|
-2.874590e+16
|
0.0000000
|
0.000000e+00
|
0.000000e+00
|
|
Gerimis
|
ddd_carTenang (Calm)
|
2.436192e+02
|
5.6522086
|
9.722936e-01
|
0.3309045
|
3.763100e-03
|
1.577176e+07
|
|
Gerimis
|
ddd_carTimur
|
1.367674e+06
|
5.0664194
|
2.788680e+00
|
0.0052923
|
6.659455e+01
|
2.808838e+10
|
|
Gerimis
|
ddd_carTimur Laut
|
1.322452e+26
|
2.9703254
|
2.024920e+01
|
0.0000000
|
3.917729e+23
|
4.464013e+28
|
|
Hujan Lebat
|
(Intercept)
|
9.064082e+01
|
0.0296593
|
1.519556e+02
|
0.0000000
|
8.552198e+01
|
9.606604e+01
|
|
Hujan Lebat
|
RR
|
7.365659e+20
|
2.3069598
|
2.082764e+01
|
0.0000000
|
8.007996e+18
|
6.774846e+22
|
|
Hujan Lebat
|
Tavg
|
1.647892e+01
|
2.5683620
|
1.091000e+00
|
0.2752730
|
1.073335e-01
|
2.530011e+03
|
|
Hujan Lebat
|
RH_avg
|
2.322360e-02
|
1.0364578
|
-3.630237e+00
|
0.0002832
|
3.045800e-03
|
1.770776e-01
|
|
Hujan Lebat
|
ss
|
1.942250e-01
|
1.3715424
|
-1.194814e+00
|
0.2321597
|
1.320830e-02
|
2.856039e+00
|
|
Hujan Lebat
|
ddd_x
|
1.206104e+00
|
0.1999834
|
9.370533e-01
|
0.3487311
|
8.150017e-01
|
1.784887e+00
|
|
Hujan Lebat
|
ff_avg
|
6.900000e-06
|
0.1527412
|
-7.780139e+01
|
0.0000000
|
5.100000e-06
|
9.300000e-06
|
|
Hujan Lebat
|
ddd_carBarat Daya
|
1.634459e+23
|
0.0046794
|
1.142264e+04
|
0.0000000
|
1.619537e+23
|
1.649518e+23
|
|
Hujan Lebat
|
ddd_carBarat Laut
|
2.229778e+12
|
0.0000000
|
1.491899e+27
|
0.0000000
|
2.229778e+12
|
2.229778e+12
|
|
Hujan Lebat
|
ddd_carSelatan
|
1.067320e-02
|
NaN
|
NaN
|
NaN
|
NaN
|
NaN
|
|
Hujan Lebat
|
ddd_carTenang (Calm)
|
2.833228e+00
|
0.7200695
|
1.446272e+00
|
0.1481008
|
6.908078e-01
|
1.161999e+01
|
|
Hujan Lebat
|
ddd_carTimur
|
0.000000e+00
|
0.1516557
|
-1.705064e+02
|
0.0000000
|
0.000000e+00
|
0.000000e+00
|
|
Hujan Lebat
|
ddd_carTimur Laut
|
2.695141e+51
|
0.0000000
|
1.817576e+52
|
0.0000000
|
2.695141e+51
|
2.695141e+51
|
|
Hujan Sedang
|
(Intercept)
|
0.000000e+00
|
0.2839951
|
-4.164417e+02
|
0.0000000
|
0.000000e+00
|
0.000000e+00
|
|
Hujan Sedang
|
RR
|
8.786949e+16
|
2.5434295
|
1.533938e+01
|
0.0000000
|
6.009897e+14
|
1.284722e+19
|
|
Hujan Sedang
|
Tavg
|
8.555627e-01
|
1.6158634
|
-9.654030e-02
|
0.9230915
|
3.604350e-02
|
2.030847e+01
|
|
Hujan Sedang
|
RH_avg
|
2.125539e+00
|
0.6986036
|
1.079333e+00
|
0.2804395
|
5.405261e-01
|
8.358371e+00
|
|
Hujan Sedang
|
ss
|
1.536756e+02
|
2.1447370
|
2.347535e+00
|
0.0188981
|
2.296157e+00
|
1.028510e+04
|
|
Hujan Sedang
|
ddd_x
|
1.099269e+00
|
0.0956769
|
9.892163e-01
|
0.3225573
|
9.113045e-01
|
1.326002e+00
|
|
Hujan Sedang
|
ff_avg
|
1.562607e+00
|
4.8699014
|
9.165600e-02
|
0.9269713
|
1.118000e-04
|
2.183320e+04
|
|
Hujan Sedang
|
ddd_carBarat Daya
|
1.560641e+15
|
6.6766896
|
5.239703e+00
|
0.0000002
|
3.236642e+09
|
7.525088e+20
|
|
Hujan Sedang
|
ddd_carBarat Laut
|
0.000000e+00
|
0.0000000
|
-2.387190e+73
|
0.0000000
|
0.000000e+00
|
0.000000e+00
|
|
Hujan Sedang
|
ddd_carSelatan
|
1.271554e+39
|
0.0000000
|
7.011979e+26
|
0.0000000
|
1.271554e+39
|
1.271554e+39
|
|
Hujan Sedang
|
ddd_carTenang (Calm)
|
1.487030e-02
|
8.6659637
|
-4.856225e-01
|
0.6272348
|
0.000000e+00
|
3.538362e+05
|
|
Hujan Sedang
|
ddd_carTimur
|
4.091328e+06
|
6.1156506
|
2.489413e+00
|
0.0127954
|
2.548080e+01
|
6.569246e+11
|
|
Hujan Sedang
|
ddd_carTimur Laut
|
2.366956e+27
|
0.5530772
|
1.139649e+02
|
0.0000000
|
8.005899e+26
|
6.997943e+27
|
# Simpan hasil tidy model ke dalam CSV
output_tidy <- tidy(fit_full, conf.int = TRUE, exponentiate = TRUE)
## Warning in sqrt(diag(vc)): NaNs produced
## Warning in sqrt(diag(vc)): NaNs produced
write.csv(output_tidy, "Interpret for the relative risk ratios.csv", row.names = FALSE)
levels(factor(data$ddd_car))
## [1] "Barat" "Barat Daya" "Barat Laut" "Selatan"
## [5] "Tenang (Calm)" "Timur" "Timur Laut"
# ---------------- EVALUASI MLR ----------------
#library yang diperlukan
library(nnet) # Untuk multinom()
#melakukan prediksi terhadap data
mlr_pred <- predict(fit_full, newdata = data_clean)
data_clean$MLR_Predicted <- mlr_pred
#membuat confusion matrix antara prediksi dan label asli
conf_matrix <- table(Predicted = mlr_pred,
Actual = data_clean$Kategori_Cuaca)
print(conf_matrix)
## Actual
## Predicted Berawan Cerah Cerah Berawan Gerimis Hujan Lebat Hujan Sedang
## Berawan 6 0 0 0 0 0
## Cerah 0 55 7 2 0 0
## Cerah Berawan 0 4 53 0 0 0
## Gerimis 0 0 0 26 0 0
## Hujan Lebat 0 0 0 0 17 0
## Hujan Sedang 0 0 0 0 0 37
#menghitung akurasi klasifikasi dari model
accuracy <- mean(mlr_pred == data_clean$Kategori_Cuaca)
cat("\nAkurasi klasifikasi MLR:", round(accuracy * 100, 2), "%\n")
##
## Akurasi klasifikasi MLR: 93.72 %
#Membuat dataframe berisi skor diskriminan dan label asli
mlr_scores <- data.frame(mlr_pred)
mlr_scores$Kategori_Cuaca <- data_clean$Kategori_Cuaca
#----------------LDA---------------
# =======================
# LDA Multikelas Cuaca
# =======================
# Pastikan library MASS tersedia
if (!require(MASS)) install.packages("MASS")
## Loading required package: MASS
##
## Attaching package: 'MASS'
##
## The following object is masked from 'package:dplyr':
##
## select
library(MASS)
# Pastikan juga variabel kategorikal (ddd_car dan Kategori_Cuaca) bertipe faktor
data_clean$ddd_car <- as.factor(data_clean$ddd_car)
data_clean$Kategori_Cuaca <- as.factor(data_clean$Kategori_Cuaca)
# LDA menggunakan semua variabel prediktor yang diminta
lda_model <- lda(Kategori_Cuaca ~ Tavg + RH_avg + RR + ss + ddd_x + ff_avg + ddd_car, data = data_clean)
# Ringkasan model
print(lda_model)
## Call:
## lda(Kategori_Cuaca ~ Tavg + RH_avg + RR + ss + ddd_x + ff_avg +
## ddd_car, data = data_clean)
##
## Prior probabilities of groups:
## Berawan Cerah Cerah Berawan Gerimis Hujan Lebat
## 0.02898551 0.28502415 0.28985507 0.13526570 0.08212560
## Hujan Sedang
## 0.17874396
##
## Group means:
## Tavg RH_avg RR ss ddd_x ff_avg
## Berawan 28.25000 80.66667 0.0000000 2.633333 108.33333 2.166667
## Cerah 29.19661 68.69492 0.0000000 8.071186 90.89831 2.847458
## Cerah Berawan 28.93167 76.50000 0.0000000 7.650000 108.50000 2.583333
## Gerimis 28.79643 82.07143 0.9678571 5.075000 162.50000 2.107143
## Hujan Lebat 28.50000 82.94118 31.5000000 4.805882 162.35294 1.941176
## Hujan Sedang 28.73514 82.21622 10.1621622 5.172973 172.70270 2.108108
## ddd_carBarat Daya ddd_carBarat Laut ddd_carSelatan
## Berawan 0.16666667 0.00000000 0.00000000
## Cerah 0.01694915 0.00000000 0.00000000
## Cerah Berawan 0.00000000 0.00000000 0.01666667
## Gerimis 0.14285714 0.00000000 0.00000000
## Hujan Lebat 0.17647059 0.05882353 0.00000000
## Hujan Sedang 0.13513514 0.00000000 0.02702703
## ddd_carTenang (Calm) ddd_carTimur ddd_carTimur Laut
## Berawan 0.33333333 0.3333333 0.00000000
## Cerah 0.03389831 0.8305085 0.11864407
## Cerah Berawan 0.05000000 0.6833333 0.11666667
## Gerimis 0.28571429 0.3214286 0.07142857
## Hujan Lebat 0.17647059 0.2352941 0.00000000
## Hujan Sedang 0.21621622 0.2972973 0.02702703
##
## Coefficients of linear discriminants:
## LD1 LD2 LD3 LD4
## Tavg 0.229325211 0.432202376 -0.230089387 -0.350835576
## RH_avg 0.087271730 0.200434166 -0.149675103 0.041040283
## RR 0.284937442 -0.107048505 -0.014890439 -0.008025864
## ss -0.146713392 -0.162785938 -0.377516773 -0.073061535
## ddd_x 0.002611645 0.001011126 -0.001934972 -0.015235921
## ff_avg -0.170157592 -0.054360293 -0.217155274 0.372316139
## ddd_carBarat Daya 0.790433879 -0.010280672 1.060621351 -1.414675873
## ddd_carBarat Laut 1.987601945 -2.592134308 0.631015334 2.905221064
## ddd_carSelatan 1.581409782 1.746998802 -2.413638672 0.097271736
## ddd_carTenang (Calm) -0.434772762 0.124862949 1.095778244 -1.254907690
## ddd_carTimur 0.068728399 -0.268408114 -0.573257075 -1.484483318
## ddd_carTimur Laut 0.225484945 -0.075320546 -1.046307996 -1.947796412
## LD5
## Tavg -0.227119208
## RH_avg -0.041359258
## RR 0.008470898
## ss -0.062862919
## ddd_x 0.000195861
## ff_avg 0.029015935
## ddd_carBarat Daya -1.406425801
## ddd_carBarat Laut -9.029519385
## ddd_carSelatan 5.200024780
## ddd_carTenang (Calm) -1.309853488
## ddd_carTimur -0.911858689
## ddd_carTimur Laut -1.928780916
##
## Proportion of trace:
## LD1 LD2 LD3 LD4 LD5
## 0.8476 0.1254 0.0202 0.0052 0.0015
# Lakukan prediksi menggunakan model LDA
lda_pred <- predict(lda_model)
# Plot hasil proyeksi ke LD1 dan LD2
plot(lda_pred$x[, 1:2],
col = as.numeric(data_clean$Kategori_Cuaca),
pch = 19,
xlab = "LD1", ylab = "LD2",
main = "LDA Scatter Plot")
legend("topright", legend = levels(data_clean$Kategori_Cuaca),
col = 1:length(levels(data_clean$Kategori_Cuaca)), pch = 19)
#----------------Evaluasi----------------
#melakukan prediksi kelas berdasarkan model LDA
lda_pred <- predict(lda_model, newdata = data_clean)
data_clean$LDA_Predicted <- lda_pred$class
#membuat confusion matrix antara prediksi dan label asli
conf_matrix <- table(Predicted = lda_pred$class,
Actual = data_clean$Kategori_Cuaca)
print(conf_matrix)
## Actual
## Predicted Berawan Cerah Cerah Berawan Gerimis Hujan Lebat Hujan Sedang
## Berawan 2 0 0 2 0 1
## Cerah 0 53 8 0 0 0
## Cerah Berawan 1 6 49 8 0 2
## Gerimis 3 0 3 18 0 5
## Hujan Lebat 0 0 0 0 16 0
## Hujan Sedang 0 0 0 0 1 29
#menghitung akurasi klasifikasi dari model
accuracy <- mean(lda_pred$class == data_clean$Kategori_Cuaca)
cat("\nAkurasi klasifikasi LDA:", round(accuracy * 100, 2), "%\n")
##
## Akurasi klasifikasi LDA: 80.68 %
manova_model <- manova(cbind(Tavg, RH_avg, RR, ss, ddd_x, ff_avg, as.numeric(ddd_car)) ~ Kategori_Cuaca, data = data_clean)
# Summary dengan Wilks' Lambda
summary(manova_model, test = "Wilks")
## Df Wilks approx F num Df den Df Pr(>F)
## Kategori_Cuaca 5 0.035282 28.548 35 822.72 < 2.2e-16 ***
## Residuals 201
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#----------------PERCEPTUAL MAPPINNG---------------
# Prediksi dan Konfusi untuk LDA
# Prediksi menggunakan LDA
lda_pred <- predict(lda_model, data_clean)
lda_class <- lda_pred$class
# Ubah ke angka untuk MDS
lda_dist <- dist(scale(as.matrix(lda_pred$x[, 1:2])))
lda_mds <- cmdscale(lda_dist)
# Dataframe untuk plot LDA
lda_df <- data.frame(
Dim1 = lda_mds[, 1],
Dim2 = lda_mds[, 2],
Predicted = lda_class
)
# Prediksi dan Konfusi untuk MLR
# Prediksi menggunakan Multinomial Logistic Regression
mlr_pred <- predict(fit_full, newdata = data_clean)
mlr_prob <- predict(fit_full, newdata = data_clean, type = "probs")
# Gunakan probabilitas (jarak antar prediksi) untuk MDS
mlr_dist <- dist(scale(as.matrix(mlr_prob)))
mlr_mds <- cmdscale(mlr_dist)
# Dataframe untuk plot MLR
mlr_df <- data.frame(
Dim1 = mlr_mds[, 1],
Dim2 = mlr_mds[, 2],
Predicted = mlr_pred
)
#Plot Perceptual Mapping
library(ggplot2)
# Plot MDS LDA
plot_lda <- ggplot(lda_df, aes(x = Dim1, y = Dim2, color = Predicted)) +
geom_point(size = 2) +
labs(title = "MDS Mapping Berdasarkan Hasil Klasifikasi LDA") +
theme_minimal()
# Plot MDS MLR
plot_mlr <- ggplot(mlr_df, aes(x = Dim1, y = Dim2, color = Predicted)) +
geom_point(size = 2) +
labs(title = "MDS Mapping Berdasarkan Hasil Multinomial Logistic Regression") +
theme_minimal()
# Tampilkan dua plot secara bersamaan
library(gridExtra)
grid.arrange(plot_lda, plot_mlr, ncol = 2)
