# Data Frame untuk Variabel Kategori
df_kategori <- data.frame(
  Variabel = c("Region", "InsurancePlan", "Gender", "SmokerStatus", "EmploymentType", "PolicyCategory"),
  Deskripsi = c(
    "Wilayah geografis pemegang polis.",
    "Jenis paket asuransi yang dimiliki.",
    "Jenis kelamin pemegang polis.",
    "Status merokok pemegang polis.",
    "Jenis pekerjaan pemegang polis.",
    "Variabel turunan yang mengelompokkan plan menjadi dua kategori untuk analisis."
  ),
  Level.Kategori.Kunci = c(
    "Urban, Suburban, Rural",
    "Basic, Silver, Gold, Platinum",
    "Male, Female",
    "Smoker, Non-Smoker",
    "Private, Government, Self-Employed, Unemployed",
    "Premium (Gold/Platinum), Standar (Basic/Silver)"
  )
)

# Menampilkan tabel 
datatable(
  df_kategori,
  caption = 'Deskripsi Variabel Kategori',
  rownames = FALSE,
  colnames = c('Variabel', 'Deskripsi', 'Level / Kategori Kunci'),
  options = list(
    dom = 't', 
    ordering = FALSE
  )
)

# Data Frame untuk Variabel Numerik
df_numerik <- data.frame(
  Variabel = c("Age", "BMI", "Dependents", "AnnualIncome", "ExerciseHours", "HealthScore", "ClaimAmount", "RiskScore"),
  Deskripsi = c(
    "Usia pemegang polis.",
    "Body Mass Index (Indeks Massa Tubuh).",
    "Jumlah tanggungan dalam keluarga (anak, dll.).",
    "Pendapatan tahunan pemegang polis.",
    "Rata-rata jam olahraga per minggu.",
    "Skor kesehatan gabungan yang ditetapkan oleh perusahaan asuransi.",
    "Jumlah klaim asuransi yang diajukan oleh pemegang polis.",
    "Skor risiko keseluruhan yang dihitung."
  ),
  Satuan.Skala = c("Tahun", "Skor (kg/m²)", "Jumlah", "Ribuan USD", "Jam", "Skor (0-100)", "Ribuan USD", "Skor"),
  Relevansi.dalam.Asuransi = c(
    "Faktor utama dalam risiko dan penentuan premi.",
    "Indikator kesehatan; nilai tinggi (>25) dapat meningkatkan risiko.",
    "Mempengaruhi ukuran keluarga dan potensi klaim.",
    "Indikator kemampuan finansial.",
    "Indikator gaya hidup sehat; nilai tinggi dapat menurunkan risiko.",
    "Penilaian risiko kesehatan secara umum.",
    "Variabel hasil utama (target); mengukur biaya asuransi.",
    "Variabel hasil; hasil akhir dari semua faktor risiko."
  )
)

# Menampilkan tabel
datatable(
  df_numerik,
  caption = 'Deskripsi Variabel Numerik',
  rownames = FALSE,
  colnames = c('Variabel', 'Deskripsi', 'Satuan / Skala', 'Relevansi dalam Asuransi'),
  options = list(
    dom = 't', 
    ordering = FALSE
  )
)

# Hitung Mean, Median, dan Modus ClaimAmount
mean_claim <- mean(df$ClaimAmount, na.rm = TRUE)
median_claim <- median(df$ClaimAmount, na.rm = TRUE)
mode_claim <- get_mode(df$ClaimAmount) 

  
# Membuat data frame untuk Garis Vertikal
vlines_df <- data.frame(
  Value = c(mean_claim, median_claim, mode_claim),
  Label = c("Mean", "Median", "Mode"),
  Color = c("#0072B2", "#009E73", "#FF6347"),
  Linetype = c("dashed", "solid", "dotted")
)

# Fungsi untuk menghitung Mode
get_mode <- function(v) {
  v_filtered <- v[!is.na(v)] 
  if(length(v_filtered) == 0) return(NA)
  uniq_v <- unique(v_filtered)
  mode_val <- uniq_v[which.max(tabulate(match(v_filtered, uniq_v)))]
  return(mode_val[1]) 
}

# Hitung Mean, Median, dan Modus ClaimAmount
mean_claim <- mean(df$ClaimAmount, na.rm = TRUE)
median_claim <- median(df$ClaimAmount, na.rm = TRUE)
mode_claim <- get_mode(df$ClaimAmount) 

  
# Membuat data frame untuk Garis Vertikal
vlines_df <- data.frame(
  Value = c(mean_claim, median_claim, mode_claim),
  Label = c("Mean", "Median", "Mode"),
  Color = c("#0072B2", "#009E73", "#FF6347"),
  Linetype = c("dashed", "solid", "dotted")
)

# Plot Dasar dengan ggplot
p_dist <- ggplot(df, aes(x = ClaimAmount)) + 
  
  # Histogram
  geom_histogram(
    aes(y = after_stat(density),    
        text = paste0("Jumlah Data: ",
                      after_stat(count))),   
    binwidth = 3,    
    fill = "#337AB7",    
    color = "white",    
    alpha = 0.7
  ) +
  
  # Kurva Kepadatan (Density)
  geom_density(aes(text = paste0("Kepadatan: ",
                                 round(after_stat(density),
                                       4))),
    fill = "#D9534F",
    color = "#D9534F",
    linewidth = 1.2,
    alpha = 0.2
  )+
  
  # Penyesuaian batas axis-Y
  ylim(0, 0.043) +
  
  # Garis Vertikal: Memetakan color dan linetype ke Label
  geom_vline(
    data = vlines_df,
    aes(xintercept = Value, color = Label, linetype = Label, 
        text = paste0(Label, ": ", round(Value, 2))), 
    linewidth = 1
  ) +
  
  # Kustomisasi Skala Warna (Untuk Legenda)
  scale_color_manual(
    name = "Tendensi Sentral",
    values = setNames(vlines_df$Color, vlines_df$Label)
  ) +
  
  # Kustomisasi Skala Tipe Garis 
  scale_linetype_manual(
    name = "Tendensi Sentral",
    values = setNames(vlines_df$Linetype, vlines_df$Label)
  ) +
  
  # Label dan Tema
  labs(
    title = "Distribusi Jumlah Klaim",
    subtitle = paste0("Distribusi Condong Kanan: Mode (", round(mode_claim, 2), ") < Median (", round(median_claim, 2), ") < Mean (", round(mean_claim, 2), ")"),
    x = "Jumlah Klaim (Rp Ribuan)",
    y = "Kepadatan Probabilitas"
  ) +
  theme_minimal(base_size = 14) +
  theme(
    plot.title = element_text(hjust = 0.5, face = "bold"),
    plot.subtitle = element_text(hjust = 0.5, face = "italic"),
    legend.position = "right"   
  )

# Konversi ke Plotly
p_plotly <- ggplotly(p_dist, tooltip = "text", width = 750, height = 500)

# Penghapus duplikat legenda 
num_traces <- length(p_plotly$x$data)
# Trace untuk 'linetype'
traces_to_hide <- (num_traces - 2):num_traces

for (i in traces_to_hide) {
  p_plotly$x$data[[i]]$showlegend = FALSE
}

# Tampilkan Plotly
p_plotly %>%
  layout(hovermode = 'x unified') %>%
  config(displayModeBar = TRUE, modeBarButtonsToRemove = c('zoomIn2d', 'zoomOut2d', 'autoScale2d', 'toggleSpikelines', 'select2d', 'lasso2d'))

# Hitung Mean,Median dan Modus Age
mean_age <- mean(df$Age, na.rm = TRUE)
median_age <- median(df$Age, na.rm = TRUE)
mode_age <- get_mode(df$Age)

ggplot(df, aes(x = Age)) +
  
  # Histogram
  geom_histogram(
    aes(y = after_stat(density)), 
    binwidth = 5, 
    fill = "#5CB85C",
    color = "white", 
    alpha = 0.7
  ) +
  
  # Kurva Kepadatan
  geom_density(color = "#D9534F",
               fill = "#D9534F",
               alpha = 0.2,
               linewidth = 1.2) +
  
  # Garis Vertikal Mean 
  geom_vline(
    aes(xintercept = mean_age, color = "Mean"),
    linetype = "dashed",
    linewidth = 1
  ) +
  
  # Garis Vertikal Median
  geom_vline(
    aes(xintercept = median_age, color = "Median"),
    linetype = "solid",
    linewidth = 1
  ) +
  
  # Garis Vertikal MODUS 
  geom_vline(
    aes(xintercept = mode_age, color = "Mode"),
    linetype = "dotted", 
    linewidth = 1
  ) +
  
  # Penyesuaian batas axis-Y
  ylim(0, 0.021) +
  
  # Anotasi Mean dan Median
  annotate("text", 
           x = mean_age + 5, y = 0.018, 
           label = paste0("Mean: ", round(mean_age, 2)), 
           color = "#0072B2", fontface = "bold", size = 4) +
  annotate("text", 
           x = median_age - 5, y = 0.02,
           label = paste0("Median: ", round(median_age, 2)), 
           color = "#009E73", fontface = "bold", size = 4) +
  
  # Anotasi MODUS 
  annotate("text", 
           x = mode_age + 5, y = 0.013,
           label = paste0("Mode: ", round(mode_age, 2)), 
           color = "#FF6347", fontface = "bold", size = 4) + 
  
  # Kustomisasi Skala Warna untuk Garis VLine
  scale_color_manual(
    name = "Tendensi Sentral",
    values = c("Mean" = "#0072B2", "Median" = "#009E73", "Mode" = "#FF6347")
  ) +
  
  # Label dan Tema
  labs(
    title = "Distribusi Usia (Age)",
    subtitle = paste0("Mean (", round(mean_age, 2), ") ≈ Median (", round(median_age, 2), ") ≈ Mode (", round(mode_age, 2), ")"),
    x = "Usia Pemegang Polis",
    y = "Kepadatan Probabilitas"
  ) +
  theme_minimal(base_size = 14) +
  theme(
    plot.title = element_text(hjust = 0.5, face = "bold"),
    plot.subtitle = element_text(hjust = 0.5, face = "italic"),
    legend.position = "none"
  )

# Hitung rata-rata klaim per kelompok usia (bin 5 tahun)
df_agg <- df %>%
  mutate(age_group = cut(Age, breaks = seq(15, 85, by = 5), include.lowest = TRUE, right = FALSE)) %>%
  group_by(age_group, PolicyCategory) %>%
  summarise(
    mean_claim = mean(ClaimAmount, na.rm = TRUE),
    .groups = 'drop'
  )

# Plot Dasar dengan ggplot
p <- ggplot(df_agg, aes(
    x = age_group, 
    y = mean_claim, 
    group = PolicyCategory, 
    color = PolicyCategory,
    # Menambahkan teks untuk tooltip yang lebih informatif
    text = paste0(
      "Grup Usia: ", age_group, "<br>",
      "Kategori: ", PolicyCategory, "<br>",
      "Klaim Rata-rata: ", scales::dollar(mean_claim, prefix = "Rp", scale = 1000, big.mark = ".")
    )
  )) +
  geom_line(linewidth = 1.2) +
  geom_point(size = 3) +
  scale_y_continuous(labels = scales::label_dollar(prefix = "Rp", scale = 1000, big.mark = ".")) +
  labs(
    title = "Tren Biaya Klaim Rata-rata Seiring Usia",
    subtitle = "Segmentasi Premium (Gold/Platinum) vs. Standar (Basic/Silver)",
    x = "Kelompok Usia (Tahun)",
    y = "Rata-rata Jumlah Klaim (Rp Ribuan)",
    color = "Kategori Polis"
  ) +
  theme_minimal(base_size = 14) +
  theme(
    plot.title = element_text(hjust = 0.5, face = "bold"),
    axis.text.x = element_text(angle = 45, hjust = 1)
  ) +
  scale_color_manual(values = c("Premium (Gold/Platinum)" = "#FFC300", "Standar (Basic/Silver)" = "#0057B7"))

# Konversi ke Plotly
ggplotly(p, tooltip = "text", width = 800, height = 500) %>%
  layout(
    xaxis = list(title = "Kelompok Usia (Tahun)"),
    yaxis = list(title = "Rata-rata Jumlah Klaim (Rp Ribuan)")
  )

# Variabel 1: Median Smoker
median_smoker <- df %>%
  filter(SmokerStatus == "Smoker") %>%
  summarise(median_value = median(ClaimAmount, na.rm = TRUE)) %>%
  pull(median_value)

# Variabel 2: Median Non-Smoker 
median_non_smoker <- df %>%
  filter(SmokerStatus == "Non-Smoker") %>%
  summarise(median_value = median(ClaimAmount, na.rm = TRUE)) %>%
  pull(median_value)

# Variabel 3: Mean Smoker
mean_smoker <- df %>%
  filter(SmokerStatus == "Smoker") %>%
  summarise(mean_value = mean(ClaimAmount, na.rm = TRUE)) %>%
  pull(mean_value)


# Hitung Mean dan Median per kelompok untuk anotasi 
claim_stats_values <- df %>%
    group_by(SmokerStatus) %>%
    summarise(
      Median_Value = median(ClaimAmount, na.rm = TRUE),
      Mean_Value = mean(ClaimAmount, na.rm = TRUE),
      .groups = 'drop'
    ) %>%
    mutate(
      Median_Label = paste0("Median: Rp", round(Median_Value, 1), "K"),
      Mean_Label = paste0("Mean: Rp", round(Mean_Value, 1), "K")
    )

# Plot Dasar 
p_smoker <- ggplot(df, aes(x = SmokerStatus, y = ClaimAmount, fill = SmokerStatus)) +
  
  # Violin Plot
  geom_violin(
    trim = FALSE,
    alpha = 0.2, 
    color = "gray50",
    lwd = 0.5
  ) +
  
  # Box Plot Utama
  geom_boxplot(
    width = 0.25, 
    alpha = 0.7, 
    fill = "white",
    outlier.color = "#FF4500", 
    outlier.shape = 19, 
    color = "black"
  ) +
  
  # Titik Median
  geom_point(
    data = claim_stats_values,
    aes(y = Median_Value, 
        text = Median_Label), 
    shape = 18, 
    size = 4,
    color = "#3498db" 
  ) +
  
  # Titik Mean
  geom_point(
    data = claim_stats_values,
    aes(y = Mean_Value,
        text = Mean_Label),
    shape = 17, 
    size = 4,
    color = "#FF9800" 
  ) +
  
  # Transformasi Skala Y Logaritmik
  scale_y_log10(
    labels = scales::label_dollar(prefix = "Rp", scale = 1000, big.mark = "."), 
    breaks = c(10, 20, 50, 100, 200, 500) 
  ) +
  
  # Label dan Tema
  labs(
    title = "Klaim Berdasarkan Status Perokok",
    subtitle = "Dispersi Klaim Perokok Jauh Lebih Tinggi & Sangat Miring ke Kanan",
    x = "Status Perokok",
    y = "Jumlah Klaim (Rp)"
  ) +
  theme_minimal(base_size = 14) + 
  theme(
    legend.position = "none", 
    plot.title = element_text(hjust = 0.5, face = "bold", size = 18),
    plot.subtitle = element_text(hjust = 0.5, face = "italic", color = "gray50"),
    axis.title.y = element_text(face = "bold")
  ) +
  scale_fill_manual(values = c("Non-Smoker" = "#76d7c4", "Smoker" = "#f1c40f"))


# Konversi ke Plotly dan Anotasi 
p_plotly <- ggplotly(p_smoker, tooltip = "text", width = 800, height = 600) %>%
  layout(hovermode = 'closest') %>%
  config(
    displayModeBar = TRUE, 
    modeBarButtonsToRemove = c('zoomIn2d', 'zoomOut2d', 'autoScale2d', 'toggleSpikelines', 'select2d', 'lasso2d')
  )

p_plotly

# Hitung nilai mean per InsurancePlan
mean_age_by_plan <- df %>%
  group_by(InsurancePlan) %>%
  summarise(Mean_Age = mean(Age, na.rm = TRUE))

ggplot(df, aes(x = InsurancePlan, y = Age, fill = InsurancePlan)) +
  
  # Violin Plot
  geom_violin(
    trim = FALSE, 
    alpha = 0.4, 
    color = "gray40"
  ) +
  
  # Box Plot
  geom_boxplot(
    width = 0.25, 
    color = "black", 
    alpha = 0.8, 
    outlier.shape = NA
  ) +
  
  # Titik Mean 
  geom_point(
    data = mean_age_by_plan, 
    aes(y = Mean_Age, fill = InsurancePlan), 
    shape = 23, 
    size = 3.5, 
    color = "black"
  ) +
  
  # Label Angka Mean
  geom_text(
    data = mean_age_by_plan,
    aes(y = Mean_Age, label = paste0("Mean: ", round(Mean_Age, 1))) ,
    color = "blue", size = 3.5, fontface = "bold",
    vjust = -1.5
  ) +
  
  # Label dan Tema
  labs(
    title = "Distribusi Usia Berdasarkan Rencana Asuransi",
    subtitle = "Angka di atas titik menunjukkan Mean Usia",
    x = "Rencana Asuransi",
    y = "Usia Pemegang Polis (Tahun)"
  ) +
  theme_minimal(base_size = 14) +
  theme(
    axis.text.x = element_text(angle = 60, hjust = 1, face = "bold"),
    legend.position = "none",
    plot.title = element_text(hjust = 0.5, face = "bold"),
    plot.subtitle = element_text(hjust = 0.5, face = "italic", color = "gray50")
  ) +
  scale_fill_brewer(palette = "Set2")

# Mengelompokkan (Binning) variabel Age dengan lebih detail
df_binned <- df %>%
  mutate(
    AgeGroup = case_when(
      Age < 25 ~ "1. Young Adult (< 25 thn)",
      Age >= 25 & Age < 35 ~ "2. Adult (25-34 thn)",
      Age >= 35 & Age < 50 ~ "3. Middle Age (35-49 thn)",
      Age >= 50 & Age < 65 ~ "4. Senior (50-64 thn)",
      Age >= 65 ~ "5. Elderly (≥ 65 thn)",
      TRUE ~ "Lainnya"
    ),
    AgeGroup = factor(AgeGroup, levels = c("1. Young Adult (< 25 thn)", "2. Adult (25-34 thn)", 
                                          "3. Middle Age (35-49 thn)", "4. Senior (50-64 thn)", 
                                          "5. Elderly (≥ 65 thn)"))
  )

# Agregasi Data dengan lebih banyak metrik
age_plan_data <- df_binned %>%
  group_by(AgeGroup) %>%
  summarise(
    Count = n(),
    TotalClaimAmount = sum(ClaimAmount, na.rm = TRUE),
    MeanAge = mean(Age, na.rm = TRUE),
    AvgClaimAmount = mean(ClaimAmount, na.rm = TRUE),
    MaxClaimAmount = max(ClaimAmount, na.rm = TRUE),
    MinClaimAmount = min(ClaimAmount, na.rm = TRUE),
    .groups = 'drop'
  ) %>%
  # Hitung berbagai persentase dan metrik
  mutate(
    Percentage = Count / sum(Count) * 100,
    ClaimPercentage = TotalClaimAmount / sum(TotalClaimAmount) * 100,
    TotalClaimFormatted = paste0("Rp", format(round(TotalClaimAmount/1000, 0), big.mark = ".", scientific = FALSE), "K"),
    AvgClaimFormatted = paste0("Rp", format(round(AvgClaimAmount/1000, 0), big.mark = ".", scientific = FALSE), "K"),
    TooltipText = paste0(
      "<b>", AgeGroup, "</b><br>",
      "Jumlah Polis: <b>", Count, "</b> (", round(Percentage, 1), "%)<br>",
      "Total Klaim: <b>", TotalClaimFormatted, "</b><br>",
      "Rata-rata Klaim: ", AvgClaimFormatted, "<br>",
      "Klaim Tertinggi: Rp", format(round(MaxClaimAmount/1000, 0), big.mark = ".", scientific = FALSE), "K<br>",
      "Rata-rata Usia: ", round(MeanAge, 1), " tahun"
    )
  )

# Warna custom yang lebih menarik
custom_colors <- c('#1f77b4', '#ff7f0e', '#2ca02c', '#d62728', '#9467bd', '#8c564b')

# Buat Donut Chart Interaktif dengan Plotly
age_donut_interactive <- plot_ly(
  age_plan_data,
  labels = ~AgeGroup,
  values = ~Count,
  width = 800,
  height = 500,
  type = 'pie',
  hole = 0.6,
  hoverinfo = 'text',
  text = ~TooltipText,
  textinfo = 'percent',
  textposition = 'inside',
  insidetextorientation = 'radial',
  marker = list(
    colors = custom_colors,
    line = list(color = '#FFFFFF', width = 2),
    pattern = list(
      shape = c("/", "x", "-", "\\", "|", "+"),
      solidity = 0.7,
      fgcolor = "white",
      fgopacity = 0.5
    )
  ),
  sort = FALSE,
  direction = 'clockwise',
  rotation = -90
) %>%
layout(
  title = list(
    text = '<b>DISTRIBUSI PEMEGANG POLIS BERDASARKAN KELOMPOK USIA</b>',
    x = 0.5,
    y = 0.98,
    font = list(size = 20, family = "Arial", color = '#2c3e50')
  ),
  font = list(family = "Arial", size = 12),
  showlegend = TRUE,
  legend = list(
    orientation = "v",
    x = 1.05,
    y = 0.5,
    bgcolor = 'rgba(255,255,255,0.8)',
    bordercolor = '#CCCCCC',
    borderwidth = 1,
    font = list(size = 12, family = "Arial"),
    itemclick = 'toggle',
    itemdoubleclick = 'toggleothers'
  ),
  
  # Annotations untuk ringkasan di tengah donut
  annotations = list(
    list(
      text = paste0(
        "TOTAL<br>",
        "<b>", sum(age_plan_data$Count), "</b><br>",
        "POLIS"
      ),
      x = 0.5, y = 0.5,
      font = list(size = 18, color = '#2c3e50', family = "Arial"),
      showarrow = FALSE,
      xref = 'paper', yref = 'paper'
    )
  ),
  margin = list(l = 50, r = 150, b = 50, t = 100),
  paper_bgcolor = 'rgba(0,0,0,0)',
  plot_bgcolor = 'rgba(0,0,0,0)'
) %>%
  
# Tambahkan dropdown menu untuk metrik berbeda
layout(
  updatemenus = list(
    list(
      type = "dropdown",
      x = 1.15, y = 0.9,
      buttons = list(
        list(
          method = "restyle",
          args = list("values", list(age_plan_data$Count)),
          label = "Jumlah Polis"
        ),
        list(
          method = "restyle",
          args = list("values", list(age_plan_data$TotalClaimAmount)),
          label = "Total Klaim"
        ),
        list(
          method = "restyle",
          args = list("values", list(age_plan_data$AvgClaimAmount)),
          label = "Rata-rata Klaim"
        )
      ),
      bgcolor = '#f8f9fa',
      bordercolor = '#dee2e6',
      font = list(size = 11)
    )
  )
) %>%
  
# Konfigurasi mode bar
config(
  displayModeBar = TRUE,
  modeBarButtonsToAdd = list(
    "drawline",
    "drawopenpath",
    "drawcircle",
    "drawrect",
    "eraseshape"
  ),
  modeBarButtonsToRemove = c(
    'sendDataToCloud', 'hoverClosestPie', 'toggleSpikelines',
    'select2d', 'lasso2d', 'zoomIn2d', 'zoomOut2d', 'autoScale2d'
  ),
  scrollZoom = TRUE,
  displaylogo = FALSE
)

# efek highlight on click
age_donut_interactive <- age_donut_interactive %>%
  highlight(
    on = "plotly_click",
    off = "plotly_doubleclick",
    persistent = TRUE,
    dynamic = TRUE,
    selectize = FALSE,
    color = "yellow",
    opacityDim = 0.1
  )

# event handler untuk interaksi tambahan
age_donut_interactive <- age_donut_interactive %>%htmlwidgets::onRender("
    function(el, x) {
      var id = el.getAttribute('id');
      var gd = document.getElementById(id);
      
      // Custom hover effects
      gd.on('plotly_hover', function(data) {
        var annotations = [];
        if(data.points && data.points.length > 0) {
          var point = data.points[0];
          annotations.push({
            text: '<b>' + point.label + '</b><br>' + point.value + ' polis',
            x: 0.5,
            y: 0.3,
            xref: 'paper',
            yref: 'paper',
            showarrow: false,
            font: {size: 14, color: point.fullData.marker.colors[point.i]}
          });
        }
        Plotly.relayout(gd, {annotations: annotations});
      });
      
      // Clear annotation on unhover
      gd.on('plotly_unhover', function(data) {
        Plotly.relayout(gd, {
          annotations: [{
            text: 'TOTAL<br><b>' + x.data[0].values.reduce((a, b) => a + b, 0) + '</b><br>POLIS',
            x: 0.5,
            y: 0.5,
            xref: 'paper',
            yref: 'paper',
            showarrow: false,
            font: {size: 18, color: '#2c3e50'}
          }]
        });
      });
    }
  ")

# Tampilkan plot
age_donut_interactive

# Hitung Koefisien Korelasi Pearson dan statistik lainnya
correlation_value <- cor(df$BMI, df$RiskScore, use = "complete.obs")
cor_test <- cor.test(df$BMI, df$RiskScore)
n_obs <- nrow(na.omit(df[, c("BMI", "RiskScore")]))

# Buat model untuk mendapatkan persamaan garis
lm_model <- lm(RiskScore ~ BMI, data = df)
lm_eq <- paste0("y = ", round(coef(lm_model)[1], 2), " + ", round(coef(lm_model)[2], 2), "x")
r_squared <- round(summary(lm_model)$r.squared, 3)

# Plot Dasar
bmi_risk_ggplot <- ggplot(df, aes(x = BMI, y = RiskScore, color = ExerciseHours)) +
  
  # Scatter Plot Utama dengan size berdasarkan ExerciseHours
  geom_point(
    alpha = 0.7,
    size = 4,
    aes(text = paste0(
      "ID: ", rownames(df), "<br>",
      "BMI: ", round(BMI, 1), "<br>",
      "Risk Score: ", round(RiskScore, 1), "<br>",
      "Jam Olahraga: ", ExerciseHours, " jam/minggu<br>",
      "Kategori BMI: ", cut(BMI, breaks = c(0, 18.5, 25, 30, Inf), 
                          labels = c("Underweight", "Normal", "Overweight", "Obese"))
    ))
  ) +
  
  # Garis Tren Linear (LM)
  geom_smooth(
    method = "lm", 
    se = TRUE, 
    alpha = 0.2, 
    color = "#292B2C", 
    linewidth = 1.2,
    linetype = "solid",
    aes(fill = "Linear Trend")
  ) +
  
  # Garis Tren LOESS 
  geom_smooth(
    method = "loess", 
    se = TRUE, 
    alpha = 0.2, 
    color = "#D9534F", 
    linetype = "dashed", 
    linewidth = 1,
    aes(fill = "Local Trend")
  ) +
  
  # Scale untuk fill
  scale_fill_manual(
    name = "Trend Lines",
    values = c("Linear Trend" = "#292B2C", "Local Trend" = "#D9534F")
  ) +

  # Kustomisasi Gradien Warna
  scale_color_gradientn(
    colors = c("#5BC0DE", "#292B2C", "#D9534F"),  
    name = "Exercise Hours\n(per week)",
    breaks = seq(0, max(df$ExerciseHours, na.rm = TRUE), by = 2)
  ) +
  
  # Labels dan tema
  labs(
    title = paste0("BMI vs Risk Score Correlation (r = ", round(correlation_value, 2), ")"),
    subtitle = paste0("Linear: ", lm_eq, " | R² = ", r_squared, " | p-value = ", round(cor_test$p.value, 4)),
    x = "Body Mass Index (BMI)",
    y = "Risk Score",
    caption = paste("N =", n_obs, "observations | Click and drag to zoom | Double-click to reset")
  ) +
  
  theme_minimal(base_size = 14) +
  theme(
    plot.title = element_text(hjust = 0.5, face = "bold", size = 16),
    plot.subtitle = element_text(hjust = 0.5, size = 12),
    legend.position = "right",
    panel.grid.minor = element_blank()
  )

# Konversi ke Plotly
bmi_risk_interactive <- ggplotly(bmi_risk_ggplot, tooltip = "text", width = 800, height = 600) %>%
  layout(
    title = list(
      text = paste0(
        "BMI vs Risk Score Correlation<br>",
        "<sub>r = ", round(correlation_value, 2), 
        " | p = ", round(cor_test$p.value, 4),
        " | N = ", n_obs, "</sub>"
      ),
      x = 0.05
    ),
    hovermode = 'closest',
    xaxis = list(
      title = "Body Mass Index (BMI)",
      spikemode = "across",
      showspikes = TRUE,
      spikethickness = 1,
      spikecolor = "#999999"
    ),
    yaxis = list(
      title = "Risk Score",
      spikemode = "across", 
      showspikes = TRUE,
      spikethickness = 1,
      spikecolor = "#999999"
    ),
    annotations = list(
      list(
        x = 1, y = 0, 
        text = "Source: Your Dataset",
        showarrow = FALSE, 
        xref = "paper", yref = "paper",
        xanchor = "right", yanchor = "bottom",
        font = list(size = 10, color = "gray")
      )
    )
  ) %>%
  config(
    displayModeBar = TRUE,
    modeBarButtonsToAdd = list(
      "drawline",
      "drawopenpath",
      "drawcircle",
      "drawrect",
      "eraseshape"
    ),
    modeBarButtonsToRemove = c(
      'zoomIn2d', 'zoomOut2d', 'autoScale2d', 
      'toggleSpikelines', 'select2d', 'lasso2d'
    ),
    scrollZoom = TRUE,
    displaylogo = FALSE
  )

# range slider untuk BMI
bmi_risk_interactive <- bmi_risk_interactive %>%
  layout(
    xaxis = list(
      rangeslider = list(
        type = "linear",
        thickness = 0.08,
        bgcolor = "lightgray"
      )
    )
  )

# fitur highlight on click
bmi_risk_interactive <- bmi_risk_interactive %>%
  highlight(
    on = "plotly_click",
    off = "plotly_doubleclick",
    color = "red",
    opacityDim = 0.3
  )

# fitur selection
bmi_risk_interactive <- bmi_risk_interactive %>%
  layout(
    dragmode = "select"
  ) %>%
  event_register("plotly_selected")

# Function untuk menangani event selection
observe_selection <- function(p) {
  p %>% htmlwidgets::onRender("
    function(el, x) {
      var id = el.getAttribute('id');
      var gd = document.getElementById(id);
      
      gd.on('plotly_click', function(data) {
        var point = data.points[0];
        var customText = 'Clicked: BMI=' + point.x + ', Risk=' + point.y;
        console.log(customText);
      });
      
      gd.on('plotly_selected', function(data) {
        var selectedPoints = data.points;
        console.log('Selected points:', selectedPoints.length);
      });
    }
  ")
}

# observasi selection
bmi_risk_interactive <- observe_selection(bmi_risk_interactive)

# Tampilkan plot
bmi_risk_interactive