---
title: "Flex Dashboards QC Seven Tools"
output: 
  flexdashboard::flex_dashboard:
    vertical_layout: scroll
    theme: yeti
    source_code: embed
---

```{r setup, include=FALSE}
library(dplyr)
library(RSQLite)
library(flexdashboard)
library(tidyverse)
library(tidyr)
library(highcharter) #Buat Visual
library(gt)
library(htmltools)
library(DT)
library(readr)
library(stringr)
library(qcc)
library(DiagrammeR)
library(moments)
library(viridis)
library(forcats)
library(DBI)
library(plotly)
library(glue)
```

Data
======================================================================

Column {.tabset .tabset-fade data-width=1000}
-----------------------------------------------------------------------

```{r}
library(DBI)
library(RSQLite)
library(DT)

# Koneksi ke database (sesuaikan nama file SQLite)
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Cek daftar tabel
dbListTables(con)  # Ini untuk verifikasi jika "survey" memang ada

# Ambil seluruh isi tabel survey
data_survey <- dbReadTable(con, "survey")

# Tutup koneksi setelah selesai ambil data
dbDisconnect(con)

# Tampilkan dengan datatable interaktif scrollable
DT::datatable(
  data_survey,
  options = list(
    pageLength = 10,       
    scrollY = "400px",     
    scrollX = TRUE         
  ),
  caption = htmltools::tags$caption(
    style = 'caption-side: top; text-align: left; font-weight: bold;',
    "Tabel Data Survei"
  )
)

```


Check Sheet
======================================================================

Column {.tabset .tabset-fade data-width=500}
-----------------------------------------------------------------------

### Frekuensi Hambatan dalam Menjalani Hidup Sehat

```{r}
# Koneksi ke SQLite
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Query produk favorit
df <- dbGetQuery(con, "SELECT produk_favorit FROM survey")

# Hitung frekuensi dan susun
check_sheet <- df %>%
  group_by(produk_favorit) %>%
  summarise(Jumlah = n()) %>%
  arrange(desc(Jumlah))

# Tampilkan dalam tabel interaktif
DT::datatable(
  check_sheet,
  options = list(pageLength = 5),
  caption = "Frekuensi Produk Favorit"
)

# Tutup koneksi database
dbDisconnect(con)
```

Column {.tabset .tabset-fade data-width=500}
---------------------------------------------------------------------

### Visualisasi Pie Chart Hambatan dalam Menjalani Hidup Sehat

```{r}
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

data_produk_fav <- dbGetQuery(con, "SELECT produk_favorit FROM survey")

data_pie <- data_produk_fav %>%
  group_by(produk_favorit) %>%
  summarise(Jumlah = n()) %>%
  mutate(Persentase = round(Jumlah / sum(Jumlah) * 100, 2)) %>%
  arrange(desc(Jumlah))

n_colors <- nrow(data_pie)
custom_colors <- viridis::inferno(n = n_colors)

hc <- highchart() %>%
  hc_chart(type = "pie") %>%
  hc_title(text = "Distribusi Produksi Favorit") %>%
  hc_subtitle(text = "Berdasarkan Frekuensi Responden") %>%
  hc_add_series(
    name = "Produk",
    data = list_parse(data.frame(
      name = data_pie$produk_favorit,
      y = data_pie$Jumlah,
      color = custom_colors
    ))
  ) %>%
  hc_tooltip(pointFormat = "<b>{point.name}</b>: {point.y} responden ({point.percentage:.2f}%)") %>%
  hc_plotOptions(
    pie = list(
      allowPointSelect = TRUE,
      cursor = "pointer",
      innerSize = "50%",
      dataLabels = list(
        enabled = TRUE,
        format = "<b>{point.name}</b>: {point.percentage:.2f}%"
      )
    )
  )

hc

dbDisconnect(con)
```

Histogram
======================================================================

Column {data-width=650}
-----------------------------------------------------------------------
### Berapa rata-rata pengeluaran bulanan mahasiswa di Shopee?

```{r}
# Koneksi ke database
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil data pengeluaran
df_hist <- dbGetQuery(con, "SELECT pengeluaran_bulanan FROM survey")

# Bersihkan: pastikan kolom numerik dan tidak NA
df_hist <- df_hist %>%
  filter(!is.na(pengeluaran_bulanan)) %>%
  mutate(pengeluaran_bulanan = as.numeric(pengeluaran_bulanan))

# Tampilkan 10 data pertama untuk pengecekan
DT::datatable(
  head(df_hist, 10),
  caption = "Sampel Data Pengeluaran Bulanan"
)

# Tutup koneksi
dbDisconnect(con)
```

Column {data-width=650}
-----------------------------------------------------------------------

### Visualisasi rata-rata pengeluaran mahasiswa

```{r}
# Koneksi ke database
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil data pengeluaran
df_hist <- dbGetQuery(con, "SELECT pengeluaran_bulanan FROM survey") %>%
  filter(!is.na(pengeluaran_bulanan)) %>%
  mutate(pengeluaran_bulanan = as.numeric(pengeluaran_bulanan))

dbDisconnect(con)

# Buat breaks dan label custom
breaks <- c(0, 500000, 1500000, 2500000, 5000000, Inf)
labels <- c("Rp100rb–500rb", "Rp501rb–1,5jt", "Rp1,5jt–2,5jt", "Rp2,5jt–5jt", "> Rp5jt")

# Histogram binned
df_hist_binned <- df_hist %>%
  mutate(bin = cut(pengeluaran_bulanan, breaks = breaks, labels = labels, include.lowest = TRUE)) %>%
  group_by(bin) %>%
  summarise(Jumlah = n())

# Hitung density keseluruhan (tanpa bin)
density_data <- density(df_hist$pengeluaran_bulanan)

# Normalisasi density agar cocok skala histogram
density_scaled <- density_data$y * max(df_hist_binned$Jumlah) / max(density_data$y)

# Visualisasi
highchart() %>%
  hc_chart(type = "column") %>%
  hc_title(text = "Distribusi Pengeluaran Bulanan Mahasiswa") %>%
  hc_xAxis(
    categories = df_hist_binned$bin,
    title = list(text = "Rentang Pengeluaran")
  ) %>%
  hc_yAxis_multiples(
    list(title = list(text = "Jumlah Responden"), opposite = FALSE),
    list(title = list(text = "Kepadatan (Density)"), opposite = TRUE)
  ) %>%
  hc_add_series(
    name = "Responden",
    data = df_hist_binned$Jumlah,
    type = "column",
    yAxis = 0,
    colorByPoint = TRUE
  ) %>%
  hc_add_series(
  name = "Kurva Density",
  type = "spline",
  data = list_parse2(data.frame(
    x = scales::rescale(density_data$x, to = c(0, length(labels) - 1)),
    y = round(density_scaled, 2)
  )),
  yAxis = 0,
  color = "#E74C3C",          # 🔴 Merah
  dashStyle = "Dash",         # ➖ Putus-putus
  marker = list(enabled = FALSE),
  tooltip = list(pointFormat = "<b>Kepadatan</b>: {point.y:.2f}")
)

```



### Hubungan Diskon yang Ditunggu dengan Frekuensi Pembelian

```{r}
library(dplyr)
library(DBI)
library(RSQLite)
library(highcharter)

# Koneksi database
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil data dua variabel
df_scatter <- dbGetQuery(con, "SELECT persen_diskon_tunggu, frekuensi_pembelian FROM survey") %>%
  filter(!is.na(persen_diskon_tunggu), !is.na(frekuensi_pembelian)) %>%
  mutate(
    persen_diskon_tunggu = as.numeric(persen_diskon_tunggu),
    frekuensi_pembelian = as.numeric(frekuensi_pembelian)
  )

dbDisconnect(con)

# Format data ke list of lists
data_points <- df_scatter %>%
  transmute(x = persen_diskon_tunggu, y = frekuensi_pembelian) %>%
  list_parse()

# Scatter plot
highchart() %>%
  hc_chart(type = "scatter") %>%
  hc_title(text = "Hubungan Diskon yang Ditunggu dengan Frekuensi Pembelian") %>%
  hc_xAxis(title = list(text = "Persen Diskon yang Ditunggu (%)")) %>%
  hc_yAxis(title = list(text = "Frekuensi Pembelian")) %>%
  hc_add_series(
    data = data_points,
    type = "scatter",
    name = "Responden",
    marker = list(symbol = "circle", radius = 5)
  ) %>%
  hc_tooltip(pointFormat = "Diskon Ditunggu: <b>{point.x}%</b><br>Frekuensi: <b>{point.y}</b>")
```

Scatter Plot
======================================================================

Column
----------------------------------------------------------------------

### Hubungan Persen Diskon Tunggu dengan Frekuensi Pembelian

```{r scatter-frekuensi-diskon, warning=FALSE, message=FALSE}
library(DBI)
library(RSQLite)
library(dplyr)
library(highcharter)

# Koneksi ke database SQLite
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil data dari tabel 'survey'
df_scatter <- dbGetQuery(con, "
  SELECT persen_diskon_tunggu, frekuensi_pembelian
  FROM survey
  WHERE persen_diskon_tunggu IS NOT NULL AND frekuensi_pembelian IS NOT NULL
")

# Tutup koneksi
dbDisconnect(con)

# Siapkan data dan pastikan numerik
df_scatter <- df_scatter %>%
  mutate(
    persen_diskon_tunggu = as.numeric(persen_diskon_tunggu),
    frekuensi_pembelian = as.numeric(frekuensi_pembelian),
    Responden = paste("Responden", row_number())
  )

# Titik data
scatter_data <- df_scatter %>%
  transmute(
    name = Responden,
    x = persen_diskon_tunggu,
    y = frekuensi_pembelian
  ) %>%
  list_parse()

# Regresi linier
model <- lm(frekuensi_pembelian ~ persen_diskon_tunggu, data = df_scatter)
x_seq <- seq(min(df_scatter$persen_diskon_tunggu), max(df_scatter$persen_diskon_tunggu), length.out = 100)
y_pred <- predict(model, newdata = data.frame(persen_diskon_tunggu = x_seq))
line_data <- data.frame(x = x_seq, y = y_pred) %>% list_parse()

# Visualisasi dengan highcharter
highchart() %>%
  hc_chart(type = "scatter") %>%
  hc_title(text = "Scatter Plot: Persen Diskon Tunggu vs Frekuensi Pembelian") %>%
  hc_xAxis(title = list(text = "Persen Diskon Tunggu (%)")) %>%
  hc_yAxis(title = list(text = "Frekuensi Pembelian (per Bulan)")) %>%
  hc_add_series(data = scatter_data, type = "scatter", name = "Responden", color = "#1abc9c") %>%
  hc_add_series(data = line_data, type = "line", name = "Regresi Linear", color = "#e74c3c", dashStyle = "ShortDash") %>%
  hc_tooltip(pointFormat = "<b>{point.name}</b><br>Diskon Tunggu: {point.x}%<br>Frekuensi: {point.y}") %>%
  hc_add_theme(hc_theme_gridlight())
```

Line Chart
======================================================================

Column
----------------------------------------------------------------------

### Perbandingan Pentingnya Layanan Pelanggan dan Skor Kepuasan

```{r linechart-penting-kepuasan, warning=FALSE, message=FALSE}
# Koneksi ke SQLite
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil data dari tabel 'survey'
df_line <- dbGetQuery(con, "
  SELECT penting_layanan_pelanggan, skor_kepuasan
  FROM survey
  WHERE penting_layanan_pelanggan IS NOT NULL AND skor_kepuasan IS NOT NULL
")

# Tutup koneksi
dbDisconnect(con)

# Pastikan data numerik dan beri urutan
df_line <- df_line %>%
  mutate(
    penting_layanan_pelanggan = as.numeric(penting_layanan_pelanggan),
    skor_kepuasan = as.numeric(skor_kepuasan),
    urutan = row_number()
  )

# Data untuk dua garis
data_penting <- df_line %>%
  transmute(x = urutan, y = penting_layanan_pelanggan) %>%
  list_parse()

data_kepuasan <- df_line %>%
  transmute(x = urutan, y = skor_kepuasan) %>%
  list_parse()

# Highchart line plot
highchart() %>%
  hc_chart(type = "line") %>%
  hc_title(text = "Perbandingan Pentingnya Layanan vs Skor Kepuasan") %>%
  hc_xAxis(title = list(text = "Responden"), allowDecimals = FALSE) %>%
  hc_yAxis(title = list(text = "Skor (1–10)")) %>%
  hc_add_series(data = data_penting, name = "Pentingnya Layanan", color = "#2980B9") %>%
  hc_add_series(data = data_kepuasan, name = "Skor Kepuasan", color = "#E67E22") %>%
  hc_tooltip(shared = TRUE, crosshairs = TRUE, valueDecimals = 1) %>%
  hc_add_theme(hc_theme_flat())
```

Pie Chart
======================================================================

Column {.tabset .tabset-fade data-width=500}
----------------------------------------------------------------------

### Frekuensi Produk Favorit

```{r}
# Koneksi ke database
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil data check sheet dari SQL
check_sheet <- dbGetQuery(con, "
  SELECT produk_favorit, COUNT(*) AS Jumlah
  FROM survey
  WHERE produk_favorit IS NOT NULL
  GROUP BY produk_favorit
  ORDER BY Jumlah DESC
")

# Tutup koneksi
dbDisconnect(con)

# Tampilkan datatable
datatable(
  check_sheet,
  options = list(pageLength = 5),
  caption = "Frekuensi Produk Favorit Responden Shopee"
)
```

Column {.tabset .tabset-fade data-width=500}
----------------------------------------------------------------------

### Distribusi Produk Favorit Responden

```{r}
library(viridis)

con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

data_produk_fav <- dbGetQuery(con, "SELECT produk_favorit FROM survey")

dbDisconnect(con)

data_pie <- data_produk_fav %>%
  group_by(produk_favorit) %>%
  summarise(Jumlah = n()) %>%
  mutate(Persentase = round(Jumlah / sum(Jumlah) * 100, 2)) %>%
  arrange(desc(Jumlah))

n_colors <- nrow(data_pie)
custom_colors <- viridis::inferno(n = n_colors)

highchart() %>%
  hc_chart(type = "pie") %>%
  hc_title(text = "Distribusi Produk Favorit") %>%
  hc_subtitle(text = "Berdasarkan Frekuensi Responden") %>%
  hc_add_series(
    name = "Produk",
    data = list_parse(data.frame(
      name = data_pie$produk_favorit,
      y = data_pie$Jumlah,
      color = custom_colors
    ))
  ) %>%
  hc_tooltip(pointFormat = "<b>{point.name}</b>: {point.y} responden ({point.percentage:.2f}%)") %>%
  hc_plotOptions(
    pie = list(
      allowPointSelect = TRUE,
      cursor = "pointer",
      innerSize = "50%",
      dataLabels = list(
        enabled = TRUE,
        format = "<b>{point.name}</b>: {point.percentage:.2f}%"
      )
    )
  )
```


Column {.tabset .tabset-fade data-width=500}
----------------------------------------------------------------------

### Produk Favorit
```{r}
# Koneksi ke database
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil dan olah data dari SQL
data_pie <- dbGetQuery(con, "
  SELECT produk_favorit, COUNT(*) AS Jumlah
  FROM survey
  WHERE produk_favorit IS NOT NULL
  GROUP BY produk_favorit
  ORDER BY Jumlah DESC
") %>%
  mutate(Persentase = round(Jumlah / sum(Jumlah) * 100, 2))

# Tutup koneksi
dbDisconnect(con)

# Warna untuk tiap kategori
n_colors <- nrow(data_pie)
custom_colors <- viridis::inferno(n = n_colors)

# Buat donut chart
hc <- highchart() %>%
  hc_chart(type = "pie") %>%
  hc_title(text = "Distribusi Produk Favorit Responden Shopee") %>%
  hc_subtitle(text = "Berdasarkan Pilihan Responden") %>%
  hc_add_series(
    name = "Produk Favorit",
    data = list_parse(data.frame(
      name = data_pie$produk_favorit,
      y = data_pie$Jumlah,
      color = custom_colors
    ))
  ) %>%
  hc_tooltip(
    pointFormat = "<b>{point.name}</b>: {point.y} responden ({point.percentage:.2f}%)"
  ) %>%
  hc_plotOptions(
    pie = list(
      allowPointSelect = TRUE,
      cursor = "pointer",
      innerSize = "50%",
      dataLabels = list(
        enabled = TRUE,
        format = "<b>{point.name}</b>: {point.percentage:.2f}%"
      )
    )
  )

hc
```

Pareto Chart
======================================================================

Column
----------------------------------------------------------------------

### Sumber Informasi Produk saat berbelanja

```{r}
# Koneksi ke database
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil data sumber_info_produk dari database
pareto_data <- dbGetQuery(con, "
  SELECT sumber_info_produk, COUNT(*) AS n
  FROM survey
  WHERE sumber_info_produk IS NOT NULL
  GROUP BY sumber_info_produk
  ORDER BY n DESC
") %>%
  mutate(
    cum_freq = cumsum(n) / sum(n) * 100
  )

dbDisconnect(con)

# Warna
colors <- colorRampPalette(RColorBrewer::brewer.pal(12, "Set3"))(nrow(pareto_data))

# Buat plot Pareto
fig <- plot_ly()

# Bar chart
fig <- fig %>% add_bars(
  x = ~reorder(pareto_data$sumber_info_produk, -pareto_data$n),
  y = ~pareto_data$n,
  name = 'Jumlah Responden',
  marker = list(color = colors),
  yaxis = "y1"
)

# Garis kumulatif
fig <- fig %>% add_lines(
  x = ~reorder(pareto_data$sumber_info_produk, -pareto_data$n),
  y = ~pareto_data$cum_freq,
  name = 'Kumulatif (%)',
  yaxis = "y2",
  line = list(color = 'red', dash = 'dash')
)

# Garis batas 80%
fig <- fig %>% add_lines(
  x = ~reorder(pareto_data$sumber_info_produk, -pareto_data$n),
  y = rep(80, nrow(pareto_data)),
  name = 'Cut-off 80%',
  yaxis = "y2",
  line = list(color = 'green', dash = 'dot')
)

# Layout
fig <- fig %>% layout(
  title = "Pareto Chart – Sumber Informasi Produk yang Paling Berpengaruh",
  xaxis = list(
    title = "Sumber Informasi Produk",
    tickangle = -45,  # Miringkan label
    tickfont = list(size = 11),
    automargin = TRUE
  ),
  yaxis = list(
    title = "Jumlah Responden"
  ),
  yaxis2 = list(
    title = "Kumulatif (%)",
    overlaying = "y",
    side = "right",
    range = c(0, 100)
  ),
  legend = list(
    x = 0.5,
    y = -0.2,
    orientation = "h",
    xanchor = "center"
  ),
  margin = list(b = 100)  # Tambahkan margin bawah agar label tidak terpotong
)

## Tampilan Plot
fig

```


Fishbone
======================================================================

Column
----------------------------------------------------------------------

### Penyebab Kepuasan Layanan Shopee Rendah

```{r}
library(DBI)
library(RSQLite)
library(DiagrammeR)

# Koneksi ke database
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")
data_fishbone <- dbReadTable(con, "fishbone")
dbDisconnect(con)

# Buat node unik
nodes <- unique(c(data_fishbone$kategori, data_fishbone$sub_penyebab, "KEPUASAN LAYANAN SHOPEE RENDAH"))

# Buat warna per kategori
warna_kategori <- c(
  "MANPOWER" = "lightblue",
  "METHOD" = "lightgreen",
  "MACHINE" = "khaki",
  "MATERIAL" = "plum",
  "MOTHER_NATURE" = "orange",
  "MEASUREMENT" = "lightsalmon"
)

# Buat graph script dalam format DOT
edges <- ""
for (i in 1:nrow(data_fishbone)) {
  kategori <- data_fishbone$kategori[i]
  penyebab <- data_fishbone$sub_penyebab[i]
  edges <- paste0(edges, "\"", penyebab, "\" -> \"", kategori, "\";\n")
}

# Tambahkan koneksi kategori ke masalah utama
kategori_unik <- unique(data_fishbone$kategori)
for (k in kategori_unik) {
  edges <- paste0(edges, "\"", k, "\" -> \"KEPUASAN LAYANAN SHOPEE RENDAH\";\n")
}

# Tambahkan style node
style <- ""
for (k in kategori_unik) {
  style <- paste0(style, "\"", k, "\" [style=filled, fillcolor=", warna_kategori[[k]], "];\n")
}
style <- paste0(style, "\"KEPUASAN LAYANAN SHOPEE RENDAH\" [style=filled, shape=ellipse, fillcolor=deeppink, fontcolor=white];\n")

# Gabungkan
script <- paste0("digraph fishbone {\nrankdir=LR;\nnode [shape=box];\n", edges, style, "}")

# Tampilkan
grViz(script)

```

```{r}
# Koneksi ke DB
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil data fishbone
fish_data <- dbGetQuery(con, "SELECT * FROM fishbone")

# Tutup koneksi
dbDisconnect(con)

# Kelompokkan berdasarkan kategori
nodes <- fish_data %>%
  group_by(kategori) %>%
  summarise(subs = paste0("  ", kategori, tolower(letters[1:length(sub_penyebab)]), 
                          " [label='", sub_penyebab, "']\n",
                          kategori, tolower(letters[1:length(sub_penyebab)]), 
                          " -> ", kategori, "\n", collapse = ""))

# Bangun bagian diagram fishbone secara dinamis
sub_nodes_code <- paste(nodes$subs, collapse = "\n")

# Buat template diagram
diagram_code <- glue('
digraph fishbone {{
  graph [layout = dot, rankdir = LR]
  node [fontname=Helvetica, fontsize=20, style=filled]

  Problem [label="KEPUASAN LAYANAN SHOPEE RENDAH", shape=ellipse, fillcolor=deeppink, width=6.5, height=1.2]

  node [shape=hexagon, width=2.4, height=0.9]
  {paste(unique(fish_data$kategori), collapse = " -> Problem\n")} -> Problem

  node [shape=rectangle, width=3, height=0.5, fillcolor=lightgray]
  {sub_nodes_code}
}}
')

# Tampilkan diagram
grViz(diagram_code)
```

Flowchart
======================================================================

Column
----------------------------------------------------------------------

### Hubungan diskon, pelayanan, dan pengalaman negatif terhadap kepuasan shopee

```{r}
cols <- viridis(6)

# Koneksi database
con <- dbConnect(SQLite(), "data-survei-2025-06-13 (2).sqlite")

# Ambil ringkasan data
summary_data <- list(
  penting_diskon = dbGetQuery(con, "
    SELECT AVG(CASE WHEN penting_diskon = 'ya' THEN 1 ELSE 0 END) as rata
    FROM survey")$rata,

  tunggu_diskon_tinggi = dbGetQuery(con, "
    SELECT AVG(CASE WHEN persen_diskon_tunggu > 50 THEN 1 ELSE 0 END) as rata
    FROM survey")$rata,

  penting_pelayanan = dbGetQuery(con, "
    SELECT AVG(CASE WHEN penting_layanan_pelanggan > 7 THEN 1 ELSE 0 END) as rata
    FROM survey")$rata,

  pengalaman_buruk_tinggi = dbGetQuery(con, "
    SELECT AVG(CASE WHEN pengaruh_pengalaman_negatif > 6 THEN 1 ELSE 0 END) as rata
    FROM survey")$rata
)

dbDisconnect(con)

# Konversi ke jawaban teks
check <- function(x) ifelse(x > 0.5, "Ya", "Tidak")

grViz(glue("
digraph flow_shopee {{
  graph [layout = dot, rankdir = TB, nodesep=0.6, ranksep=0.9]
  node [style = filled, fontname = Helvetica, fontsize = 14, penwidth = 1.2]

  Start  [label = 'Mulai', shape = ellipse, fillcolor = '#FFFFFF']
  End    [label = 'Selesai', shape = ellipse, fillcolor = '#FFFFFF']

  CekDiskon  [label = 'Diskon Penting?\\n({check(summary_data$penting_diskon)})', shape = diamond, fillcolor = '{cols[5]}']
  DiskonTinggi [label = 'Diskon Ditunggu > 50%?\\n({check(summary_data$tunggu_diskon_tinggi)})', shape = diamond, fillcolor = '{cols[3]}']
  PentingPelayanan [label = 'Pelayanan Penting?\\n({check(summary_data$penting_pelayanan)})', shape = diamond, fillcolor = '{cols[4]}']
  PengalamanBuruk [label = 'Pengalaman Negatif > 6?\\n({check(summary_data$pengalaman_buruk_tinggi)})', shape = diamond, fillcolor = '{cols[6]}']

  TidakPuas [label = 'Resiko Tidak Puas Tinggi', shape = rectangle, fillcolor = '#FFCCCC']
  CukupPuas [label = 'Puas Tapi Tidak Optimal', shape = rectangle, fillcolor = '#FFE066']
  SangatPuas [label = 'Sangat Puas & Loyal', shape = rectangle, fillcolor = '#B2F7B8']

  Start -> CekDiskon

  CekDiskon -> DiskonTinggi [label = 'Ya']
  CekDiskon -> PentingPelayanan [label = 'Tidak']

  DiskonTinggi -> PentingPelayanan [label = 'Ya']
  DiskonTinggi -> SangatPuas [label = 'Tidak']

  PentingPelayanan -> PengalamanBuruk [label = 'Ya']
  PentingPelayanan -> SangatPuas [label = 'Tidak']

  PengalamanBuruk -> TidakPuas [label = 'Ya']
  PengalamanBuruk -> CukupPuas [label = 'Tidak']

  TidakPuas -> End
  CukupPuas -> End
  SangatPuas -> End
}}
"))
```