# ============================================================
# TASK 1: Dynamic Multi-Formula Function
# ============================================================

compute_formula <- function(x, formula) {
  # Validasi input formula
  valid_formulas <- c("linear", "quadratic", "cubic", "exponential")
  if (!(formula %in% valid_formulas)) {
    stop(paste("❌ Formula tidak valid:", formula,
               "\nPilih dari:", paste(valid_formulas, collapse = ", ")))
  }
  
  # Hitung berdasarkan jenis formula
  result <- switch(formula,
    "linear"      = 2 * x + 1,              # f(x) = 2x + 1
    "quadratic"   = x^2 + 3*x + 2,          # f(x) = x² + 3x + 2
    "cubic"       = x^3 - 2*x^2 + x,        # f(x) = x³ - 2x² + x
    "exponential" = exp(0.3 * x)             # f(x) = e^(0.3x)
  )
  return(result)
}

# ── Nested loop: hitung semua formula untuk x = 1:20 ──
x_values <- 1:20
formulas <- c("linear", "quadratic", "cubic", "exponential")

# Buat data frame menggunakan nested loop
df_formulas <- data.frame(x = x_values)

for (formula in formulas) {
  values <- c()
  for (x in x_values) {
    values <- c(values, compute_formula(x, formula))  # nested loop
  }
  df_formulas[[formula]] <- values
}

# Tampilkan tabel hasil
df_formulas %>%
  mutate(across(where(is.numeric), ~ round(., 2))) %>%
  kable(caption = "📊 Hasil Komputasi Multi-Formula (x = 1 sampai 20)",
        align = "c") %>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed"),
                full_width = FALSE) %>%
  column_spec(1, bold = TRUE, color = "#2c3e50")

library(plotly)

# ── Reshape ke format long ──
df_long <- df_formulas %>%
  pivot_longer(cols = -x, names_to = "formula", values_to = "y") %>%
  mutate(formula = factor(formula,
                          levels = c("linear","quadratic","cubic","exponential"),
                          labels = c("Linear: f(x)=2x+1",
                                     "Quadratic: f(x)=x²+3x+2",
                                     "Cubic: f(x)=x³-2x²+x",
                                     "Exponential: f(x)=e^0.3x")))

# ── Plot ggplot ──
p <- ggplot(df_long, aes(x = x, y = y, color = formula, group = formula,
                         text = paste("x:", x, "<br>y:", round(y,2)))) +
  geom_line(linewidth = 1.2) +
  geom_point(size = 2.5, alpha = 0.8) +
  scale_color_manual(values = c("#2196F3","#4CAF50","#FF5722","#9C27B0")) +
  scale_x_continuous(breaks = 1:20) +
  labs(
    title    = "📈 Task 1: Perbandingan Multi-Formula",
    x        = "x",
    y        = "f(x)",
    color    = "Formula"
  ) +
  theme_minimal()

ggplotly(p, tooltip = "text")

# ============================================================
# TASK 2: Nested Simulation — Multi-Sales & Discounts
# ============================================================

# Nested helper function: tentukan discount berdasarkan sales amount
apply_discount <- function(sales_amount) {
  if (sales_amount >= 900) return(0.15)
  else if (sales_amount >= 700) return(0.10)
  else if (sales_amount >= 500) return(0.07)
  else if (sales_amount >= 300) return(0.05)
  else return(0.00)
}

# Fungsi utama simulasi penjualan
simulate_sales <- function(n_salesperson, days) {
  records <- list()
  idx     <- 1
  
  # Nested loop: per salesperson, per day
  for (sp_id in 1:n_salesperson) {
    for (day in 1:days) {
      sales_amount  <- round(runif(1, min = 100, max = 1000), 2)
      discount_rate <- apply_discount(sales_amount)  # nested function call
      net_sales     <- round(sales_amount * (1 - discount_rate), 2)
      
      records[[idx]] <- data.frame(
        sales_id      = sp_id,
        day           = day,
        sales_amount  = sales_amount,
        discount_rate = discount_rate,
        net_sales     = net_sales
      )
      idx <- idx + 1
    }
  }
  return(do.call(rbind, records))
}

# ── Simulasi: 5 salesperson × 30 hari ──
set.seed(42)
df_sales <- simulate_sales(n_salesperson = 5, days = 30)

# ── Hitung cumulative sales per salesperson ──
df_sales <- df_sales %>%
  group_by(sales_id) %>%
  mutate(cumulative_sales = cumsum(net_sales)) %>%
  ungroup()

# ── Summary statistics ──
summary_sales <- df_sales %>%
  group_by(sales_id) %>%
  summarise(
    Total_Sales   = round(sum(sales_amount), 2),
    Total_Net     = round(sum(net_sales), 2),
    Avg_Discount  = paste0(round(mean(discount_rate) * 100, 1), "%"),
    Avg_Daily_Sales = round(mean(sales_amount), 2),
    .groups = "drop"
  ) %>%
  rename("Salesperson ID" = sales_id)

summary_sales %>%
  kable(caption = "📊 Summary Statistik per Salesperson (30 Hari)",
        align = "c") %>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed"),
                full_width = FALSE) %>%
  row_spec(0, background = "#3498db", color = "white")

# ── Plot 1: Cumulative Sales per Salesperson ──
p1 <- ggplot(df_sales,
             aes(x = day, y = cumulative_sales,
                 color = factor(sales_id), group = factor(sales_id))) +
  geom_line(linewidth = 1.1) +
  geom_point(size = 1.5, alpha = 0.7) +
  scale_color_brewer(palette = "Set1") +
  labs(
    title  = "📈 Cumulative Net Sales per Salesperson",
    x      = "Day", y = "Cumulative Net Sales",
    color  = "Salesperson"
  ) +
  theme_minimal(base_size = 12) +
  theme(plot.title = element_text(face = "bold"),
        legend.position = "bottom")

# ── Plot 2: Total Net Sales per Salesperson ──
total_net <- df_sales %>%
  group_by(sales_id) %>%
  summarise(total_net = sum(net_sales), .groups = "drop")

p2 <- ggplot(total_net, aes(x = factor(sales_id), y = total_net,
                             fill = factor(sales_id))) +
  geom_col(width = 0.65, show.legend = FALSE) +
  geom_text(aes(label = paste0("$", format(round(total_net), big.mark=","))),
            vjust = -0.4, size = 3.8, fontface = "bold") +
  scale_fill_brewer(palette = "Set1") +
  scale_x_discrete(labels = paste0("SP ", 1:5)) +
  labs(
    title = "💰 Total Net Sales per Salesperson",
    x     = "Salesperson", y = "Total Net Sales"
  ) +
  theme_minimal(base_size = 12) +
  theme(plot.title = element_text(face = "bold"),
        panel.grid.major.x = element_blank())

grid.arrange(p1, p2, ncol = 2)

# ============================================================
# TASK 3: Multi-Level Performance Categorization
# ============================================================

categorize_performance <- function(sales_amount) {
  # Kategorisasi berdasarkan sales amount
  if      (sales_amount >= 850) return("Excellent")
  else if (sales_amount >= 700) return("Very Good")
  else if (sales_amount >= 500) return("Good")
  else if (sales_amount >= 300) return("Average")
  else                          return("Poor")
}

# ── Loop through vector untuk kategorisasi ──
set.seed(99)
sales_vector <- runif(200, min = 100, max = 1000)

categories <- c()
for (amount in sales_vector) {              # loop through vector
  categories <- c(categories, categorize_performance(amount))
}

df_perf <- data.frame(sales_amount = sales_vector, category = categories)

# ── Hitung persentase per kategori ──
cat_order   <- c("Excellent", "Very Good", "Good", "Average", "Poor")
cat_summary <- df_perf %>%
  group_by(category) %>%
  summarise(Count = n(), .groups = "drop") %>%
  mutate(
    category   = factor(category, levels = cat_order),
    Percentage = round(Count / sum(Count) * 100, 2),
    Label      = paste0(Count, " (", Percentage, "%)")
  ) %>%
  arrange(category)

cat_summary %>%
  select(Category = category, Count, Percentage) %>%
  mutate(Percentage = paste0(Percentage, "%")) %>%
  kable(caption = "📊 Distribusi Kategori Performa",
        align = "c") %>%
  kable_styling(bootstrap_options = c("striped", "hover"),
                full_width = FALSE) %>%
  row_spec(1, background = "#d5f5e3") %>%
  row_spec(2, background = "#a9dfbf") %>%
  row_spec(3, background = "#fdebd0") %>%
  row_spec(4, background = "#fad7a0") %>%
  row_spec(5, background = "#f5b7b1")

colors_perf <- c("Excellent" = "#1a9641",
                 "Very Good" = "#a6d96a",
                 "Good"      = "#ffffbf",
                 "Average"   = "#fdae61",
                 "Poor"      = "#d7191c")

# ── Bar Chart ──
p_bar <- ggplot(cat_summary,
                aes(x = category, y = Count, fill = category)) +
  geom_col(width = 0.65, show.legend = FALSE) +
  geom_text(aes(label = paste0(Count, "\n(", Percentage, "%)")),
            vjust = -0.3, size = 3.8, fontface = "bold") +
  scale_fill_manual(values = colors_perf) +
  scale_x_discrete(limits = cat_order) +
  labs(title = "📊 Bar Chart: Distribusi Kategori Performa",
       x = "Kategori", y = "Jumlah") +
  theme_minimal(base_size = 12) +
  theme(plot.title = element_text(face = "bold"),
        panel.grid.major.x = element_blank())

# ── Pie Chart ──
p_pie <- ggplot(cat_summary,
                aes(x = "", y = Count, fill = category)) +
  geom_col(width = 1, color = "white", linewidth = 0.8) +
  coord_polar("y", start = 0) +
  geom_text(aes(label = paste0(Percentage, "%")),
            position = position_stack(vjust = 0.5),
            size = 4, fontface = "bold") +
  scale_fill_manual(values = colors_perf,
                    limits = cat_order) +
  labs(title = "🥧 Pie Chart: Distribusi Kategori Performa",
       fill = "Kategori") +
  theme_void(base_size = 12) +
  theme(plot.title = element_text(face = "bold", hjust = 0.5),
        legend.position = "right")

grid.arrange(p_bar, p_pie, ncol = 2)

# ============================================================
# TASK 4: Multi-Company Dataset Simulation
# ============================================================

departments <- c("Engineering", "Marketing", "Finance", "HR", "Operations")

generate_company_data <- function(n_company, n_employees) {
  records <- list()
  emp_id  <- 1
  
  # Nested loop: per company, per employee
  for (c_id in 1:n_company) {
    for (e in 1:n_employees) {
      salary            <- max(2000, round(rnorm(1, mean = 5000, sd = 1500), 2))
      performance_score <- round(runif(1, 50, 100), 2)
      kpi_score         <- round(runif(1, 50, 100), 2)
      department        <- sample(departments, 1)
      
      # Conditional logic: top performers (KPI > 90)
      is_top <- kpi_score > 90
      
      records[[emp_id]] <- data.frame(
        company_id        = paste0("C", sprintf("%02d", c_id)),
        employee_id       = paste0("E", sprintf("%04d", emp_id)),
        salary            = salary,
        department        = department,
        performance_score = performance_score,
        KPI_score         = kpi_score,
        top_performer     = is_top
      )
      emp_id <- emp_id + 1
    }
  }
  return(do.call(rbind, records))
}

# ── Generate: 4 perusahaan × 50 karyawan ──
set.seed(7)
df_company <- generate_company_data(n_company = 4, n_employees = 50)

# ── Summary per company ──
company_summary <- df_company %>%
  group_by(company_id) %>%
  summarise(
    Avg_Salary      = round(mean(salary), 2),
    Avg_Performance = round(mean(performance_score), 2),
    Max_KPI         = round(max(KPI_score), 2),
    Top_Performers  = sum(top_performer),
    Total_Employees = n(),
    .groups = "drop"
  )

company_summary %>%
  kable(caption = "📊 Summary per Perusahaan",
        align = "c",
        col.names = c("Company", "Avg Salary", "Avg Performance",
                      "Max KPI", "Top Performers", "Total Employees")) %>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed"),
                full_width = FALSE) %>%
  row_spec(0, background = "#2c3e50", color = "white")

pal4 <- brewer.pal(4, "Set2")

# 1. Avg Salary per Company
p4a <- ggplot(company_summary,
              aes(x = company_id, y = Avg_Salary, fill = company_id)) +
  geom_col(show.legend = FALSE, width = 0.6) +
  geom_text(aes(label = paste0("$", format(Avg_Salary, big.mark=","))),
            vjust = -0.3, size = 4, fontface = "bold") +
  scale_fill_manual(values = pal4) +
  labs(title = "💵 Avg Salary per Company", x = "Company", y = "Avg Salary") +
  theme_minimal() + theme(plot.title = element_text(face="bold"),
                          panel.grid.major.x = element_blank())

# 2. Avg Performance Score
p4b <- ggplot(company_summary,
              aes(x = company_id, y = Avg_Performance, fill = company_id)) +
  geom_col(show.legend = FALSE, width = 0.6) +
  geom_text(aes(label = round(Avg_Performance, 1)),
            vjust = -0.3, size = 4, fontface = "bold") +
  scale_fill_manual(values = pal4) +
  labs(title = "🏆 Avg Performance Score", x = "Company", y = "Score") +
  theme_minimal() + theme(plot.title = element_text(face="bold"),
                          panel.grid.major.x = element_blank())

# 3. KPI Score — Boxplot
p4c <- ggplot(df_company,
              aes(x = company_id, y = KPI_score, fill = company_id)) +
  geom_boxplot(show.legend = FALSE, alpha = 0.85, outlier.color = "red") +
  scale_fill_manual(values = pal4) +
  labs(title = "📦 KPI Score Distribution (Boxplot)",
       x = "Company", y = "KPI Score") +
  theme_minimal() + theme(plot.title = element_text(face="bold"))

# 4. Top Performers Count
p4d <- ggplot(company_summary,
              aes(x = company_id, y = Top_Performers, fill = company_id)) +
  geom_col(show.legend = FALSE, width = 0.6) +
  geom_text(aes(label = Top_Performers), vjust = -0.3, size = 4.5, fontface = "bold") +
  scale_fill_manual(values = pal4) +
  labs(title = "⭐ Top Performers (KPI > 90)", x = "Company", y = "Count") +
  theme_minimal() + theme(plot.title = element_text(face="bold"),
                          panel.grid.major.x = element_blank())

grid.arrange(p4a, p4b, p4c, p4d, ncol = 2,
             top = "Task 4: Multi-Company Dataset Analysis")

# ============================================================
# TASK 5: Monte Carlo Simulation — Pi & Probability
# ============================================================

monte_carlo_pi <- function(n_points, seed = 42) {
  set.seed(seed)
  inside_circle  <- 0
  in_subsquare   <- 0
  
  x_in <- c(); y_in <- c()
  x_out <- c(); y_out <- c()
  
  # Loop untuk iterasi
  for (i in 1:n_points) {
    x <- runif(1, -1, 1)
    y <- runif(1, -1, 1)
    
    # Cek apakah di dalam unit circle
    if (x^2 + y^2 <= 1) {
      inside_circle <- inside_circle + 1
      x_in <- c(x_in, x); y_in <- c(y_in, y)
    } else {
      x_out <- c(x_out, x); y_out <- c(y_out, y)
    }
    
    # Cek sub-square [0, 0.5] × [0, 0.5]
    if (x >= 0 && x <= 0.5 && y >= 0 && y <= 0.5) {
      in_subsquare <- in_subsquare + 1
    }
  }
  
  pi_estimate    <- 4 * inside_circle / n_points
  subsquare_prob <- in_subsquare / n_points
  
  list(
    pi_estimate    = pi_estimate,
    inside_circle  = inside_circle,
    outside_circle = n_points - inside_circle,
    subsquare_prob = subsquare_prob,
    x_in = x_in, y_in = y_in,
    x_out = x_out, y_out = y_out
  )
}

# ── Uji dengan berbagai n_points ──
n_list    <- c(100, 500, 1000, 5000, 10000)
pi_results <- data.frame(
  n_points    = n_list,
  pi_estimate = sapply(n_list, function(n) monte_carlo_pi(n)$pi_estimate)
) %>%
  mutate(error = round(abs(pi_estimate - pi), 6),
         pi_estimate = round(pi_estimate, 6))

pi_results %>%
  kable(caption = paste("🎯 Konvergensi Estimasi π (nilai asli π =", round(pi, 6), ")"),
        align = "c",
        col.names = c("n Points", "π Estimate", "Absolute Error")) %>%
  kable_styling(bootstrap_options = c("striped", "hover"),
                full_width = FALSE) %>%
  row_spec(5, bold = TRUE, background = "#d5f5e3")

final <- monte_carlo_pi(10000)
cat(sprintf("\n🎯 Estimasi π (n=10.000) : %.5f\n", final$pi_estimate))

cat(sprintf("📦 Probabilitas Sub-square: %.4f  (expected: 0.0625)\n",
            final$subsquare_prob))

# ── Plot 1: Scatter Inside vs Outside Circle ──
df_circle <- rbind(
  data.frame(x = final$x_in,  y = final$y_in,  status = "Inside Circle"),
  data.frame(x = final$x_out, y = final$y_out, status = "Outside Circle")
)

# Buat circle & sub-square data
theta  <- seq(0, 2*pi, length.out = 300)
circle_df  <- data.frame(x = cos(theta), y = sin(theta))
square_df  <- data.frame(
  x = c(0, 0.5, 0.5, 0, 0),
  y = c(0, 0, 0.5, 0.5, 0)
)

p5a <- ggplot(df_circle, aes(x = x, y = y, color = status)) +
  geom_point(size = 0.3, alpha = 0.4) +
  geom_path(data = circle_df, aes(x, y),
            color = "navy", linewidth = 1.2, inherit.aes = FALSE) +
  geom_path(data = square_df, aes(x, y),
            color = "orange", linewidth = 1.2, linetype = "dashed",
            inherit.aes = FALSE) +
  scale_color_manual(values = c("Inside Circle" = "#27ae60",
                                "Outside Circle" = "#e74c3c")) +
  coord_fixed() +
  labs(
    title = sprintf("🎯 Monte Carlo π ≈ %.5f (n=10.000)", final$pi_estimate),
    color = NULL
  ) +
  theme_minimal() +
  theme(plot.title = element_text(face = "bold"),
        legend.position = "bottom")

# ── Plot 2: Konvergensi Estimasi π ──
p5b <- ggplot(pi_results, aes(x = n_points, y = pi_estimate)) +
  geom_line(color = "#2196F3", linewidth = 1.2) +
  geom_point(color = "#2196F3", size = 3) +
  geom_hline(yintercept = pi, color = "red",
             linetype = "dashed", linewidth = 1) +
  annotate("text", x = max(n_list)*0.6, y = pi + 0.015,
           label = paste0("True π = ", round(pi, 5)),
           color = "red", size = 3.8) +
  scale_x_continuous(labels = scales::comma) +
  labs(
    title = "📉 Konvergensi Estimasi π",
    x     = "Jumlah Titik",
    y     = "Estimasi π"
  ) +
  theme_minimal() +
  theme(plot.title = element_text(face = "bold"))

grid.arrange(p5a, p5b, ncol = 2)

# ============================================================
# TASK 6: Advanced Data Transformation & Feature Engineering
# ============================================================

normalize_columns <- function(df, cols) {
  # Min-Max normalization menggunakan loop
  for (col in cols) {
    col_min <- min(df[[col]], na.rm = TRUE)
    col_max <- max(df[[col]], na.rm = TRUE)
    df[[paste0(col, "_norm")]] <- (df[[col]] - col_min) / (col_max - col_min)
  }
  return(df)
}

z_score <- function(df, cols) {
  # Z-score standardization menggunakan loop
  for (col in cols) {
    mu    <- mean(df[[col]], na.rm = TRUE)
    sigma <- sd(df[[col]], na.rm = TRUE)
    df[[paste0(col, "_zscore")]] <- (df[[col]] - mu) / sigma
  }
  return(df)
}

# ── Gunakan df_company dari Task 4 ──
df_t6 <- df_company

# ── Feature Engineering ──
# 1. performance_category
df_t6$performance_category <- ifelse(
  df_t6$performance_score >= 80, "High",
  ifelse(df_t6$performance_score >= 65, "Mid", "Low")
)

# 2. salary_bracket
df_t6$salary_bracket <- cut(
  df_t6$salary,
  breaks = c(0, 3000, 5000, 7000, Inf),
  labels = c("Low", "Medium", "High", "Very High"),
  include.lowest = TRUE
)

# ── Terapkan transformasi ──
num_cols <- c("salary", "performance_score", "KPI_score")
df_t6    <- normalize_columns(df_t6, num_cols)
df_t6    <- z_score(df_t6, num_cols)

# Tampilkan preview
df_t6 %>%
  select(employee_id, salary, salary_norm, salary_zscore,
         performance_score, performance_score_norm,
         performance_category, salary_bracket) %>%
  head(8) %>%
  mutate(across(where(is.numeric), ~round(., 3))) %>%
  kable(caption = "📊 Preview Data Setelah Transformasi & Feature Engineering",
        align = "c") %>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed"),
                full_width = TRUE, font_size = 12) %>%
  row_spec(0, background = "#8e44ad", color = "white")

# ── Histogram: Before & After untuk 3 kolom ──
plot_list <- list()
col_labels <- c("Salary", "Performance Score", "KPI Score")

for (i in seq_along(num_cols)) {
  col <- num_cols[i]
  lbl <- col_labels[i]
  
  # Original
  p_orig <- ggplot(df_t6, aes_string(x = col)) +
    geom_histogram(bins = 15, fill = "#3498db", color = "white", alpha = 0.85) +
    labs(title = paste(lbl, "— Original"), x = col, y = "Freq") +
    theme_minimal(base_size = 11) +
    theme(plot.title = element_text(face = "bold", size = 10))
  
  # Normalized
  p_norm <- ggplot(df_t6, aes_string(x = paste0(col, "_norm"))) +
    geom_histogram(bins = 15, fill = "#2ecc71", color = "white", alpha = 0.85) +
    labs(title = paste(lbl, "— Min-Max Norm"), x = "Normalized", y = "Freq") +
    theme_minimal(base_size = 11) +
    theme(plot.title = element_text(face = "bold", size = 10))
  
  # Z-Score
  p_z <- ggplot(df_t6, aes_string(x = paste0(col, "_zscore"))) +
    geom_histogram(bins = 15, fill = "#e74c3c", color = "white", alpha = 0.85) +
    labs(title = paste(lbl, "— Z-Score"), x = "Z-Score", y = "Freq") +
    theme_minimal(base_size = 11) +
    theme(plot.title = element_text(face = "bold", size = 10))
  
  plot_list <- c(plot_list, list(p_orig, p_norm, p_z))
}

do.call(grid.arrange, c(plot_list, ncol = 3,
        top = "Task 6: Distribusi Sebelum & Sesudah Transformasi"))

ggplot(df_t6, aes(x = salary_bracket, y = KPI_score, fill = salary_bracket)) +
  geom_boxplot(alpha = 0.8, outlier.color = "red", show.legend = FALSE) +
  scale_fill_brewer(palette = "Set3") +
  scale_x_discrete(limits = c("Low", "Medium", "High", "Very High")) +
  labs(
    title = "📦 KPI Score berdasarkan Salary Bracket",
    x     = "Salary Bracket",
    y     = "KPI Score"
  ) +
  theme_minimal(base_size = 13) +
  theme(plot.title = element_text(face = "bold"))

# ============================================================
# TASK 7: Mini Project — Company KPI Dashboard
# ============================================================

set.seed(2024)
df_dash <- generate_company_data(n_company = 7, n_employees = 80)

# ── KPI Tier Categorization (loop-based) ──
kpi_tier <- function(score) {
  if      (score >= 90) return("Platinum")
  else if (score >= 75) return("Gold")
  else if (score >= 60) return("Silver")
  else                  return("Bronze")
}

kpi_tiers <- c()
for (score in df_dash$KPI_score) {   # loop per employee
  kpi_tiers <- c(kpi_tiers, kpi_tier(score))
}
df_dash$KPI_tier <- factor(kpi_tiers,
                            levels = c("Platinum","Gold","Silver","Bronze"))

# ── Summary per Company ──
dash_summary <- df_dash %>%
  group_by(company_id) %>%
  summarise(
    Avg_Salary    = round(mean(salary), 0),
    Avg_KPI       = round(mean(KPI_score), 2),
    Top_Performers = sum(top_performer),
    Total_Emp     = n(),
    .groups = "drop"
  )

dash_summary %>%
  kable(caption = "📋 Company KPI Summary (7 Perusahaan × 80 Karyawan)",
        align   = "c",
        col.names = c("Company","Avg Salary","Avg KPI","Top Performers","Total Emp")) %>%
  kable_styling(bootstrap_options = c("striped","hover","condensed"),
                full_width = FALSE) %>%
  row_spec(0, background = "#2c3e50", color = "white") %>%
  column_spec(3, bold = TRUE, color = "#2980b9")

tier_colors <- c("Platinum" = "#8e44ad",
                 "Gold"     = "#f39c12",
                 "Silver"   = "#95a5a6",
                 "Bronze"   = "#a04000")

# ── Grouped Bar: KPI Tier per Company ──
tier_counts <- df_dash %>%
  group_by(company_id, KPI_tier) %>%
  summarise(count = n(), .groups = "drop")

p7a <- ggplot(tier_counts,
              aes(x = company_id, y = count, fill = KPI_tier)) +
  geom_col(position = "dodge", width = 0.75) +
  scale_fill_manual(values = tier_colors) +
  labs(title = "🏅 KPI Tier Distribution per Company",
       x = "Company", y = "Count", fill = "KPI Tier") +
  theme_minimal(base_size = 12) +
  theme(plot.title = element_text(face="bold"),
        legend.position = "bottom")

# ── Top Performers per Company ──
p7b <- ggplot(dash_summary,
              aes(x = company_id, y = Top_Performers, fill = company_id)) +
  geom_col(show.legend = FALSE, width = 0.6) +
  geom_text(aes(label = Top_Performers), vjust = -0.3,
            size = 4.5, fontface = "bold") +
  scale_fill_brewer(palette = "Set1") +
  labs(title = "⭐ Top Performers per Company",
       x = "Company", y = "Count") +
  theme_minimal(base_size = 12) +
  theme(plot.title = element_text(face="bold"),
        panel.grid.major.x = element_blank())

grid.arrange(p7a, p7b, ncol = 2)

ggplot(df_dash, aes(x = salary, y = KPI_score,
                    color = company_id)) +
  geom_point(size = 1.8, alpha = 0.55) +
  geom_smooth(method = "lm", se = TRUE, linewidth = 1,
              aes(fill = company_id), alpha = 0.1) +
  scale_color_brewer(palette = "Set1") +
  scale_fill_brewer(palette = "Set1") +
  labs(
    title    = "📊 Salary vs KPI Score dengan Regression Lines per Company",
    subtitle = "Setiap warna mewakili 1 perusahaan; garis = regresi linear",
    x        = "Salary",
    y        = "KPI Score",
    color    = "Company",
    fill     = "Company"
  ) +
  theme_minimal(base_size = 12) +
  theme(plot.title    = element_text(face = "bold"),
        legend.position = "bottom") +
  guides(color = guide_legend(nrow = 1))

# ── Boxplot Salary per Company ──
p_sal <- ggplot(df_dash,
                aes(x = company_id, y = salary, fill = company_id)) +
  geom_boxplot(alpha = 0.85, show.legend = FALSE,
               outlier.color = "red", outlier.size = 1.5) +
  scale_fill_brewer(palette = "Set1") +
  labs(title = "💰 Distribusi Salary per Company",
       x = "Company", y = "Salary") +
  theme_minimal() +
  theme(plot.title = element_text(face = "bold"))

# ── Department: Avg KPI per Company (Grouped Bar) ──
dept_kpi <- df_dash %>%
  group_by(department, company_id) %>%
  summarise(avg_kpi = round(mean(KPI_score), 2), .groups = "drop")

p_dept <- ggplot(dept_kpi,
                 aes(x = department, y = avg_kpi, fill = company_id)) +
  geom_col(position = "dodge", width = 0.75) +
  scale_fill_brewer(palette = "Set1") +
  labs(title = "🏢 Avg KPI per Department per Company",
       x = "Department", y = "Avg KPI Score",
       fill = "Company") +
  theme_minimal() +
  theme(plot.title = element_text(face = "bold"),
        axis.text.x = element_text(angle = 15, hjust = 1),
        legend.position = "bottom")

grid.arrange(p_sal, p_dept, nrow = 2,
             top = "Task 7: Company KPI Dashboard — Salary & Department Analysis")

# ============================================================
# TASK 8 (BONUS): Automated Report Generation
# ============================================================

# Fungsi: buat summary per company secara otomatis
generate_company_report <- function(df, c_id) {
  df_sub <- df %>% filter(company_id == c_id)
  
  # Statistik utama
  stats <- list(
    company      = c_id,
    n_emp        = nrow(df_sub),
    avg_salary   = round(mean(df_sub$salary), 0),
    avg_kpi      = round(mean(df_sub$KPI_score), 2),
    avg_perf     = round(mean(df_sub$performance_score), 2),
    n_top        = sum(df_sub$top_performer),
    max_kpi      = round(max(df_sub$KPI_score), 2),
    min_salary   = round(min(df_sub$salary), 0),
    max_salary   = round(max(df_sub$salary), 0)
  )
  
  # Top 5 performers
  top5 <- df_sub %>%
    arrange(desc(KPI_score)) %>%
    select(employee_id, department, salary, performance_score, KPI_score) %>%
    head(5) %>%
    mutate(across(where(is.numeric), ~round(., 2)))
  
  # KPI Tier distribution
  tier_dist <- df_sub %>%
    group_by(KPI_tier) %>%
    summarise(count = n(), pct = round(n()/nrow(df_sub)*100, 1), .groups="drop") %>%
    arrange(KPI_tier)
  
  # Department summary
  dept_sum <- df_sub %>%
    group_by(department) %>%
    summarise(
      Avg_KPI    = round(mean(KPI_score), 2),
      Avg_Salary = round(mean(salary), 0),
      Count      = n(),
      .groups = "drop"
    ) %>%
    arrange(desc(Avg_KPI))
  
  return(list(stats = stats, top5 = top5,
              tier_dist = tier_dist, dept_sum = dept_sum,
              data = df_sub))
}

cat("✅ Fungsi generate_company_report() siap digunakan.\n")

cat("✅ Loop akan dijalankan di bawah untuk menghasilkan report semua perusahaan.\n")

company_ids <- sort(unique(df_dash$company_id))
pal_tier    <- c("Platinum"="#8e44ad","Gold"="#f39c12","Silver"="#95a5a6","Bronze"="#a04000")

# ── LOOP: Generate report untuk setiap perusahaan ──
for (c_id in company_ids) {
  
  rpt <- generate_company_report(df_dash, c_id)
  s   <- rpt$stats
  
  # ── Section Header ──
  cat(sprintf("\n\n### 📌 %s — Automated Report\n\n", c_id))
  
  # ── KPI Stat Cards (dalam tabel kecil) ──
  stats_df <- data.frame(
    Metric = c("Total Karyawan", "Avg Salary", "Avg KPI Score",
               "Avg Performance", "Top Performers", "Max KPI",
               "Min Salary", "Max Salary"),
    Value  = c(s$n_emp,
               paste0("$", format(s$avg_salary, big.mark=",")),
               s$avg_kpi, s$avg_perf, s$n_top, s$max_kpi,
               paste0("$", format(s$min_salary, big.mark=",")),
               paste0("$", format(s$max_salary, big.mark=",")))
  )
  
  print(
    stats_df %>%
      kable(align = "c",
            caption = paste("📊 Key Metrics —", c_id)) %>%
      kable_styling(bootstrap_options = c("striped","hover"),
                    full_width = FALSE, font_size = 12) %>%
      row_spec(0, background = "#2c3e50", color = "white") %>%
      column_spec(2, bold = TRUE, color = "#2980b9")
  )
  
  # ── Top 5 Performers ──
  cat("\n**🥇 Top 5 Performers:**\n\n")
  print(
    rpt$top5 %>%
      kable(align = "c",
            col.names = c("Employee ID","Department","Salary","Performance","KPI Score")) %>%
      kable_styling(bootstrap_options = c("striped","hover","condensed"),
                    full_width = FALSE, font_size = 12) %>%
      row_spec(1, bold = TRUE, background = "#d5f5e3")
  )
  
  # ── KPI Tier Distribution ──
  cat("\n**🏷️ KPI Tier Breakdown:**\n\n")
  print(
    rpt$tier_dist %>%
      mutate(label = paste0(count, " karyawan (", pct, "%)")) %>%
      select(KPI_tier, label) %>%
      kable(align = "c", col.names = c("KPI Tier","Count & Percentage")) %>%
      kable_styling(bootstrap_options = c("striped"),
                    full_width = FALSE, font_size = 12)
  )
  
  # ── Department Summary ──
  cat("\n**🏢 Department Summary:**\n\n")
  print(
    rpt$dept_sum %>%
      kable(align = "c",
            col.names = c("Department","Avg KPI","Avg Salary","Employee Count")) %>%
      kable_styling(bootstrap_options = c("striped","hover"),
                    full_width = FALSE, font_size = 12)
  )
  
  # ── Plot untuk company ini ──
  df_sub <- rpt$data
  
  p_hist <- ggplot(df_sub, aes(x = KPI_score)) +
    geom_histogram(bins = 15, fill = "#3498db", color = "white", alpha = 0.85) +
    geom_vline(xintercept = mean(df_sub$KPI_score),
               color = "red", linetype = "dashed", linewidth = 1) +
    annotate("text", x = mean(df_sub$KPI_score) + 1.5, y = Inf,
             label = paste0("Mean: ", round(mean(df_sub$KPI_score),1)),
             vjust = 1.5, color = "red", size = 3.5) +
    labs(title = paste(c_id, "— KPI Score Distribution"),
         x = "KPI Score", y = "Count") +
    theme_minimal(base_size = 11) +
    theme(plot.title = element_text(face = "bold"))
  
  p_dept2 <- ggplot(rpt$dept_sum,
                    aes(x = reorder(department, Avg_KPI), y = Avg_KPI,
                        fill = Avg_KPI)) +
    geom_col(show.legend = FALSE) +
    geom_text(aes(label = Avg_KPI), hjust = -0.2, size = 3.5, fontface = "bold") +
    scale_fill_gradient(low = "#f39c12", high = "#27ae60") +
    coord_flip() +
    labs(title = paste(c_id, "— Avg KPI per Department"),
         x = "Department", y = "Avg KPI") +
    theme_minimal(base_size = 11) +
    theme(plot.title = element_text(face = "bold"))
  
  grid.arrange(p_hist, p_dept2, ncol = 2)
  
  cat("\n---\n")
}