import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from nba_api.stats.endpoints import (
    leaguedashptstats,
    teamdashboardbygeneralsplits,
    playerdashboardbygeneralsplits,
    leaguedashteamstats
)
from nba_api.stats.static import teams
import time

# Set visualization style
sns.set_style("whitegrid")
plt.rcParams['figure.figsize'] = (14, 8)

def get_transition_data(season='2023-24'):
    """
    Retrieve transition play data for all NBA teams
    """
    print(f"Fetching transition data for {season} season...")

    # Get transition frequency and efficiency using PT Stats
    transition_stats = leaguedashptstats.LeagueDashPtStats(
        season=season,
        per_mode_simple='PerGame',
        pt_measure_type='Transition'
    )

    time.sleep(1)  # Rate limiting

    df_transition = transition_stats.get_data_frames()[0]

    # Calculate additional metrics
    df_transition['TRANSITION_PPP'] = (
        df_transition['PTS'] / df_transition['POSS']
    )
    df_transition['TRANSITION_EFG'] = (
        (df_transition['FGM'] + 0.5 * df_transition['FG3M']) /
        df_transition['FGA']
    )
    df_transition['TRANSITION_FREQ'] = (
        df_transition['POSS'] / df_transition['GP']
    )

    return df_transition

def get_player_transition_stats(season='2023-24', min_possessions=50):
    """
    Get player-level transition statistics
    """
    print(f"Fetching player transition data for {season}...")

    player_transition = leaguedashptstats.LeagueDashPtStats(
        season=season,
        per_mode_simple='Totals',
        pt_measure_type='Transition',
        player_or_team='Player'
    )

    time.sleep(1)

    df_players = player_transition.get_data_frames()[0]

    # Filter by minimum possessions
    df_players = df_players[df_players['POSS'] >= min_possessions].copy()

    # Calculate efficiency metrics
    df_players['TRANSITION_PPP'] = df_players['PTS'] / df_players['POSS']
    df_players['TRANSITION_EFG'] = (
        (df_players['FGM'] + 0.5 * df_players['FG3M']) /
        df_players['FGA']
    )
    df_players['AST_RATE'] = df_players['AST'] / df_players['POSS'] * 100
    df_players['TOV_RATE'] = df_players['TOV'] / df_players['POSS'] * 100

    return df_players

def analyze_transition_efficiency(df_transition):
    """
    Comprehensive analysis of transition efficiency across teams
    """
    print("\n=== TRANSITION EFFICIENCY ANALYSIS ===\n")

    # Sort by transition PPP
    df_sorted = df_transition.sort_values('TRANSITION_PPP', ascending=False)

    print("Top 10 Teams by Transition Efficiency (PPP):")
    print(df_sorted[['TEAM_NAME', 'POSS', 'PTS', 'TRANSITION_PPP',
                     'FG_PCT', 'TRANSITION_EFG']].head(10).to_string(index=False))

    print("\n\nTop 10 Teams by Transition Frequency:")
    df_freq = df_transition.sort_values('TRANSITION_FREQ', ascending=False)
    print(df_freq[['TEAM_NAME', 'GP', 'POSS', 'TRANSITION_FREQ',
                   'PTS']].head(10).to_string(index=False))

    # Statistical summary
    print("\n\nLeague-Wide Transition Statistics:")
    stats_summary = df_transition[[
        'POSS', 'PTS', 'TRANSITION_PPP', 'FG_PCT', 'FG3_PCT'
    ]].describe()
    print(stats_summary)

    return df_sorted

def create_transition_visualizations(df_transition):
    """
    Generate comprehensive visualizations of transition play
    """
    fig, axes = plt.subplots(2, 2, figsize=(16, 12))

    # 1. Transition Efficiency vs Volume
    ax1 = axes[0, 0]
    scatter = ax1.scatter(
        df_transition['TRANSITION_FREQ'],
        df_transition['TRANSITION_PPP'],
        s=df_transition['PTS'] * 2,
        alpha=0.6,
        c=df_transition['FG_PCT'],
        cmap='RdYlGn'
    )
    ax1.set_xlabel('Transition Frequency (Possessions per Game)', fontsize=11)
    ax1.set_ylabel('Transition Efficiency (PPP)', fontsize=11)
    ax1.set_title('Transition Volume vs Efficiency', fontsize=13, fontweight='bold')
    ax1.axhline(df_transition['TRANSITION_PPP'].mean(),
                color='red', linestyle='--', alpha=0.5, label='League Avg PPP')
    ax1.axvline(df_transition['TRANSITION_FREQ'].mean(),
                color='blue', linestyle='--', alpha=0.5, label='League Avg Freq')
    ax1.legend()
    ax1.grid(True, alpha=0.3)
    plt.colorbar(scatter, ax=ax1, label='FG%')

    # 2. Top Teams by Transition PPP
    ax2 = axes[0, 1]
    top_teams = df_transition.nlargest(15, 'TRANSITION_PPP')
    ax2.barh(range(len(top_teams)), top_teams['TRANSITION_PPP'],
             color=plt.cm.viridis(np.linspace(0, 1, len(top_teams))))
    ax2.set_yticks(range(len(top_teams)))
    ax2.set_yticklabels(top_teams['TEAM_NAME'], fontsize=9)
    ax2.set_xlabel('Points Per Possession', fontsize=11)
    ax2.set_title('Top 15 Teams: Transition Efficiency', fontsize=13, fontweight='bold')
    ax2.axvline(df_transition['TRANSITION_PPP'].mean(),
                color='red', linestyle='--', alpha=0.7, label='League Avg')
    ax2.legend()
    ax2.invert_yaxis()
    ax2.grid(True, axis='x', alpha=0.3)

    # 3. Distribution of Transition Efficiency
    ax3 = axes[1, 0]
    ax3.hist(df_transition['TRANSITION_PPP'], bins=20,
             color='steelblue', alpha=0.7, edgecolor='black')
    ax3.axvline(df_transition['TRANSITION_PPP'].mean(),
                color='red', linestyle='--', linewidth=2, label='Mean')
    ax3.axvline(df_transition['TRANSITION_PPP'].median(),
                color='orange', linestyle='--', linewidth=2, label='Median')
    ax3.set_xlabel('Transition PPP', fontsize=11)
    ax3.set_ylabel('Number of Teams', fontsize=11)
    ax3.set_title('Distribution of Transition Efficiency', fontsize=13, fontweight='bold')
    ax3.legend()
    ax3.grid(True, alpha=0.3)

    # 4. Shot Selection in Transition
    ax4 = axes[1, 1]
    metrics = ['FG2_PCT', 'FG3_PCT', 'FG_PCT']
    averages = [df_transition[m].mean() for m in metrics]
    colors = ['#FF6B6B', '#4ECDC4', '#45B7D1']
    bars = ax4.bar(range(len(metrics)), averages, color=colors, alpha=0.7, edgecolor='black')
    ax4.set_xticks(range(len(metrics)))
    ax4.set_xticklabels(['2PT%', '3PT%', 'FG%'], fontsize=11)
    ax4.set_ylabel('Percentage', fontsize=11)
    ax4.set_title('Average Shooting Efficiency in Transition', fontsize=13, fontweight='bold')
    ax4.set_ylim(0, 100)
    ax4.grid(True, axis='y', alpha=0.3)

    # Add value labels on bars
    for bar, val in zip(bars, averages):
        height = bar.get_height()
        ax4.text(bar.get_x() + bar.get_width()/2., height,
                f'{val:.1f}%', ha='center', va='bottom', fontsize=10, fontweight='bold')

    plt.tight_layout()
    plt.savefig('transition_offense_analysis.png', dpi=300, bbox_inches='tight')
    print("\nVisualization saved as 'transition_offense_analysis.png'")
    plt.show()

def analyze_player_transition_leaders(df_players, top_n=20):
    """
    Identify and analyze top transition players
    """
    print("\n=== TOP TRANSITION PLAYERS ===\n")

    # Top scorers in transition
    print(f"Top {top_n} Transition Scorers:")
    top_scorers = df_players.nlargest(top_n, 'PTS')
    print(top_scorers[['PLAYER_NAME', 'TEAM_ABBREVIATION', 'POSS', 'PTS',
                       'TRANSITION_PPP', 'FG_PCT']].to_string(index=False))

    # Most efficient players (min 100 possessions)
    print(f"\n\nMost Efficient Transition Players (min 100 possessions):")
    efficient = df_players[df_players['POSS'] >= 100].nlargest(top_n, 'TRANSITION_PPP')
    print(efficient[['PLAYER_NAME', 'TEAM_ABBREVIATION', 'POSS',
                     'TRANSITION_PPP', 'TRANSITION_EFG', 'FG_PCT']].to_string(index=False))

    # Best playmakers in transition
    print(f"\n\nTop Transition Playmakers (by Assist Rate):")
    playmakers = df_players[df_players['POSS'] >= 100].nlargest(top_n, 'AST_RATE')
    print(playmakers[['PLAYER_NAME', 'TEAM_ABBREVIATION', 'AST',
                      'AST_RATE', 'TRANSITION_PPP']].to_string(index=False))

def create_player_transition_viz(df_players):
    """
    Visualize player transition performance
    """
    # Filter for high-volume players
    df_viz = df_players[df_players['POSS'] >= 100].copy()

    fig, axes = plt.subplots(1, 2, figsize=(16, 6))

    # 1. Player Efficiency vs Volume
    ax1 = axes[0]
    scatter = ax1.scatter(
        df_viz['POSS'],
        df_viz['TRANSITION_PPP'],
        s=df_viz['PTS'],
        alpha=0.6,
        c=df_viz['TRANSITION_EFG'],
        cmap='plasma'
    )
    ax1.set_xlabel('Transition Possessions', fontsize=11)
    ax1.set_ylabel('Transition PPP', fontsize=11)
    ax1.set_title('Player Transition Efficiency vs Volume', fontsize=13, fontweight='bold')
    ax1.axhline(df_viz['TRANSITION_PPP'].mean(),
                color='red', linestyle='--', alpha=0.5, label='League Avg')
    ax1.legend()
    ax1.grid(True, alpha=0.3)
    plt.colorbar(scatter, ax=ax1, label='eFG%')

    # 2. Top 15 Players by Transition PPP
    ax2 = axes[1]
    top_15 = df_viz.nlargest(15, 'TRANSITION_PPP')
    bars = ax2.barh(range(len(top_15)), top_15['TRANSITION_PPP'],
                    color=plt.cm.plasma(np.linspace(0.3, 0.9, len(top_15))))
    ax2.set_yticks(range(len(top_15)))
    ax2.set_yticklabels(top_15['PLAYER_NAME'], fontsize=9)
    ax2.set_xlabel('Points Per Possession', fontsize=11)
    ax2.set_title('Top 15 Players: Transition Efficiency (min 100 poss)',
                  fontsize=13, fontweight='bold')
    ax2.invert_yaxis()
    ax2.grid(True, axis='x', alpha=0.3)

    plt.tight_layout()
    plt.savefig('player_transition_analysis.png', dpi=300, bbox_inches='tight')
    print("\nPlayer visualization saved as 'player_transition_analysis.png'")
    plt.show()

def compare_transition_defense(season='2023-24'):
    """
    Analyze transition defense effectiveness
    """
    print("\n=== TRANSITION DEFENSE ANALYSIS ===\n")

    # Get opponent transition stats (defense)
    print("Fetching transition defense data...")

    # Note: This would require opponent stats which may need different endpoint
    # For now, we'll use the regular stats as a template

    print("Best Transition Defense (Points Allowed):")
    print("Analysis requires opponent-specific transition data")

# Main execution
if __name__ == "__main__":
    # Set season
    SEASON = '2023-24'

    print("=" * 60)
    print("NBA TRANSITION OFFENSE ANALYSIS")
    print("=" * 60)

    # Get team transition data
    df_transition = get_transition_data(SEASON)

    # Analyze efficiency
    df_analyzed = analyze_transition_efficiency(df_transition)

    # Create visualizations
    create_transition_visualizations(df_transition)

    # Get player data
    df_players = get_player_transition_stats(SEASON, min_possessions=50)

    # Analyze top players
    analyze_player_transition_leaders(df_players, top_n=15)

    # Create player visualizations
    create_player_transition_viz(df_players)

    print("\n" + "=" * 60)
    print("Analysis complete!")
    print("=" * 60)

# Load required libraries
library(hoopR)
library(tidyverse)
library(ggplot2)
library(scales)
library(gridExtra)
library(corrplot)
library(broom)

# Set theme for visualizations
theme_set(theme_minimal(base_size = 12))

#' Get NBA play-by-play data with transition identification
#'
#' @param seasons Vector of seasons to analyze
#' @return Data frame with play-by-play data including transition flags
get_transition_pbp <- function(seasons = 2024) {
  cat("Fetching play-by-play data for transition analysis...\n")

  # Load play-by-play data
  pbp_data <- hoopR::load_nba_pbp(seasons = seasons)

  # Identify transition plays
  pbp_transition <- pbp_data %>%
    filter(!is.na(shooting_play), shooting_play == TRUE) %>%
    arrange(game_id, game_play_number) %>%
    group_by(game_id) %>%
    mutate(
      time_since_last = lag(clock_seconds) - clock_seconds,
      possession_changed = lag(team_id) != team_id,
      # Transition if shot within 8 seconds of possession change
      is_transition = if_else(
        possession_changed & abs(time_since_last) <= 8,
        TRUE,
        FALSE
      ),
      is_transition = replace_na(is_transition, FALSE),
      # Early offense if shot within 8-15 seconds
      is_early_offense = if_else(
        possession_changed & abs(time_since_last) > 8 & abs(time_since_last) <= 15,
        TRUE,
        FALSE
      )
    ) %>%
    ungroup()

  return(pbp_transition)
}

#' Calculate team transition efficiency metrics
#'
#' @param pbp_data Play-by-play data with transition flags
#' @return Data frame of team-level transition statistics
calculate_team_transition_stats <- function(pbp_data) {
  cat("Calculating team transition statistics...\n")

  team_stats <- pbp_data %>%
    group_by(team_id, pos_team_name) %>%
    summarise(
      total_possessions = n(),
      transition_possessions = sum(is_transition, na.rm = TRUE),
      early_offense_possessions = sum(is_early_offense, na.rm = TRUE),
      transition_freq = transition_possessions / total_possessions * 100,

      # Transition scoring
      transition_points = sum(if_else(is_transition, scoring_play, 0), na.rm = TRUE),
      transition_fgm = sum(is_transition & score_value > 0, na.rm = TRUE),
      transition_fga = sum(is_transition, na.rm = TRUE),
      transition_fg_pct = transition_fgm / transition_fga * 100,

      # Transition efficiency
      transition_ppp = transition_points / transition_possessions,

      # Shot types in transition
      transition_rim_att = sum(is_transition & shot_distance <= 4, na.rm = TRUE),
      transition_rim_pct = transition_rim_att / transition_fga * 100,
      transition_three_att = sum(is_transition & score_value == 3, na.rm = TRUE),
      transition_three_pct = transition_three_att / transition_fga * 100,

      # Half-court comparison
      halfcourt_possessions = total_possessions - transition_possessions,
      halfcourt_points = sum(if_else(!is_transition, scoring_play, 0), na.rm = TRUE),
      halfcourt_ppp = halfcourt_points / halfcourt_possessions,

      # Efficiency advantage
      efficiency_advantage = transition_ppp - halfcourt_ppp,

      .groups = 'drop'
    ) %>%
    arrange(desc(transition_ppp))

  return(team_stats)
}

#' Analyze factors driving transition success
#'
#' @param pbp_data Play-by-play data
#' @return Regression model and analysis results
analyze_transition_factors <- function(pbp_data) {
  cat("\n=== TRANSITION SUCCESS FACTORS ===\n\n")

  # Prepare data for regression
  transition_plays <- pbp_data %>%
    filter(is_transition == TRUE) %>%
    mutate(
      points_scored = if_else(score_value > 0, score_value, 0),
      scored = if_else(points_scored > 0, 1, 0),
      is_assisted = !is.na(assist_player_id),
      is_three = score_value == 3,
      is_rim_shot = shot_distance <= 4,
      shot_speed_fast = time_since_last <= 4,
      period_late = period >= 3
    )

  # Logistic regression for scoring probability
  model_score <- glm(
    scored ~ shot_distance + is_assisted + shot_speed_fast + period_late,
    data = transition_plays,
    family = binomial(link = "logit")
  )

  cat("Logistic Regression: Factors Affecting Transition Scoring\n")
  print(summary(model_score))

  # Odds ratios
  odds_ratios <- exp(coef(model_score))
  cat("\n\nOdds Ratios (likelihood of scoring):\n")
  print(odds_ratios)

  # Linear regression for points scored
  model_points <- lm(
    points_scored ~ shot_distance + is_assisted + is_three + shot_speed_fast,
    data = transition_plays
  )

  cat("\n\nLinear Regression: Factors Affecting Points Scored\n")
  print(summary(model_points))

  return(list(
    score_model = model_score,
    points_model = model_points,
    data = transition_plays
  ))
}

#' Create comprehensive transition visualizations
#'
#' @param team_stats Team transition statistics
#' @param pbp_data Play-by-play data
visualize_transition_analysis <- function(team_stats, pbp_data) {
  cat("\nCreating visualizations...\n")

  # 1. Transition Efficiency vs Frequency
  p1 <- ggplot(team_stats, aes(x = transition_freq, y = transition_ppp)) +
    geom_point(aes(size = transition_possessions, color = efficiency_advantage),
               alpha = 0.7) +
    geom_smooth(method = "lm", se = TRUE, color = "red", linetype = "dashed") +
    geom_hline(yintercept = mean(team_stats$transition_ppp, na.rm = TRUE),
               linetype = "dashed", color = "blue", alpha = 0.5) +
    geom_vline(xintercept = mean(team_stats$transition_freq, na.rm = TRUE),
               linetype = "dashed", color = "blue", alpha = 0.5) +
    scale_color_gradient2(low = "red", mid = "yellow", high = "green",
                         midpoint = 0, name = "Efficiency\nAdvantage") +
    scale_size_continuous(name = "Transition\nPossessions") +
    labs(
      title = "Transition Efficiency vs Frequency",
      subtitle = "Bubble size represents transition volume",
      x = "Transition Frequency (%)",
      y = "Transition Points Per Possession"
    ) +
    theme_minimal() +
    theme(legend.position = "right")

  # 2. Top teams by efficiency advantage
  p2 <- team_stats %>%
    top_n(15, efficiency_advantage) %>%
    ggplot(aes(x = reorder(pos_team_name, efficiency_advantage),
               y = efficiency_advantage)) +
    geom_col(aes(fill = transition_freq), alpha = 0.8) +
    geom_hline(yintercept = 0, linetype = "solid", color = "black") +
    scale_fill_gradient(low = "lightblue", high = "darkblue",
                       name = "Transition\nFreq (%)") +
    coord_flip() +
    labs(
      title = "Top 15 Teams: Transition Efficiency Advantage",
      subtitle = "Transition PPP minus Half-Court PPP",
      x = NULL,
      y = "Efficiency Advantage (PPP)"
    ) +
    theme_minimal() +
    theme(axis.text.y = element_text(size = 9))

  # 3. Shot selection in transition vs half-court
  shot_comparison <- pbp_data %>%
    mutate(context = if_else(is_transition, "Transition", "Half-Court")) %>%
    group_by(context) %>%
    summarise(
      rim_freq = sum(shot_distance <= 4, na.rm = TRUE) / n() * 100,
      mid_freq = sum(shot_distance > 4 & shot_distance < 22, na.rm = TRUE) / n() * 100,
      three_freq = sum(shot_distance >= 22, na.rm = TRUE) / n() * 100,
      .groups = 'drop'
    ) %>%
    pivot_longer(cols = ends_with("_freq"),
                 names_to = "shot_type",
                 values_to = "frequency") %>%
    mutate(shot_type = recode(shot_type,
                              "rim_freq" = "Rim",
                              "mid_freq" = "Mid-Range",
                              "three_freq" = "Three-Point"))

  p3 <- ggplot(shot_comparison, aes(x = context, y = frequency, fill = shot_type)) +
    geom_col(position = "dodge", alpha = 0.8) +
    scale_fill_manual(values = c("Rim" = "#2E7D32",
                                 "Mid-Range" = "#F57C00",
                                 "Three-Point" = "#1976D2"),
                     name = "Shot Type") +
    labs(
      title = "Shot Selection: Transition vs Half-Court",
      x = NULL,
      y = "Frequency (%)"
    ) +
    theme_minimal() +
    theme(legend.position = "bottom")

  # 4. Transition efficiency distribution
  p4 <- ggplot(team_stats, aes(x = transition_ppp)) +
    geom_histogram(aes(y = ..density..), bins = 15,
                   fill = "steelblue", alpha = 0.7, color = "black") +
    geom_density(color = "red", size = 1.2) +
    geom_vline(aes(xintercept = mean(transition_ppp, na.rm = TRUE)),
               color = "darkred", linetype = "dashed", size = 1) +
    labs(
      title = "Distribution of Team Transition Efficiency",
      x = "Transition Points Per Possession",
      y = "Density"
    ) +
    theme_minimal()

  # Combine plots
  combined_plot <- grid.arrange(p1, p2, p3, p4, ncol = 2)

  ggsave("transition_analysis_r.png", combined_plot,
         width = 16, height = 12, dpi = 300)

  cat("Visualizations saved as 'transition_analysis_r.png'\n")

  return(list(p1 = p1, p2 = p2, p3 = p3, p4 = p4))
}

#' Analyze player transition performance
#'
#' @param pbp_data Play-by-play data
#' @return Player transition statistics
analyze_player_transition <- function(pbp_data) {
  cat("\n=== PLAYER TRANSITION ANALYSIS ===\n\n")

  player_stats <- pbp_data %>%
    filter(is_transition == TRUE, !is.na(athlete_id_1)) %>%
    group_by(athlete_id_1, athlete_display_name_1) %>%
    summarise(
      transition_att = n(),
      transition_fgm = sum(score_value > 0, na.rm = TRUE),
      transition_pts = sum(score_value, na.rm = TRUE),
      transition_fg_pct = transition_fgm / transition_att * 100,
      avg_shot_distance = mean(shot_distance, na.rm = TRUE),
      rim_frequency = sum(shot_distance <= 4, na.rm = TRUE) / transition_att * 100,
      assisted_rate = sum(!is.na(assist_player_id), na.rm = TRUE) / transition_att * 100,
      .groups = 'drop'
    ) %>%
    filter(transition_att >= 50) %>%
    arrange(desc(transition_pts))

  cat("Top 20 Players by Transition Points:\n")
  print(head(player_stats, 20))

  cat("\n\nMost Efficient Transition Scorers (min 50 attempts):\n")
  efficient_players <- player_stats %>%
    arrange(desc(transition_fg_pct)) %>%
    head(20)
  print(efficient_players)

  return(player_stats)
}

#' Generate comprehensive report
#'
#' @param team_stats Team statistics
#' @param analysis_results Analysis results
generate_transition_report <- function(team_stats, analysis_results) {
  cat("\n\n" , rep("=", 60), "\n", sep = "")
  cat("COMPREHENSIVE TRANSITION ANALYSIS REPORT\n")
  cat(rep("=", 60), "\n\n", sep = "")

  # League-wide statistics
  cat("LEAGUE-WIDE TRANSITION STATISTICS\n")
  cat(rep("-", 40), "\n", sep = "")
  cat(sprintf("Average Transition Frequency: %.2f%%\n",
              mean(team_stats$transition_freq, na.rm = TRUE)))
  cat(sprintf("Average Transition PPP: %.3f\n",
              mean(team_stats$transition_ppp, na.rm = TRUE)))
  cat(sprintf("Average Half-Court PPP: %.3f\n",
              mean(team_stats$halfcourt_ppp, na.rm = TRUE)))
  cat(sprintf("Average Efficiency Advantage: %.3f\n\n",
              mean(team_stats$efficiency_advantage, na.rm = TRUE)))

  # Top performers
  cat("TOP 5 TEAMS BY TRANSITION EFFICIENCY\n")
  cat(rep("-", 40), "\n", sep = "")
  top_5 <- team_stats %>%
    select(pos_team_name, transition_ppp, transition_freq, efficiency_advantage) %>%
    head(5)
  print(top_5)

  cat("\n\nTOP 5 TEAMS BY TRANSITION VOLUME\n")
  cat(rep("-", 40), "\n", sep = "")
  top_volume <- team_stats %>%
    arrange(desc(transition_freq)) %>%
    select(pos_team_name, transition_freq, transition_possessions, transition_ppp) %>%
    head(5)
  print(top_volume)

  # Correlations
  cat("\n\nKEY CORRELATIONS\n")
  cat(rep("-", 40), "\n", sep = "")

  cor_freq_eff <- cor(team_stats$transition_freq, team_stats$transition_ppp,
                      use = "complete.obs")
  cat(sprintf("Frequency vs Efficiency: r = %.3f\n", cor_freq_eff))

  cor_rim_eff <- cor(team_stats$transition_rim_pct, team_stats$transition_ppp,
                     use = "complete.obs")
  cat(sprintf("Rim Frequency vs Efficiency: r = %.3f\n", cor_rim_eff))

  cor_advantage <- cor(team_stats$transition_freq, team_stats$efficiency_advantage,
                       use = "complete.obs")
  cat(sprintf("Frequency vs Efficiency Advantage: r = %.3f\n", cor_advantage))

  cat("\n", rep("=", 60), "\n\n", sep = "")
}

# Main execution
main <- function() {
  # Set season
  SEASON <- 2024

  cat("Starting NBA Transition Analysis...\n\n")

  # Load data
  pbp_data <- get_transition_pbp(seasons = SEASON)

  # Calculate team statistics
  team_stats <- calculate_team_transition_stats(pbp_data)

  # Analyze factors
  analysis_results <- analyze_transition_factors(pbp_data)

  # Create visualizations
  plots <- visualize_transition_analysis(team_stats, pbp_data)

  # Player analysis
  player_stats <- analyze_player_transition(pbp_data)

  # Generate report
  generate_transition_report(team_stats, analysis_results)

  cat("\nAnalysis complete!\n")

  # Return results
  return(list(
    team_stats = team_stats,
    player_stats = player_stats,
    analysis = analysis_results,
    plots = plots
  ))
}

# Run analysis
results <- main()