RAPTOR Shiny

#Data from 538 and Neil Paine

library(shiny)
library(ggplot2)
library(dplyr)
library(tidyverse)
library(hoopR)
library(png)
library(grid)
library(ggimage)
library(plotly)



# Load data
latest_data <- read.csv("C:\\Users\\imksy\\Downloads\\latest_RAPTOR_by_player.csv") #2023 season 538 RAPTOR data
historical_data <- read.csv("C:\\Users\\imksy\\Downloads\\historical_RAPTOR_by_player.csv") #previous decade 538 RAPTOR data
estimated_data <- read.csv("C:\\Users\\imksy\\Downloads\\2024-NBA-Stat-Sheet - Players by Player.csv") #scraped data eRT from Neil Paine

# Get NBA logos
tictoc::tic()
progressr::with_progress({
  nba_logos <- hoopR::espn_nba_teams()
})


# Update team abbreviations
team_logos <- nba_logos %>%
  select(abbreviation, logo) %>%
  rename(team_abbreviation = abbreviation) %>%
  mutate(team_abbreviation = recode(team_abbreviation,
                                    "CHA" = "CHO",
                                    "BKN" = "BRK",
                                    "PHX" = "PHO",
                                    "NY" = "NYK",
                                    "GS" = "GSW",
                                    "UTAH" = "UTA",
                                    "NO" = "NOP",
                                    "WSH" = "WAS", 
                                    "SA" = "SAS"))
  

# Merge with estimated_data
merged_data <- estimated_data %>%
  left_join(team_logos, by = c("Franch" = "team_abbreviation"))

# Combine data for the second tab
combined_data <- bind_rows(latest_data, historical_data) %>%
  group_by(player_name) %>%
  mutate(career_RAPTOR_total = sum(raptor_total, na.rm = TRUE),
            total_minutes = sum(mp, na.rm = TRUE),
            RAPTOR_per_minute = sum(raptor_total, na.rm = TRUE) / sum(mp, na.rm = TRUE)) %>%
  filter(total_minutes >= 10000) %>%  # Include only players with at least 10,000 minutes
  mutate(RAPTOR_per_minute = format(RAPTOR_per_minute, digits = 10, scientific = FALSE) %>% as.numeric())

# Define UI
ui <- fluidPage(
  titlePanel("NBA RAPTOR Statistics"),
  navbarPage("Data",
             tabPanel("Estimated Raptor Data by Season",
                      sidebarLayout(
                        sidebarPanel(
                          selectInput("Player", "Select Player:", choices = NULL)
                        ),
                        mainPanel(
                          plotOutput("raptorPlot"),
                          tableOutput("raptorTable"),
                          downloadButton("downloadData", "Download Data")
                        )
                      )),
             tabPanel("538 Raptor Data Careers",
                        mainPanel(
                          plotlyOutput("combinedPlot"),
                          tableOutput("combinedTable"),
                          downloadButton("downloadCombinedData", "Download Data")
                        )
                      ))
  )


# Define server logic
server <- function(input, output, session) {
  
  # Update select input for estimated data tab
  observe({
    updateSelectizeInput(session, "Player",
                         choices = unique(merged_data$Player),
                         server = TRUE)
  })
  
  # Prepare combined data
  combined_data <- reactive({
    bind_rows(latest_data, historical_data) %>%
      group_by(player_name) %>%
      summarise(career_RAPTOR_total = sum(raptor_total, na.rm = TRUE),
                total_minutes = sum(mp, na.rm = TRUE),
                RAPTOR_per_minute = sum(raptor_total, na.rm = TRUE) / sum(mp, na.rm = TRUE)) %>%
      filter(total_minutes >= 10000) %>%
      mutate(RAPTOR_per_minute = as.numeric(format(RAPTOR_per_minute, digits = 10, scientific = FALSE)))
  })
  
  # Plot for estimated data
  output$raptorPlot <- renderPlot({
    selected_player_data <- merged_data %>%
      filter(Player == input$Player, Type == "RS") %>%
      arrange(Year)
    
    if (nrow(selected_player_data) == 0) {
      return(ggplot() + 
               labs(title = "No data available", x = NULL, y = NULL) + 
               theme_void())
    }
    
    min_eRT <- min(selected_player_data$eRT, na.rm = TRUE)
    max_eRT <- max(selected_player_data$eRT, na.rm = TRUE)
    
    ggplot(selected_player_data, aes(x = Year, y = eRT, group = Player)) +
      geom_line() +
      geom_image(aes(image = logo), size = 0.1) +
      scale_x_continuous(breaks = unique(selected_player_data$Year), 
                         labels = unique(selected_player_data$Year)) +
      scale_y_continuous(breaks = scales::pretty_breaks(n = 10),
                         limits = c(min_eRT, max_eRT)) +
      theme_minimal() +
      labs(title = paste("Year-by-Year RAPTOR for", input$Player),
           x = "Year", y = "RAPTOR Total") +
      theme(legend.position = "bottom") +
      guides(shape = guide_legend(title = "Team")) +
      theme(panel.grid.major = element_line(color = "grey"),
            panel.grid.minor = element_line(color = "lightgrey")) + 
      annotate("text", x = Inf, y = -Inf, label = "Data courtesy of Neil Paine",
               hjust = 1, vjust = -1, size = 3, color = "black", fontface = "italic")
  })
  
  # Table for estimated data
  output$raptorTable <- renderTable({
    selected_player_data <- merged_data %>%
      filter(Player == input$Player, Type == "RS") %>%
      arrange(Year)
    
    selected_player_data
  })
  
  # Download data for estimated data
  output$downloadData <- downloadHandler(
    filename = function() { "career_RAPTOR_data.csv" },
    content = function(file) {
      write.csv(merged_data, file)
    }
  )
  
  # Plot for combined data
  output$combinedPlot <- renderPlotly({
    data <- combined_data()
    
    # Calculate medians
    x_median <- median(data$RAPTOR_per_minute, na.rm = TRUE)
    y_median <- median(data$career_RAPTOR_total, na.rm = TRUE)
    
    # Assign quadrant colors
    data <- data %>%
      mutate(color = case_when(
        RAPTOR_per_minute > x_median & career_RAPTOR_total > y_median ~ "green",
        RAPTOR_per_minute > x_median & career_RAPTOR_total <= y_median ~ "orange",
        RAPTOR_per_minute <= x_median & career_RAPTOR_total > y_median ~ "yellow",
        TRUE ~ "red"
      ))
    
    p <- plot_ly(data, x = ~RAPTOR_per_minute, y = ~career_RAPTOR_total, 
                 type = 'scatter', mode = 'markers',
                 text = ~paste("Player: ", player_name, "<br>Total RAPTOR: ", career_RAPTOR_total, "<br>RAPTOR per Minute: ", RAPTOR_per_minute, "<br>Minutes Played: ", total_minutes),
                 hoverinfo = 'text',
                 marker = list(size = 10, color = ~color, opacity = 0.6)) %>%
      layout(title = "RAPTOR per Minute vs. Total RAPTOR",
             xaxis = list(title = "RAPTOR per Minute",
                          gridcolor = 'lightgrey', # Color of the gridlines
                          gridwidth = 1),          # Width of the gridlines
             yaxis = list(title = "Total RAPTOR",
                          gridcolor = 'lightgrey', # Color of the gridlines
                          gridwidth = 1),          # Width of the gridlines
             hovermode = 'closest') %>%
      add_segments(x = x_median, xend = x_median,
                   y = min(data$career_RAPTOR_total, na.rm = TRUE), yend = max(data$career_RAPTOR_total, na.rm = TRUE),
                   line = list(color = 'grey', width = 2, dash = 'dot'), # Dotted line
                   name = 'X Median') %>%
      add_segments(x = min(data$RAPTOR_per_minute, na.rm = TRUE), xend = max(data$RAPTOR_per_minute, na.rm = TRUE),
                   y = y_median, yend = y_median,
                   line = list(color = 'grey', width = 2, dash = 'dot'), # Dotted line
                   name = 'Y Median') %>%
      event_register("plotly_hover") %>%
      event_register("plotly_unhover")
    
    p
  })
  
  # # Table for combined data with selected player's statistics
  # output$combinedTable <- renderTable({
  #   click_data <- event_data("plotly_click")
  #   
  #   # Handle click data and errors
  #   if (is.null(click_data) || is.null(click_data$text)) {
  #     return(NULL)
  #   }
  #   
  #   # Extract player name from the click event data
  #   clicked_player <- stringr::str_extract(click_data$text, "Player: (.+?)<br>") %>%
  #     stringr::str_remove("Player: ")
  #   
  #   if (!is.na(clicked_player) && clicked_player != "") {
  #     filtered_data <- combined_data() %>%
  #       filter(player_name == clicked_player)
  #     
  #     if (nrow(filtered_data) > 0) {
  #       filtered_data %>%
  #         mutate(RAPTOR_per_minute = sprintf("%.10f", RAPTOR_per_minute))
  #     } else {
  #       return(NULL)
  #     }
  #   } else {
  #     return(NULL)
  #   }
  # })
  
  # Download data for combined data
  output$downloadCombinedData <- downloadHandler(
    filename = function() { "combined_RAPTOR_data.csv" },
    content = function(file) {
      write.csv(combined_data(), file)
    }
  )
}

# Run the app
shinyApp(ui = ui, server = server)