app.R

library(shiny)
library(tidyverse)
library(shinyWidgets)
library(purrr)
library(rmarkdown)
library(here)
library(shinycssloaders)

# remove stings in Runtime column
movies <- read.csv("data/imdb_top_1000.csv", stringsAsFactors = FALSE) %>%
  mutate(Runtime = as.numeric(gsub("([0-9]+).*$", "\\1", Runtime))) %>%
  mutate(Gross = as.numeric(gsub(",", "", Gross)) / 1000000)

plot_movies <- separate_rows(movies, Genre, sep = ",")
plot_movies$Genre <- trimws(plot_movies$Genre)
plot_movies$Series_Title <- trimws(plot_movies$Series_Title)

# Define side panel
sidepanels <- sidebarPanel(
  pickerInput("genre", "Genre(s):",
              choices = sort(unique(str_replace_all(string = unlist(strsplit(movies$Genre, ",")), pattern = " ", repl = ""))),
              options = list(`actions-box` = TRUE), multiple = T, selected = "Drama"
  ),
  # Select single star
  pickerInput("star", "Star(s):",
              choices = sort(unique(unlist(select(movies, Star1:Star4) %>% as.list()))),
              options = list(`actions-box` = TRUE, `live-search` = TRUE), multiple = T, selected = "Morgan Freeman"
  ),
  # select min movie Revenue
  sliderInput("minRevenue", "Min Revenue in Million:",
              min = 0,
              max = round(max(na.omit(movies$Gross))) - (round(max(na.omit(movies$Gross))) %% 100),
              value = median(na.omit(movies$Gross)), step = 50, pre = "$", post = "M"
  ),
  # select range of released year
  sliderInput("year", "Released Year Range:",
              min = suppressWarnings(min(na.omit(parse_number(movies$Released_Year)))),
              max = suppressWarnings(max(na.omit(parse_number(movies$Released_Year)))),
              value = c(1920, 2020)
  ),
  # Select range of runtime
  sliderInput("runtimes", "Runtimes Range (Minutes):",
              min = min(movies$Runtime),
              max = max(movies$Runtime),
              value = c(70, 200)
  )
)

# Create navbar page with tabs
ui <- navbarPage(
  "IMDB_Viz_R", # app name
  # First tab with shared side panel 
  tabPanel(
    "IMDB Movie", 
    sidebarLayout(
      sidepanels,
      mainPanel(
        tabsetPanel( 
          tabPanel("Top 3 Movie Recommendations",
                   shinycssloaders::withSpinner(
                     htmlOutput("picture")
                   ),
                   downloadButton("download", "Download .tsv"),  # UI for download button
                   downloadButton("report", "Download Report")
                  ),
          
          tabPanel(
            "Top Rated movies by Genre",
            shinycssloaders::withSpinner(
              plotOutput("movie_genre", height = "500px", width = "800px")
            ),
            br(),
            
            tags$h2( style = "font-size: 15px; font-weight: bold;",textOutput("note"))
            
          ),
          
          tabPanel(
            "Ratings by Genre",
            shinycssloaders::withSpinner(
              plotOutput("boxplot_rg", height = "500px", width = "800px")
            )
          ),
          
          tabPanel(
            "Runtimes by Genre",
            shinycssloaders::withSpinner(
              plotOutput("boxplot", height = "500px", width = "800px")
            )
          ),
          
          tabPanel(
            "Movies by Genre",
            shinycssloaders::withSpinner(
              plotOutput("barplot", height = "500px", width = "800px")
            )
            )
        )
      )
    )
  )
  

)



wrangled_data<-function(df,input){reactive({
  req(input$minRevenue, input$year, input$runtimes, input$genre, input$star)
  df %>%
    filter(.data$Gross >= input$minRevenue) %>% # filter by min gross revenue
    filter(.data$Released_Year >= input$year[1] & .data$Released_Year <= input$year[2]) %>% # filter by released year range
    filter(.data$Runtime >= input$runtimes[1] & .data$Runtime <= input$runtimes[2]) %>% # filter by runtimes range
    filter(str_detect(.data$Genre, paste(input$genre, collapse = "|"))) %>% # filter by genre(s)
    filter(.data$Star1 %in% input$star | .data$Star2 %in% input$star | .data$Star3 %in% input$star | .data$Star4 %in% input$star) # filter by actor(s)
})}

genre_wrangled_data<-function(df,input){reactive({
  req(input$genre)
  df %>%
    filter(str_detect(.data$Genre, paste(input$genre, collapse = "|")))  %>%
    distinct( Series_Title, .keep_all = TRUE) %>%
  arrange(desc(IMDB_Rating)) %>%
            head(5) 
  
})}


# Server function
server <- function(input, output) {
  # use reactive to avoid duplication
  filtered_data <- wrangled_data(movies,input)
  plot_data <- wrangled_data(plot_movies, input)
  genre_data <- genre_wrangled_data(plot_movies, input)
  
  output$note <- renderText({
    req(genre_data()) 
    paste("Note: This plot only shows the top-rated movies for a specific genre and is not based on the rest of the filters.")
  })
  
  # output for movie recommendations

  output$picture <- renderUI({
    req(filtered_data())
    urls <- filtered_data() %>%
      select(Poster_Link) %>%
      head(3) %>%
      pull() # get top 3 movie poster urls
    titles <- filtered_data() %>%
      select(Series_Title) %>%
      head(3) %>%
      pull() # get top 3 movie titles
    overviews <- filtered_data() %>%
      select(Overview) %>%
      head(3) %>%
      pull() # get top 3 movie overviews
    img_tags <- mapply(function(url, caption, text) {
      img_tag <- paste0("<img src='", url, "' style='width: 100px; height: auto; display:block; margin:0 auto;'>")
      title_tag <- paste0("<div style='text-align:center; margin-bottom:5px; word-wrap: break-word; word-break: break-all; font-weight:bold; font-size:12px;'>", caption, "</div>")
      overview_tag <- paste0("<div style='text-align:center; word-wrap: break-word; word-break: break-all; max-width: 120px; line-height: 1.4; margin-top: 5px; font-size:10px;'>", text, "</div>")
      paste0("<div class='column'","<div style='display:inline-block; margin:50px; vertical-align:top;'>", title_tag, img_tag, overview_tag, "</div>", "</div>")
    }, urls, titles, overviews, SIMPLIFY = FALSE)
    HTML(paste0("<div class='row'", img_tags, "</div>"))
  })

  # output for movie ratings distribution plot

  output$boxplot_rg <- renderPlot({
    req(plot_data())
    ggplot(
      plot_data(),
      aes(x = IMDB_Rating, y = Genre, fill = Genre)
    ) +
      geom_boxplot() +
      theme(
        plot.title = element_text(size = 20, face = "bold"),
        axis.title = element_text(size = 15, face = "bold"),
        axis.text = element_text(size = 12, face = "bold")
      ) +
      labs(x = "IMDB Rating", y = "Selected Genres") +
      ggtitle("Distribution of IMDB Ratings by Genre") +
      scale_x_continuous(breaks = seq(7.0, 10.0, by = 0.2))
  })
  
  # output for movie runtime distribution plot
  
  output$boxplot <- renderPlot({
    req(plot_data())
    ggplot(
      plot_data(),
      aes(x = Runtime, y = Genre, fill = Genre)
    ) +
      geom_boxplot() +
      theme(
        plot.title = element_text(size = 20, face = "bold"),
        axis.title = element_text(size = 15, face = "bold"),
        axis.text = element_text(size = 12, face = "bold")
      ) +
      labs(x = "Movie Runtime(mins)", y = "Selected Genres") +
      ggtitle("Distribution of Runtimes by Genre") +
      scale_x_continuous(breaks = seq(40, 321, by = 10))
  })
  
  # output for movie number count barplot
  output$barplot <- renderPlot({
    req(plot_data())
    ggplot(
      plot_data(),
      aes(y = Genre, fill = Genre)
    ) +
      geom_bar() +
      theme(
        plot.title = element_text(size = 20, face = "bold"),
        axis.title = element_text(size = 15, face = "bold"),
        axis.text = element_text(size = 12, face = "bold")
      ) +
      labs(x = "Number of movies", y = "Selected Genres") +
      ggtitle("Number of Movies by Genres")
  })
  
  # output for top 5 movies by genre
  output$movie_genre <- renderPlot({
    req(genre_data()) 
    
    ggplot(genre_data(), aes(x = IMDB_Rating, y = reorder(Series_Title, IMDB_Rating), fill = IMDB_Rating)) +
      geom_bar(stat = "identity") +
      geom_text(aes(label = Series_Title), hjust = 1.1, size = 7) +
      scale_fill_gradient(low = "lightyellow", high = "red") +
      theme(axis.text.y = element_blank()) +
      theme(
        plot.title = element_text(size = 20, face = "bold"),
        axis.title = element_text(size = 15, face = "bold"),
        axis.text = element_text(size = 12, face = "bold")
      ) +
      guides(fill = FALSE)+
      coord_cartesian(xlim = c(7, 9.5)) +
      labs(x = "Rating", y = "") +
      ggtitle(paste("Top 5 Movies by Rating in the selected Genre"))
  })
  
  
  # output for downloading filtered dataset
  output$download <- downloadHandler(
    filename = 'filtered_IMDB.tsv',
    content = function(file) {
      vroom::vroom_write(filtered_data(), file)
    }
  )

  # output for downloading the report
  output$report <- downloadHandler(
    filename = "IMDB_Viz_R_App_Report.html",
    content = function(file){
      
      tempdirpath <- file.path(tempdir(), "Report.Rmd")
      file.copy("reports/Report.Rmd", tempdirpath, overwrite = TRUE)
      
      params <- list(
        genre = input$genre,
        star = input$star,
        minRevenue = input$minRevenue,
        year = input$year,
        runtimes = input$runtimes)
      
      rmarkdown::render(tempdirpath,
                        output_file = file,
                        params = params,
                        envir = new.env(parent = globalenv()))
    }
  )
}

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