app.R

library(shiny)
library(shinythemes)
library(shinydashboard)
library(shinyWidgets)
library(dashboardthemes)
library(dplyr)
library(ggplot2)
library(tidyr)
library(shinycssloaders)
library(ggalluvial)
library(jsonlite)
library(ggridges)
library(ggthemes)
library(networkD3)
library(igraph)
library(ggrepel)
library(bslib)
library(lubridate)

# Randomowe imiona (w rzeczywistosci to z nimi mamy kontakt ale nie chcemy wyjsc
# na fameow...)
imiona <- read.csv("Zeszyt1.csv", sep = ";") %>%
  select(2) %>%
  tail(-1)


#Potrzebne, żeby nazwy miesięcy działały
Sys.setlocale("LC_ALL","English") # na Windowsie
Sys.setlocale("LC_TIME", "en_US.UTF-8") # i na Linuxie



#Dane
m_data <- fromJSON("m_user_data.json")
z_data <- fromJSON("z_user_data.json")
a_data <- fromJSON("a_user_data.json")

#Podstawowy theme do wykresów
my_theme <- theme(plot.background = element_rect(fill = "#010101",color = "#010101"),
                  panel.background = element_rect(fill = "#010101"),
                  legend.background = element_rect(fill = "#010101"),
                  legend.key = element_rect(fill = "#010101"),
                  legend.text = element_text(size = 16, face = "bold"),
                  legend.title = element_text(size = 18, colour = "white", face = "bold"),
                  text = element_text(colour = "white"),
                  axis.line = element_line(colour = "white"),
                  axis.text = element_text(colour = "white", size = 16),
                  axis.title = element_text(size = 18, colour = "white", face = "bold"),
                  panel.border = element_blank(),
                  panel.grid = element_blank())

#dashboardUi
motywUI <- shinyDashboardThemeDIY(
  appFontFamily = "Arial",
  appFontColor = "#FFFFFF",
  primaryFontColor = "#010101"
  ,infoFontColor = "#010101"
  ,successFontColor = "#010101"
  ,warningFontColor = "#010101"
  ,dangerFontColor = "#010101"
  ,bodyBackColor = "#010101"
  
  ### header
  ,logoBackColor = "#EE1D52"
  
  ,headerButtonBackColor = "#EE1D52"
  ,headerButtonIconColor = "#010101"
  ,headerButtonBackColorHover = "rgb(243,143,168)"
  ,headerButtonIconColorHover = "rgb(0,0,0)"
  
  ,headerBackColor = "#EE1D52"
  ,headerBoxShadowColor = "#010101"
  ,headerBoxShadowSize = "1px 1px 1px"
  
  ### sidebar
  ,sidebarBackColor = "#35DBE7"
  ,sidebarPadding = 0
  
  ,sidebarMenuBackColor = "transparent"
  ,sidebarMenuPadding = 0
  ,sidebarMenuBorderRadius = 0
  
  ,sidebarShadowRadius = "1px 1px 1px" 
  ,sidebarShadowColor = "#aaaaaa"
  
  ,sidebarUserTextColor = "rgb(255,255,255)"
  
  ,sidebarSearchBackColor = "rgb(55,72,80)"
  ,sidebarSearchIconColor = "rgb(153,153,153)"
  ,sidebarSearchBorderColor = "rgb(55,72,80)"
  
  ,sidebarTabTextColor = "#A1F0F5"
  ,sidebarTabTextSize = 13
  ,sidebarTabBorderStyle = "none none solid none"
  ,sidebarTabBorderColor = "rgb(35,106,135)"
  ,sidebarTabBorderWidth = 1
  
  ,sidebarTabBackColorSelected = "#A1F0F5"
  
  ,sidebarTabTextColorSelected = "rgb(0,0,0)"
  ,sidebarTabRadiusSelected = "0px 0px 0px 0px"
  
  ,sidebarTabBackColorHover = "#A1F0F5"
  ,sidebarTabTextColorHover = "rgb(50,50,50)"
  ,sidebarTabBorderStyleHover = "none none solid none"
  ,sidebarTabBorderColorHover = "rgb(75,126,151)"
  ,sidebarTabBorderWidthHover = 1
  ,sidebarTabRadiusHover = "0px 0px 0px 0px"
  
  ### boxes
  ,boxBackColor = "rgb(255,255,255)"
  ,boxBorderRadius = 5
  ,boxShadowSize = "0px 1px 1px"
  ,boxShadowColor = "rgba(0,0,0,.1)"
  ,boxTitleSize = 16
  ,boxDefaultColor = "rgb(210,214,220)"
  ,boxPrimaryColor = "rgba(44,222,235,1)"
  ,boxInfoColor = "rgb(210,214,220)"
  ,boxSuccessColor = "rgba(0,255,213,1)"
  ,boxWarningColor = "rgb(244,156,104)"
  ,boxDangerColor = "rgb(255,88,55)"
  
  ,tabBoxTabColor = "rgb(255,255,255)"
  ,tabBoxTabTextSize = 14
  ,tabBoxTabTextColor = "rgb(0,0,0)"
  ,tabBoxTabTextColorSelected = "rgb(0,0,0)"
  ,tabBoxBackColor = "rgb(255,255,255)"
  ,tabBoxHighlightColor = "rgba(44,222,235,1)"
  ,tabBoxBorderRadius = 5
  
  ### inputs
  ,buttonBackColor = "rgb(245,245,245)"
  ,buttonTextColor = "rgb(0,0,0)"
  ,buttonBorderColor = "rgb(200,200,200)"
  ,buttonBorderRadius = 5
  
  ,buttonBackColorHover = "rgb(235,235,235)"
  ,buttonTextColorHover = "rgb(100,100,100)"
  ,buttonBorderColorHover = "rgb(200,200,200)"
  
  ,textboxBackColor = "rgb(255,255,255)"
  ,textboxBorderColor = "rgb(200,200,200)"
  ,textboxBorderRadius = 5
  ,textboxBackColorSelect = "rgb(245,245,245)"
  ,textboxBorderColorSelect = "rgb(200,200,200)"
  
  ### tables
  ,tableBackColor = "rgb(255,255,255)"
  ,tableBorderColor = "rgb(240,240,240)"
  ,tableBorderTopSize = 1
  ,tableBorderRowSize = 1
  
  
)

#strony ui
activity <- fluidPage(
  titlePanel("Activity on TikTok App"),
  textOutput("textActivity"),
  br(),
  sliderInput("dateRange",
              "Choose date range",
              min = as.POSIXct("2022-06-01", format = "%Y-%m-%d"),
              max = as.POSIXct("2022-12-31", format = "%Y-%m-%d"),
              value = c(as.POSIXct("2022-06-01", format = "%Y-%m-%d"),
                        as.POSIXct("2022-12-31", format = "%Y-%m-%d")),
              timeFormat = "%Y-%m-%d",
              width = '100%'),
  tags$style(HTML('#dateRange-label.control-label, *{
        color: white;
  }')),
  br(),
  plotOutput("plot1")%>% 
    shinycssloaders::withSpinner(type = 6,
                                 color = "#EE1D52"),
  textOutput("plot1Comment"),
  br(),
  br(),
  plotOutput("plot7")%>% 
    shinycssloaders::withSpinner(type = 6,
                                 color = "#EE1D52"),
  textOutput("plot7Comment"),
  br(),
  br(),
  plotOutput("plot5")%>% 
    shinycssloaders::withSpinner(type = 6,
                                 color = "#EE1D52"),
  textOutput("plot5Comment"),
  br(),
  plotOutput("plot8")%>% 
    shinycssloaders::withSpinner(type = 6,
                                 color = "#EE1D52"),
  textOutput("plot8Comment"),
)

sharing <- fluidPage(
  titlePanel("How and with whom do we share TikToks?"),
  br(),
  textOutput("textshare"),
  br(),
  plotOutput("plot2")%>% 
    shinycssloaders::withSpinner(type = 6,
                                 color = "#EE1D52"),
  sliderInput("dateRange2",
              "Choose date range",
              min = as.POSIXct("2022-06-01", format = "%Y-%m-%d"),
              max = as.POSIXct("2022-12-31", format = "%Y-%m-%d"),
              value = c(as.POSIXct("2022-06-01", format = "%Y-%m-%d"),
                        as.POSIXct("2022-12-31", format = "%Y-%m-%d")),
              timeFormat = "%Y-%m-%d",
              width = '100%'),
  tags$style(HTML('#dateRange2-label.control-label, *{
        color: white;
  }')),
  textOutput("plot2Comment"),
  br(),
  br(),
  forceNetworkOutput("plot3")%>% 
    shinycssloaders::withSpinner(type = 6,
                                 color = "#EE1D52"),
  textOutput("plot3Comment"),

)

tiktoks <- fluidPage(
  titlePanel("How do we watch TikToks?"),
  br(),
  plotOutput("plot4")%>% 
    shinycssloaders::withSpinner(type = 6,
                                 color = "#EE1D52"),
  fluidRow(
    column(6,
           selectInput("plotType",
                       "Choose type of plot",
                       c("linear","alluvial"),
                       selected = "alluvial")
    ),
    column(6,
           checkboxInput("allTiktoks",
                         label = "Do you want see data for all tiktoks? (Otherwise we will show you only tiktoks liked, shared or added to favourite)",
                         value = FALSE
           )
    )
  ),
  textOutput("plot4Comment"),
  br(),
  plotOutput("plot6")%>% 
    shinycssloaders::withSpinner(type = 6,
                                 color = "#EE1D52"),
  checkboxGroupInput("watchingTime",
                     "Choose watching time ranges",
                     c("<1 sec","1-6 sec","7-15 sec","16-30 sec","31-60 sec","1-3 min",">3 min"),
                     c("<1 sec","1-6 sec","7-15 sec","16-30 sec","31-60 sec","1-3 min",">3 min"),
                     inline = TRUE),
  tags$style(HTML('#plotType-label.control-label, #watchingTime-label.control-label, *{
        color: white;
                  .selectize-dropdown-content {
                  color: black')),
  textOutput("plot6Comment"),
)

shinyUI <- dashboardPage(
  dashboardHeader(title = tags$a(tags$img(src='logo.png',width='100%'))), 
  # dashboardHeader(title = tags$li(a(tags$img(src='logo.png',width='100%')), class = "dropdown"),
  #                 tags$img(src='logo_MINI.png',width='50%', align = "right")
  #                 ),
  
  dashboardSidebar(
    sidebarMenu(
      id = 'tabs',
      menuItem("Activity", tabName = 'activity', icon = icon("chart-line")),
      menuItem("Sharing", tabName = 'sharing', icon = icon("share")),
      menuItem("TikToks", tabName = 'tikToks', icon = icon("tiktok")),
      uiOutput('sidebar'),
      
      uiOutput("changeColors")
    )
  ),
  
  dashboardBody(
    tabItems(
      tabItem(tabName = 'activity', activity),
      tabItem(tabName = 'sharing', sharing),
      tabItem(tabName = 'tikToks', tiktoks)
    )),
  
  motywUI
  
)





server <- function(input, output) {
  
  #wybor uzytkownika
  output$sidebar <- renderUI({
    selectInput("user",
                "Select user data:",
                c("Purple Monster",
                  "Orange Monster",
                  "Red Monster"),
                selected = "Purple Monster")
  })
  
  #zmiana danych w zależności od użytkownika
  data <- reactive({
    if(input$user == "Purple Monster") {
      m_data
    }
    else if(input$user == "Orange Monster") {
      z_data
    }
    else {
      a_data
    }
  })
  
  #zmiana koloru selectInput 
  bg <- reactive({
    case_when(input$user =="Purple Monster" ~ '.irs--shiny .irs-bar,.irs--shiny .irs-to,.irs--shiny .irs-from, input[type=checkbox], .selectize-dropdown-content, .selectize-input.items.has-options.full.has-items {
                                                background-color: purple;
                                                accent-color: purple;
                                                background: purple;
                                                border-top: purple;
                                                border-bottom: purple;
                                                }',
              input$user =="Orange Monster" ~ '.irs--shiny .irs-bar,.irs--shiny .irs-to,.irs--shiny .irs-from,input[type=checkbox], .selectize-dropdown-content, .selectize-input.items.has-options.full.has-items {
                                                background-color: orange;
                                                accent-color: orange;
                                                background: orange;
                                                border-top: orange;
                                                border-bottom: orange;
                                                }',
              input$user =="Red Monster" ~ '.irs--shiny .irs-bar,.irs--shiny .irs-to,.irs--shiny .irs-from,input[type=checkbox], .selectize-dropdown-content, .selectize-input.items.has-options.full.has-items {
                                                background-color: red;
                                                accent-color: red;
                                                background: red;
                                                border-top: red;
                                                border-bottom: red;
                                                }',
              TRUE  ~ ' {.irs--shiny .irs-bar,.irs--shiny .irs-to,.irs--shiny .irs-from,.selectize-dropdown-content, .selectize-input.items.has-options.full.has-items
                                                background-color: blue;
                                                accent-color: blue;
                                                background: blue;
                                                border-top: blue;
                                                border-bottom: blue;
                                                }')
  })
  
  output$changeColors <- renderUI({
    tags$style(HTML(bg()))
  })
  
  output$textActivity <- renderText("This page shows how often you are using TikTok app. Let's have a closer look. But first of all choose time range you want to focus on:")

  #Wykres 1 -barstack
  output$plot1 <- renderPlot({
    
    data()[["Activity"]][["Login History"]][["LoginHistoryList"]] %>% 
      mutate(
        NetworkType = case_when(NetworkType == "Wi-Fi" ~ "Wi-Fi",
                                TRUE ~ "Cellular"),
        Date = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S"),
        Month = factor(format(Date, format = "%b"), levels = month.abb),
        Day = format(Date, format = "%Y-%m-%d")) %>%
      mutate(Day = as.Date(Day)) %>% 
      filter(NetworkType != "") %>% 
      filter(Date <= input$dateRange[2], Date >= input$dateRange[1]) %>% 
      ggplot(aes(x = Day, fill = NetworkType))+
      geom_bar(position = "stack")+
      scale_x_date(date_breaks = "1 month", date_labels = "%b")+
      scale_y_continuous(expand = c(0,0),guide = guide_axis(title = "Number of logins"))+
      theme_gdocs()+
      theme(panel.background = element_rect(fill = "#010101"),
            panel.grid.major.y  = element_line(colour = "#787878" ),
            panel.grid.major.x  = element_blank(),
            axis.text = element_text(colour = "white"),
            axis.title = element_text(colour = "white", size = 18),
            axis.text.x = element_text(colour = "white", size = 16),
            legend.text = element_text(colour = "white", size = 16),
            legend.title = element_text(colour = "white", size = 18),
            axis.text.y = element_text(colour = "white", size = 16),
            plot.background = element_rect(fill = "#010101"),
            legend.background = element_rect(fill = "#010101"))+
      scale_fill_manual(values = c(
                                   "#EE1D52",
                                   "#35DBE7"))
  })

  output$plot1Comment <- renderText("The plot above show the cumulative sum of number of logins for each day. Just by looking at the color proportion we can observe
                                    that majority of our viewed tiktoks have been seen via Cellular of WiFi connection depending on the season.")
  output$textshare <- renderText("In this part of our app you will learn the most popular way in which you share TikToks with your friends. Not only that but you will also learn 
                                 who is your biggest friend on TikTok and who do you have spoken the most in TikTok chat. Let's have a look. You can choose period of time to see 
                                 which communicator was the most popular in the specific period of time.")
  #Wykres2 (udostepniane tiktoki)
  output$plot2 <- renderPlot({
    data()[["Activity"]][["Share History"]][["ShareHistoryList"]] %>% 
      mutate(MethodPob = strtrim(Method, 4)) %>%
      mutate(Method = ifelse(MethodPob == "chat", MethodPob, Method),
             Method = ifelse(Method == "copy", "messenger", Method)) %>%
      select(-MethodPob) %>% 
      mutate(Date = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S"),
             Month = factor(format(Date, format = "%b"), levels = month.abb),
             Day = format(Date, format = "%Y-%m-%d")) %>%
      mutate(Day = as.Date(Day)) %>% 
      filter(Date <= input$dateRange2[2], Date >= input$dateRange2[1]) %>% 
      group_by(Method) %>% 
      summarise(n = n()) %>% 
      ggplot(aes(y = factor(Method, level = c("chat", "instagram_dm", "messenger", "sms")), x = n,  fill = Method))+
      geom_col()+
      scale_fill_manual(values = c("#EE1D52",
                                   "#35DBE7",
                                   "#7D1EFF",
                                   "#FFC5F0"))+
      scale_x_continuous(expand = c(0,0),guide = guide_axis(title = "Collective sum"), limits = c(0,140))+
      scale_y_discrete(expand = c(0,0),guide = guide_axis(title = "Type of connection"),
                       breaks = rev(c("chat", "instagram_dm", "messenger", "sms")))+
      theme(panel.background = element_rect(fill = "#010101"),
            panel.grid.major.y = element_blank(),
            panel.grid.major.x  = element_line(colour = "#787878" ),
            panel.grid.minor.x  = element_line(colour = "#787878" ),
            axis.text = element_text(colour = "white"),
            axis.title = element_text(colour = "white", size = 18),
            axis.text.x = element_text(colour = "white", size = 16),
            legend.text = element_text(colour = "white", size = 16),
            legend.title = element_text(colour = "white", size = 18),
            axis.text.y = element_text(colour = "white", size = 16),
            plot.background = element_rect(fill = "#010101"),
            legend.background = element_rect(fill = "#010101"))
  })
  output$plot2Comment <- renderText("The plot above allows us to compare method in which we were sharing TikToks during the specific period of time. We can easilly notice our tnedencies and habits as well us our favourite 
                                    communicator.")
  
  #wykres 3 (graf)
  output$plot3 <- renderForceNetwork({
    len <- length(data()[["Direct Messages"]][["Chat History"]][["ChatHistory"]]);
    who <- substring(names(data()[["Direct Messages"]][["Chat History"]][["ChatHistory"]]), 19, 
                     nchar(names(data()[["Direct Messages"]][["Chat History"]][["ChatHistory"]]))-1);
    howManyChats <- c();
    for (i in 1:len) {
      howManyChats <- append(howManyChats, 
                             nrow(data()[["Direct Messages"]]
                                  [["Chat History"]]
                                  [["ChatHistory"]]
                                  [[i]]));
    } 
    chatHistory <- data.frame(who, howManyChats);
    
    # Tworzymy graf
    from <- rep(data()[["Profile"]][["Profile Information"]][["ProfileMap"]][["userName"]],
                length(data()[["Direct Messages"]][["Chat History"]][["ChatHistory"]]));
    OurGraph <- chatHistory %>%
      mutate(source = 0) %>%
      mutate(howManyChats = sqrt(howManyChats)) %>%
      mutate(who = 1:length(data()[["Direct Messages"]][["Chat History"]][["ChatHistory"]]));
    OurGraph <- OurGraph %>%
      rename("target" = "who") %>%
      rename("value" = "howManyChats") %>%
      relocate(source, target, value);
    # Wezly
    OurNodes <- data.frame(name = chatHistory$who, 
                           size = round(OurGraph$value, 1)*40, 
                           group = length(data()[["Direct Messages"]][["Chat History"]][["ChatHistory"]]));
    OurNodes <- OurNodes %>%
      mutate(group = case_when(size > 40 ~ "More than one interaction.", 
                               size <= 40 ~ "Just one sent message.")) %>%
      relocate(name, group, size);
    OurNodes <- rbind(c("me", "me", 20), OurNodes)
    len2 <- nrow(OurNodes)
    OurNodes$name <- imiona %>% head(len2)
    OurNodes$name[1, 1] <- "me"
    
    ColourScale <- 'd3.scaleOrdinal()
            .domain(["me", "More than one interaction.", "Just one sent message."])
           .range(["#FFFFFF", "#ff0050", "#00f2ea"]);'
    
    fn <- forceNetwork(Links = OurGraph, Nodes = OurNodes, Source = "source", 
                       Target = "target", Value = "value", NodeID = "name", 
                       Group = "group", opacity = 1, zoom = TRUE, Nodesize = "size",
                       linkDistance = 100, linkWidth = 2, charge = -200,
                       opacityNoHover = 1, legend = TRUE, fontFamily = "Arial", 
                       colourScale = JS(ColourScale))
    
    htmlwidgets::onRender(
      fn,
      'function(el, x) { 
    d3.select("body").style("background-color", "#000000");
    d3.selectAll(".legend text").style("fill", "white");
    d3.selectAll("text").style("fill", "white");
  }'
    )
  })
  
  output$plot3Comment <- renderText("The plot above shows by the means of graph visualisation connections with other people.")
  
  
  #wykres4 (alluvial + wersja liniowa)
  output$plot4 <- renderPlot({
    all <- data()[["Activity"]][["Video Browsing History"]][["VideoList"]]
    likes <- data()[["Activity"]][["Like List"]][["ItemFavoriteList"]] %>% 
      mutate(Like = "liked")
    shared <- data()[["Activity"]][["Share History"]][["ShareHistoryList"]] %>% 
      filter(SharedContent == "video") %>% 
      mutate(Shared = "shared") %>% 
      select(Date, Link, Shared)
    favourite <- data()[["Activity"]][["Favorite Videos"]][["FavoriteVideoList"]] %>% 
      mutate(Favourite = "favourite")
    if(input$plotType == "alluvial") {
      if(input$allTiktoks == TRUE) {
        data4 <- full_join(all,likes,by = c("Date","Link"))
      } else {
        data4 <- likes
      }
      data4 <- 
        data4 %>% 
        full_join(shared,by = c("Date","Link")) %>% 
        full_join(favourite, by = c("Date","Link")) %>% 
        mutate(Like = coalesce(Like,"not liked"),
               Shared = coalesce(Shared,"not shared"),
               Favourite = coalesce(Favourite, "not favourite")) %>%
        group_by(Like, Shared, Favourite) %>% 
        summarise(Freq = n())
      ggplot(data4,
             aes(y = Freq, axis1 = Like, axis2 = Shared, axis3 = Favourite)) +
        geom_alluvium(fill = "#ee1d52", width = 1/12) +
        geom_stratum(width = 1/12) +
        geom_label(stat = "stratum", aes(label = after_stat(stratum))) +
        labs(y = "Number of tiktoks") +
        my_theme +
        theme(axis.text = element_blank(),
              axis.ticks = element_blank())
    } else {
      likes2 <- likes %>% 
        mutate(Date = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S"),
               Month = factor(format(Date, format = "%b"),levels = month.abb)) %>% 
        group_by(Month) %>% 
        summarise(Freq = n())
      shared2 <- shared %>% 
        mutate(Date = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S"),
               Month = factor(format(Date, format = "%b"),levels = month.abb)) %>% 
        group_by(Month) %>% 
        summarise(Freq = n())
      favourite2 <- favourite %>% 
        mutate(Date = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S"),
               Month = factor(format(Date, format = "%b"),levels = month.abb)) %>% 
        group_by(Month) %>% 
        summarise(Freq = n())
      
      if(input$allTiktoks == TRUE) {
        all2 <- all %>% 
          mutate(Date = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S"),
                 Month = factor(format(Date, format = "%b"),levels = month.abb)) %>% 
          group_by(Month) %>% 
          summarise(Freq = n())
        
        likes2 %>% 
          ggplot(aes(x = Month, y = Freq)) +
          geom_line(aes(group = 1, color = "like"), size = 1.5) +
          geom_line(data = shared2, aes(group = 1, color = "shared"), size = 1.5) +
          geom_line(data = favourite2, aes(group = 1, color = "favourite"), size = 1.5) +
          geom_line(data = all2, aes(group = 1, color = "all"), size = 1.5) +
          scale_color_manual(values = c("#D3D3D3", "#EE1D52","#69C9D0","#7D1EFF")) +
          labs(color = "Option",
               y = "Number of tiktoks") +
          my_theme +
          theme(panel.grid.major  = element_line(colour = "#787878"))
      } else {
        likes2 %>% 
          ggplot(aes(x = Month, y = Freq)) +
          geom_line(aes(group = 1, color = "like"), size = 1.5) +
          geom_line(data = shared2, aes(group = 1, color = "shared"), size = 1.5) +
          geom_line(data = favourite2, aes(group = 1, color = "favourite"), size = 1.5) +
          scale_color_manual(values = c("#EE1D52","#69C9D0","#7D1EFF")) +
          labs(color = "Option",
               y = "Number of tiktoks") +
          my_theme +
          theme(panel.grid.major  = element_line(colour = "#787878"))
      }
      
    }
  })
  
  output$plot4Comment <- renderText(
    {"The plot above shows number of tiktoks with given characteristics: liked, shared or added to favourite."}
  )
  
  #wykres5 (heatmapy)
  output$plot5 <- renderPlot({
    login <- data()[["Activity"]][["Login History"]][["LoginHistoryList"]]  %>% 
      mutate(DatePOSIX = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S")) %>%
      filter(DatePOSIX <= input$dateRange[2], DatePOSIX >= input$dateRange[1]) %>%
      select(-7)
    
    login$Date <- login$Date %>%
      strtrim(19) %>%
      strptime(format = "%Y-%m-%d %H:%M:%S") 
    login$weekday <- c("Sunday", "Monday", "Tuesday",   
                       "Wednesday", "Thursday", "Friday",
                       "Saturday")[as.POSIXlt(login$Date)$wday + 1]
    loginHistory <- login %>%
      mutate(month = format(as.Date(Date, format = "%Y-%m-%d %H:%M:%S"),"%m")) %>%
      mutate(day = day(Date)) %>%
      mutate(hour = hour(Date)) %>%
      select(-2, -3, -4, -5, -6)
    case <- 0
    if (case == 0) {
      Matrix <- loginHistory %>%
        select(2, 5)
    } else {
      Matrix <- loginHistory %>%
        select(3, 4)
    }
    Matrix <- data.frame(table(Matrix)) %>%
      rename("X" = 1) %>%
      rename("Y" = 2)
    if (case == 0) {
      Matrix$X <- factor(Matrix$X,levels = c("Monday", "Tuesday", "Wednesday", "Thursday",
                                             "Friday", "Saturday", "Sunday"))
    }
    ggp <- ggplot(Matrix, aes(X, Y)) +          
      geom_tile(aes(fill = Freq)) + 
      scale_fill_gradient(low = "black", high = case_when(case == 0 ~ "#25F4EE", 
                                                          case == 1 ~ "#FE2C55")) +
      xlab(case_when(case == 0 ~ "Day of the Week", 
                     case == 1 ~ "Month")) +
      ylab(case_when(case == 0 ~ "Hour", 
                     case == 1 ~ "Day of the Month")) +
      labs(fill = "Number of Logins")  +
      my_theme
    
    ggp  
  })
  output$plot5Comment <- renderText(
    {"These plots, one above and one below, are the heatmap of a TikTok usage."}
  )
  
  #wykres8 (heatmapy) ciag dalszy
  output$plot8 <- renderPlot({
    login <- data()[["Activity"]][["Login History"]][["LoginHistoryList"]]  %>% 
      mutate(DatePOSIX = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S")) %>%
      filter(DatePOSIX <= input$dateRange[2], DatePOSIX >= input$dateRange[1]) %>%
      select(-7)
    
    login$Date <- login$Date %>%
      strtrim(19) %>%
      strptime(format = "%Y-%m-%d %H:%M:%S") 
    login$weekday <- c("Sunday", "Monday", "Tuesday",   
                       "Wednesday", "Thursday", "Friday",
                       "Saturday")[as.POSIXlt(login$Date)$wday + 1]
    login$month <- month.abb[month(as.POSIXlt(login$Date))]
    loginHistory <- login %>%
      mutate(day = day(Date)) %>%
      mutate(hour = hour(Date)) %>%
      select(-2, -3, -4, -5, -6)
    case <- 1
    if (case == 0) {
      Matrix <- loginHistory %>%
        select(2, 5)
    } else {
      Matrix <- loginHistory %>%
        select(3, 4)
    }
    Matrix <- data.frame(table(Matrix)) %>%
      rename("X" = 1) %>%
      rename("Y" = 2)
    if (case == 0) {
      Matrix$X <- factor(Matrix$X,levels = c("Monday", "Tuesday", "Wednesday", "Thursday",
                                             "Friday", "Saturday", "Sunday"))
    } else {
      Matrix$X <- factor(Matrix$X, levels = month.abb)
    }
    ggp <- ggplot(Matrix, aes(X, Y)) +          
      geom_tile(aes(fill = Freq)) + 
      scale_fill_gradient(low = "black", high = case_when(case == 0 ~ "#25F4EE", 
                                                          case == 1 ~ "#FE2C55")) +
      xlab(case_when(case == 0 ~ "Day of the Week", 
                     case == 1 ~ "Month")) +
      ylab(case_when(case == 0 ~ "Hour", 
                     case == 1 ~ "Day of the Month")) +
      labs(fill = "Number of Logins")  +
      my_theme
    
    ggp  
  })
  
  output$plot8Comment <- renderText(
    {""}
  )
  
  #wykres6 (tiktoki w)
  output$plot6 <- renderPlot({
    data6 <- data()[["Activity"]][["Video Browsing History"]][["VideoList"]] %>% 
      select(Date) %>% 
      mutate(Date = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S"),
             Month = factor(format(Date, format = "%b"), levels = month.abb)) %>% 
      transform(Time = as.numeric(difftime(Date, c(Date[-1],NA), units = "secs"))) %>% 
      head(-1) %>% 
      filter(Time <= 10*60) %>%
      mutate(Time_range = factor(case_when(Time < 1 ~ "<1 sec",
                                           Time <= 6 ~ "1-6 sec",
                                           Time <= 15 ~ "7-15 sec",
                                           Time <= 30 ~ "16-30 sec",
                                           Time <= 60 ~ "31-60 sec",
                                           Time <= 180 ~ "1-3 min",
                                           TRUE ~ ">3 min"),
                                 levels = c("<1 sec","1-6 sec","7-15 sec","16-30 sec","31-60 sec","1-3 min",">3 min"))) %>% 
      filter(Time_range %in% input$watchingTime) %>% 
      group_by(Time_range,Month) %>% 
      summarise(Freq = n()) %>% 
      ungroup() %>% 
      group_by(Month) %>% 
      summarise(Sum = sum(Freq), Time_range = Time_range, Freq = Freq/Sum)
    colours <- c("#69c9d0", "#00aff8", "#0086ff", "#7d1eff",
                 "#d600bc", "#f1007f", "#ee1d52")
    ggplot(data6, aes(y = Freq, x = Month, fill =Time_range)) +
      geom_col(position = "dodge") +
      scale_y_continuous(labels = scales::percent,expand = c(0,0)) +
      labs(y = "Percentage of tiktoks",
           fill = "Watching time") +
      scale_fill_manual(values = colours) +
      my_theme +
      theme(panel.grid.major.y  = element_line(colour = "#787878"))
  })
  
  output$plot6Comment <- renderText(
    {"The plot above shows for every month percentage of watched tiktoks with selected watching time ranges."}
  )
  
  #wykres7 (ggridges)
  output$plot7 <- renderPlot({
    data7 <- data()[["Activity"]][["Login History"]][["LoginHistoryList"]] %>% 
      select(Date, NetworkType) %>% 
      mutate(NetworkType = case_when(NetworkType == "Wi-Fi" ~ "Wi-Fi",
                                     TRUE ~ "Cellular"),
             Date = as.POSIXct(Date, format = "%Y-%m-%d %H:%M:%S"),
             Month = factor(format(Date, format = "%b"), levels = month.abb),
             Day = format(Date, format = "%Y-%m-%d")) %>%
      mutate(Day = as.Date(Day)) %>% 
      filter(Date <= input$dateRange[2], Date >= input$dateRange[1])
    ggplot(data7, aes(x = Day, y = Month, fill = NetworkType, color = NetworkType)) + 
      geom_density_ridges(scale = 2, alpha = 0.7) +
      scale_x_date(date_breaks = "1 month", date_labels = "%b") +
      scale_fill_manual(values = c("#EE1D52","#69C9D0"), name = "Network Type") +
      scale_color_manual(values = c("#EE1D52","#69C9D0"), name = "Network Type") +
      my_theme +
      theme(panel.grid.major  = element_line(colour = "#787878"))
  })
  output$plot7Comment <- renderText(
    {"The plot above shows how the number of tiktoks watched via cellular network vs Wi-FI changed on the following days of each month."}
  )
  
  
}


shinyApp(ui = shinyUI, server = server)