server.R

server <- function(input, output, session) {
  
  # READ DATA ####
  
  carbon_budgets <- importIPCCData()
  
  countries_emissions <- importCountriesEmissionData()
  
  continents_emissions <- importContinentsEmissionData()
  
  historical_temperatures <- read.csv("src/ourworldindata/temperature-anomaly.csv", stringsAsFactors = F) %>%
    filter(entity == "Global") %>%
    select(-entity) %>%
    mutate(data_id = "asd")
  
  global_emissions <- countries_emissions %>%
    group_by(year) %>%
    summarize(emissions = sum(emissions, na.rm = T))
  
  projection_data <- read.csv("src/climateactiontracker/EmissionsGaps_mod.csv", stringsAsFactors = F)
  
  # INPUTS ####
  
  # Base year: Year before allocation starts
  base_year <- reactive({
    # On page "Years left"
    allocation_date() - 1
  })
  
  
  selected_probability <- reactive({
    input$selected_probability
  })
  
  selected_warming_degrees <- reactive({
    input$selected_warming_degrees
  })
  
  calculation_approach <- reactive({
    input$selected_calculation_approach
  })
  
  # Selected countries (these inputs are synchronized by means of observers)
  
  selected_countries <- reactive({
    # On page "Years left"  
    input$selected_countries
  }) %>%
    # Debounce skips intermediary values and therefore avoids "bouncing"
    debounce(1000)
  
  selected_countries_2 <- reactive({
    # On page "Budget left"  
    input$selected_countries_2
  }) %>%
    debounce(1000)
  
  # Selected allocation date: Date when allocation starts (beginning of a year)
  # (these inputs are synchronized by means of observers)
  
  allocation_date <- reactive({
    input$alloc_date
  }) %>%
    debounce(1000)
  
  allocation_date_2 <- reactive({
    input$alloc_date_2
  }) %>%
    debounce(1000)
  
  allocation_date_3 <- reactive({
    input$alloc_date_3
  }) %>%
    debounce(1000)
  
  # OPTIONS ####
  
  options(scipen = 999) # Disable scientific notation
  
  # IMPORTANT VARIABLES ####
  
  # Date where the IPCC budgets used were published (Beginning of year)
  ipcc_budget_date <- 2018 # For the IPCC SR1.5 report
  
  # Minimum and maximum years of available emissions data
  maximum_year <- max(countries_emissions$year)
  minimum_year <- min(countries_emissions$year)
  
  # Total country emissions since allocation date
  total_country_emis_since_allocation_date <- reactive({
    countries_emissions %>%
      filter(year >= allocation_date()) %>%
      group_by(country) %>%
      summarize(total_emis_since_by_gt = sum(emissions)/1000000000) %>%
      ungroup()
  })
  
  # Cumulated country emissions since allocation date
  cumulated_country_emis_since_allocation_date <- reactive({
    countries_emissions %>%
      filter(year >= allocation_date()) %>%
      group_by(country) %>%
      mutate(cumulated_emis_since_ad_gt = cumsum(emissions)/1000000000) %>%
      ungroup() %>%
      select(country, year, cumulated_emis_since_ad_gt)
  })
  
  
  # Total global emissions between allocation date and latest year with data available
  total_global_emis_since_allocation_date <- reactive({
    global_emissions %>%
      filter(year >= allocation_date() & year <= maximum_year) %>%
      summarize(total_global_emis = sum(emissions)) %>%
      as.numeric()
  })
  
  # Total global emissions between allocation date and reference date of IPCC budgets
  total_global_emis_between_allocdate_ipccbudgetdate <- reactive({
    global_emissions %>%
    filter(year >= allocation_date() & year < ipcc_budget_date) %>% # Emissions of years before IPCC budgets apply
      summarize(total_global_emis = sum(emissions)) %>%
      as.numeric()
  })
  
  # Global emissions in base year
  global_emissions_base_year <- reactive({
    global_emissions %>%
      filter(year == base_year()) %>%
      select(emissions) %>%
      as.numeric()
  })
  
  selected_carbon_budget_sr15 <-  reactive({
    carbon_budgets %>%
      filter(probability == selected_probability(),
             warming_degrees == selected_warming_degrees()) %>%
      select(budget_gt) %>%
      as.numeric()
  })
  
  
  # All IPCC carbon budgets related to the allocation date (incl. additional emissions between allocation date and reference date of IPCC budgets)
  carbon_budgets_at_allocation_date <- reactive({
    carbon_budgets %>%
      mutate(budget_gt = budget_gt + total_global_emis_between_allocdate_ipccbudgetdate() / 1000000000)
  })
  
  selected_carbon_budget_at_allocation_date <- reactive({
    # Budget refers to all emissions between allocation date and the year when emissions reach zero (calculated_zero_emissions_year())
    # Base year emissions do not count to this budget
    carbon_budgets_at_allocation_date() %>%
      filter(probability == selected_probability(),
             warming_degrees == selected_warming_degrees()) %>%
      select(budget_gt) %>%
      as.numeric()
  })
  
  ggplot_transparent_theme <- theme_test() +
    theme(
      panel.background = element_rect(fill = "transparent"), # bg of the panel
      panel.grid.major = element_blank(), # get rid of major grid
      panel.grid.minor = element_blank(), # get rid of minor grid
      panel.border = element_rect(colour = "transparent", fill = NA, size = 0),
      plot.background = element_rect(fill = "transparent", color = NA), # bg of the plot
      title = element_text(colour = "white"),
      axis.text = element_text(colour = "white"),
    )
  
  country_list <- unique(countries_emissions[["country"]])
  
  # VISUALIZATIONS OF STATUS QUO ####
  
    # Line chart: Historical temperatures ####
    
    linechart_temperatures <- prepareLineChartTemperatures(
      data = historical_temperatures, 
      theme = ggplot_transparent_theme
    ) %>%
      makeLabelArrows(x = 1965, xend = 1970, y = .3, yend = .1, label = "Temporary cooling, possibly\n man-made according to studies") %>%
      makeLabelArrows(x = 1985, xend = 1990, y = .6, yend = .4, label = "Rapid warming since the 90s") %>%
      makeLabelArrows(x = 1860, xend = 1865, y = .2, yend = .01, label = "This line marks the average \nbetween 1961 and 1990")
    
    output$linechart_temperatures <- renderGirafe(
      girafe(ggobj = linechart_temperatures, width_svg = 10, height_svg = 5) %>%
        girafe_options(opts_hover(css = "stroke:grey; stroke-width:2px; fill:none; fill-opacity:0") )
    )
    
    # Iframe: Sea level rise ####
    
    # output$iframe_sea_level <- renderUI({
    #   tags$iframe(src = "https://seeing.climatecentral.org/#12/40.7298/-74.0070?show=lockinAnimated&level=0&unit=feet&pois=hide", style = "height:70vh; width:70vw;")
    # })
    
    # Bar chart: Emissions per continent + year ####
    
    barchart_continents <- prepareBarChartContinents(data = continents_emissions, theme = ggplot_transparent_theme) %>%
      makeLabelArrows(x = 2006, xend = 2008.8, y = 36, yend = 31, label = "Financial crisis 2009 lets\nemissions decline temporarily") %>%
      makeLabelArrows(x = 1980, xend = 1982, y = 22, yend = 19, label = "Oil crisis") %>%
      makeLabelArrows(x = 1989, xend = 1991, y = 26, yend = 23, label = "Dissolution of the Soviet Union")
    
    output$barchart_continents <- renderGirafe({
      girafe(ggobj = barchart_continents, width_svg = 10, height_svg = 5)
    })
    
    
    # Line chart: Emissions per continent + year ####
    
    linechart_continents <- prepareLineChartContinents(continents_emissions, theme = ggplot_transparent_theme) %>%
      makeLabelArrows(x = 2002, xend = 2005, y = 12.8, yend = 10.8, label = "China's emissions start \n to rise drastically") %>%
      makeLabelArrows(x = 1989, xend = 1992, y = 9.2, yend = 7.4, label = "Europe's emissions decline after \n the dissolution of Soviet Union") %>%
      makeLabelArrows(x = 2006, xend = 2008.8, y = 9, yend = 6.8, label = "Financial crisis 2009 lets\nemissions decline temporarily")
    
    output$linechart_continents <- renderGirafe(
      girafe(ggobj = linechart_continents, width_svg = 10, height_svg = 5) %>%
        girafe_options(opts_hover(css = "stroke:grey; stroke-width:2px; fill:none;") )
    )
    
    
    # Chloropleth: Emissions per capita and country in the most recent (maximum) year ####
    
    world_map <- loadWorldMap(quality = "medium")
    
    world_map_emission_per_capita <- countries_emissions %>%
      mutate(emissions = emissions / 1000000) %>% # Convert to Million tons
      filter(year == maximum_year)
    
    world_map_emission_per_capita <- sp::merge(world_map, world_map_emission_per_capita, by.x = "admin", by.y = "country")
    
    palette_world_map_emission_per_capita <- colorNumeric(palette = c("white", "darkred", "slateblue3"),
                                                          domain = world_map_emission_per_capita@data$emissions_per_cap, na.color = "lightgrey")
    
    tooltip_world_map_emission_per_capita <- paste(
      world_map_emission_per_capita@data$admin, ": ",
      sprintf("%.1f", round(world_map_emission_per_capita@data$emis_per_capita, 1), nsmall = 0, big.mark = " ", scientific = F),
      " t CO2 per capita (",
      world_map_emission_per_capita@data$year, ")",
      sep = "")
    
    output$chloropleth_emissions_per_capita <- renderLeaflet(
      createChloroplethChart(
        data = world_map_emission_per_capita,
        data_column = "emis_per_capita",
        palette = palette_world_map_emission_per_capita,
        tooltip = tooltip_world_map_emission_per_capita,
        chart_title = "Emissions per <br>capita (t CO2)"
      )
    )
    
    # Rect chart: Emissions per capita and continent ####
    
    cols_emissions_per_cap_continent <- colorRampPalette(colors = brewer.pal(9, "Paired")[4:9])
    
    chart_emissions_per_cap_continents <- reactive({
      continents_emissions %>%
        filter(year == input$year_emissions_per_region) %>%
        mutate(population = population / 1000000000) %>%
        arrange(-emis_per_capita) %>%
        mutate(emis_per_capita = round(emis_per_capita, 1),
               population_perc = round(population / sum(population, na.rm = T) * 100, 1),
               emission_perc = round(emissions / sum(emissions, na.rm = T) * 100, 1),
               xmax = cumsum(population),
               xmin = lag(xmax, default = 0)) %>%
        mutate(pos_label = rowMeans(select(., c("xmin", "xmax")))) %>%
        createRectPlotEmissionsRegion(
          data = .,
          theme = ggplot_transparent_theme,
          cols = cols_emissions_per_cap_continent,
          year = input$year_emissions_per_region,
          pos_label = pos_label
        )
    })
    
    output$rect_emissions_per_cap <- renderGirafe(
      girafe(ggobj = chart_emissions_per_cap_continents(), width_svg = 10, height_svg = 5)
    )
    
    
    
    # Scatterplot: Emissions per GDP ####
    
    output$scatterplot_emissions_gdp_year <- renderUI({
      sliderInput("animate_emissions_gdp", label = "", min = 1971, max = 2018, step = 1, value = 1971, animate = T, sep = "")
    })
    
    scatterplot_emissions_gdp_palette <- colorRampPalette(brewer.pal(9, "Paired"))
    
    scatterplot_emissions_gdp_colors <- setNames(scatterplot_emissions_gdp_palette(length(country_list)), 
                                                 country_list) # To keep colors constant
    scatterplot_emissions_gdp_xaxis_limits <- summary(countries_emissions$gdp_per_capita[is.finite(countries_emissions$gdp_per_capita)])[c("Min.", "Max.")]
    scatterplot_emissions_gdp_yaxis_limits <- summary(countries_emissions$emis_per_capita[is.finite(countries_emissions$emis_per_capita)])[c("Min.", "Max.")]
    scatterplot_emissions_gdp_scale_size_limits <- summary(countries_emissions$emissions[is.finite(countries_emissions$emissions)])[c("Min.", "Max.")]
    
    scatterplot_emissions_gdp <- reactive({
      req(input$animate_emissions_gdp)
      countries_emissions %>%
        mutate(label = ifelse(country %in% input$selected_countries_gdp, country, "")) %>%
        filter(is.finite(gdp_per_capita),
               year == input$animate_emissions_gdp) %>%
        prepareScatterPlotGDP(data = ., 
                              theme = ggplot_transparent_theme,
                              limits_x = scatterplot_emissions_gdp_xaxis_limits,
                              limits_y = scatterplot_emissions_gdp_yaxis_limits,
                              limits_scale = scatterplot_emissions_gdp_scale_size_limits,
                              colors = scatterplot_emissions_gdp_colors,
                              year = input$animate_emissions_gdp)
    })
    
    output$scatterplot_emissions_gdp <- renderGirafe({
      girafe(ggobj = scatterplot_emissions_gdp(), width_svg = 10, height_svg = 5) %>%
        girafe_options(opts_toolbar(position = "top"))
    })
    
    
    # Point chart: Remaining budgets ####
    
    # Global emission budgets as defined in the IPCC SR1.5 report
    global_emissions_sr15_gt <- reactive({
      global_emissions %>%
        filter(year == 2018) %>%
        select(emissions) %>%
        mutate(emissions = emissions / 1000000000) %>%
        as.numeric()
    })
    
    scatterplot_carbon_budgets <- reactive({
      carbon_budgets %>%
        mutate(probability = as.character(probability),
               data_id = as.character(row_number()),
               years_left = budget_gt / global_emissions_sr15_gt()) %>%
        prepareScatterPlotBudgets(data = ., theme = ggplot_transparent_theme)
    })
    
    output$scatterplot_carbon_budgets <- renderGirafe({
      girafe(ggobj = scatterplot_carbon_budgets(), width_svg = 10, height_svg = 5)
    })
    
    # Justice Approaches compared ####
    
    # Heading of the main box
    output$justice_approaches_heading <- renderUI({
      req(input$selected_justice_approach)
      if(input$selected_justice_approach == "budget") {
        tags$div(tags$h3("Budget Approach"), style = "text-align:center;")
      } else if(input$selected_justice_approach == "convergence") {
        tags$div(tags$h3("Convergence and Contraction"), style = "text-align:center;")
      } else if(input$selected_justice_approach == "grandfathering") {
        tags$div(tags$h3("Grandfathering Approach"), style = "text-align:center;")
      } else if(input$selected_justice_approach == "other") {
        tags$div(tags$h3("Other approaches"), style = "text-align:center;")
      }
    })
    
    # Output: Text in the main box
    output$justice_approaches_text <- renderUI({
      tags$div(id = "justice_approaches_content",
               hidden(tags$div(id = "budget_content", # Gets visible by means of an observer when user selects this justice approach
                               tags$h4("Intuition", style = "font-variant:small-caps;"),
                               "Every human has an equal right to emissions, regardless of nationality.",
                               tags$h4("How it works", style = "font-variant:small-caps; margin-top:30px;"),
                               "The remaining budget is allocated based on equal per capita emissions.",
                               tags$h4("Consequences", style = "font-variant:small-caps; margin-top:30px;"),
                               "Industrialized states currently exceed their budget and accumulate large 'carbon debts'
                                                                                            due to their high per capita emissions. Developing states are allocated large emission
                                                                                            budgets. If historic emissions are taken into account, budgets of industrialized countries
                                                                                            often are already exceeded by now."
               )),
               hidden(tags$div(id = "convergence_content",
                               tags$h4("Intuition", style = "font-variant:small-caps;"),
                               "Every human has an equal right to emissions, but we should implement it gradually.",
                               tags$h4("How it works", style = "font-variant:small-caps; margin-top:30px;"),
                               "Initially, the remaining global budget is distributed based on the existing distribuion ('grandfathering'). Over the years, this is gradually replaced by a distribution based on equal per capita emissions.",
                               tags$h4("Consequences", style = "font-variant:small-caps; margin-top:30px;"),
                               "Industrialized countries have to reduce their emissions, but not as drastically as with the Budget Approach. Developing countries can moderately increase their emissions."
               )),
               hidden(tags$div(id = "grandfathering_content",
                               tags$h4("Intuition", style = "font-variant:small-caps;"),
                               "Budgets should be allocated based on pragmatic considerations. The current distribution of emissions is not totally unjust.",
                               tags$h4("How it works", style = "font-variant:small-caps; margin-top:30px;"),
                               "Each country's share on the global emissions is held constant.",
                               tags$h4("Consequences", style = "font-variant:small-caps; margin-top:30px;"),
                               "Economic development of developing countries could be hindered, as they have to hold their (low) share on global emissions constant. Developed countries are in a comfortable situation."
               )),
               hidden(tags$div(id = "other_content", # Gets visible by means of an observer when user selects this justice approach
                               tags$br(), "We might think of other approaches to allocate the carbon budget:",
                               tags$br(), tags$br(), 
                               tags$ul(tags$li(
                                 "We could distribute emission rights based on the ", tags$u("capabilities"), " of states to reduce emissions, and use a metric such as GDP to approximate this capability"
                               ),
                               tags$li(
                                 "We could have a closer look at the ", tags$u("individual situation"), " states are in: Are they located in geographical latitude where their energy consumption is
                                 necessarily higher? Did they just face a civil war and need more resources to recover?"
                                 ),
                               tags$li(
                                 "Instead of looking at where emissions accrue, we could look at ", tags$u("on whose behalf"), " they accrue - a large portion of China's emissions, for example,
                                 accrues for products which are sold in Europe or the US! In this so called ", tags$a(href = "https://ourworldindata.org/co2-and-other-greenhouse-gas-emissions#consumption-based-trade-adjusted-co2-emissions",
                                 "\"consumption-based view\""), ", even more emissions would be attributed to industrialized countries."
                               
                               )
                               ), tags$br(), tags$br()
               ))
      )
    })
    
    # Observers which trigger appearance of content based on user selection
    observeEvent(input$selected_justice_approach, {
      req(input$selected_justice_approach)
      if(input$selected_justice_approach == "budget") {
        shinyjs::hide("convergence_content")
        shinyjs::hide("grandfathering_content")
        shinyjs::hide("other_content")
        shinyjs::show("budget_content")
        shinyjs::show("exemplary_years_left")
        updateSelectInput(session, "selected_calculation_approach", selected = "budget")
      } else if(input$selected_justice_approach == "convergence") {
        shinyjs::hide("budget_content")
        shinyjs::hide("grandfathering_content")
        shinyjs::hide("other_content")
        shinyjs::show("convergence_content")
        shinyjs::show("exemplary_years_left")
        updateSelectInput(session, "selected_calculation_approach", selected = "convergence")
      } else if(input$selected_justice_approach == "grandfathering") {
        shinyjs::hide("convergence_content")
        shinyjs::hide("budget_content")
        shinyjs::hide("other_content")
        shinyjs::show("grandfathering_content")
        shinyjs::show("exemplary_years_left")
        updateSelectInput(session, "selected_calculation_approach", selected = "grandfathering")
      } else if(input$selected_justice_approach == "other") {
        shinyjs::hide("convergence_content")
        shinyjs::hide("budget_content")
        shinyjs::hide("grandfathering_content")
        shinyjs::hide("exemplary_years_left")
        shinyjs::show("other_content")
        # If "other" is selected, choose the Budget Approach as calculation method and set base year to 1990 (so that the next page "Years left" starts with this combination)
        updateSelectInput(session, "selected_calculation_approach", selected = "budget")
        updateSliderInput(session, "base_year", value = 1990)
      }
    })

    # When the user selects the inclusion of past emissions, set 1990 as base year
    observe({
      if(input$selection_past_emissions == T) {
        updateSliderInput(session, "alloc_date", value = 1990)
        shinyjs::hide("text_historical_emissions_no")
        shinyjs::show("text_historical_emissions_yes")
      } else {
        updateSliderInput(session, "alloc_date", value = maximum_year)
        shinyjs::hide("text_historical_emissions_yes")
        shinyjs::show("text_historical_emissions_no")
      }
    })
    
    # Create texts below the slider for historical emissions
    output$text_historical_emissions_yes <- renderUI({
      paste0("Allocation starts in ", allocation_date(), " ", ifelse(allocation_date() == 1990, "(this is the year where climate change became a problem widely known)", ""))
    })
    output$text_historical_emissions_no <- renderUI({
      paste0("Allocation starts in ", allocation_date())
    })
    
    
    # Bar chart: Approaches compared ####
    # Small bar chart with exemplary changes to states' budgets
    
    selected_justice_approach_text <- reactive({
      if(input$selected_justice_approach == "budget") {
        "Budget Approach"
      } else if(input$selected_justice_approach == "convergence") {
        "Convergence and Contraction"
      } else {
        "Grandfathering Approach"
      }
    })
    
    barchart_exemplary_years_left <- reactive({
      just_emission_budgets_countries() %>%
        filter(country %in% c("United States", "Mexico", "Botswana")) %>%
        left_join(countries_emissions) %>%
        mutate(emissions = emissions / 1000000000) %>%
        mutate(budget_reach = base_year() + floor(total_country_budget_gt / emissions)) %>%
        createBarChartExemplaryStates(
          data = .,
          theme = ggplot_transparent_theme,
          cols = cols_countries_years_left,
          selected_budget = selected_carbon_budget_sr15(),
          start_year = maximum_year,
          temperature_target = input$selected_warming_degrees
        )
    })
    
    output$exemplary_years_left <- renderGirafe({
      girafe(ggobj = barchart_exemplary_years_left())
    })
    
    
    
    # Chart: Future projections
    
    projection_data_2100 <- projection_data %>% filter(year == 2100) %>%
      gather(-year, key = "key", value = "value") %>%
      filter(!key %in% c("optimistic_policy", "d2_median", "d15_median", "historical")) %>%
      mutate(split = str_detect(key, "high")) %>%
      rowwise() %>%
      mutate(cat = str_sub(key, 0, str_locate(key, "_")[1]-1)) %>%
      ungroup() %>%
      select(-key) %>%
      group_by(split) %>%
      spread(split, value) %>%
      rename(ymin = "FALSE", ymax = "TRUE") %>%
      mutate(xmin = year+1, xmax = year+3) %>%
      mutate(cat = c("Baseline: 4.1 - 4.8°C", "Current policies: 2.8 - 3.2°C", "1.5°C consistent: 1.3°C", "2°C consistent: 1.7 - 1.7°C", "Pledges & Targets: 2.5 - 2.8°C"),
             desc = paste0(round(ymin, 0), " to ", round(ymax, 0), " Gt yearly emissions"))
    
    ribbon_chart_projections <- createRibbonChartScenarios(
      data_ribbon = projection_data,
      data_column = projection_data_2100,
      theme = ggplot_transparent_theme
    )
    
    output$ribbon_chart_projections <- renderGirafe({
      girafe(ggobj = ribbon_chart_projections, width_svg = 10, height_svg = 5)
    })
    
  # CALCULATION OF JUST BUDGET ####
  
  population_percentages_allyears <- countries_emissions %>%
    group_by(year) %>%
    mutate(population_percentage = population / sum(population, na.rm = T)) %>%
    ungroup() %>%
    select(country, year, population_percentage)
  
    # Calculate global emissions pathway for Contraction & Convergence ####
    # Assumption: Global pathway is a linear path to zero
    
    calculated_zero_emissions_year <- reactive({
      # The calculated zero emissions year is always an integer; in reality, it is very probable that emissions go to zero in between the year.
      # Due to this, the cumulated emissions of the global future linear pathway do not exactly meet the IPCC budget
      ceiling(
        # ceiling() rounds up to the next integer
        base_year() + 2 * selected_carbon_budget_at_allocation_date() / (global_emissions_base_year() / 1000000000)
      )
    })
    
    calculated_reduction_per_year <- reactive({
      (global_emissions_base_year() / 1000000000) / (calculated_zero_emissions_year() - base_year())
    })
    
    global_future_pathway_linear <- reactive({
      # Forecasted pathway from base year until the year when emissions reach zero
      data.frame(
        year = base_year() : calculated_zero_emissions_year(),
        cumulated_reductions_absolute = calculated_reduction_per_year(),
        global_emissions = global_emissions_base_year() / 1000000000
      ) %>%
        mutate(cumulated_reductions_absolute = ifelse(year == base_year(), 0, cumulated_reductions_absolute)) %>% # Base year: No reductions
        mutate(
          cumulated_reductions_absolute = cumsum(cumulated_reductions_absolute),
          global_emissions = global_emissions - cumulated_reductions_absolute,
          # Ct is zero for base year; base year emissions are however not part of the global carbon budget
          Ct = (year - base_year()) / (calculated_zero_emissions_year() - base_year()) 
        ) %>%
        select(-cumulated_reductions_absolute) %>%
        mutate(global_emissions = global_emissions * 1000000000)
    })
    
    # Per country: Assigned emission budgets (only for base year) ####
    
    just_emission_budgets_countries <- reactive({
      
      if(input$selected_calculation_approach == "budget") {
        countries_emissions %>%
          left_join(population_percentages_allyears) %>%
          filter(year == base_year()) %>%
          mutate(total_country_budget_gt = population_percentage * selected_carbon_budget_at_allocation_date()) %>%
          select(country, year, total_country_budget_gt, data_id)
        
      } else if(input$selected_calculation_approach == "grandfathering") {
        countries_emissions %>%
          filter(year == base_year()) %>%
          mutate(emission_share = emissions / global_emissions_base_year()) %>% 
          mutate(total_country_budget_gt = emission_share * selected_carbon_budget_at_allocation_date()) %>%
          select(country, year, total_country_budget_gt, data_id)
        
      } else if(input$selected_calculation_approach == "convergence") {
        # Very simple approach: Countries' population share is assumed to be constant between base and target year.
        # Linear version of Convergence & Contraction: Linear transition from allocation according to base year emissions to alloc. acc. to population share
        countries_emissions %>%
          left_join(population_percentages_allyears) %>%
          # Take base year emissions as starting point, which are then linearly reduced from the next year on
          filter(year == base_year()) %>%
          select(country, emissions, population_percentage, data_id) %>%
          rename(emission_forecast = emissions) %>%
          sp::merge(global_future_pathway_linear()) %>%
          group_by(country) %>%
          arrange(year) %>%
          mutate(emission_forecast = calculateContractionConvergence(emission_forecast, Ct, global_emissions, population_percentage)) %>%
          mutate(total_country_budget_gt = sum(emission_forecast[-1]) / 1000000000) %>% # Base year emissions (row 1) do not count to the remaining budget
          ungroup() %>%
          filter(year == base_year()) %>% # Choose one random year, just to remove duplicates
          select(country, year, total_country_budget_gt, data_id)
      }
    })
    
    # Per country: Budget still left per year ####
    
    just_emission_budgets_countries_left <- reactive({
      countries_emissions %>%
        filter(year >= base_year()) %>%
        left_join(just_emission_budgets_countries() %>% select(-year, -data_id)) %>% # Join with total budget per country for the base year
        left_join(cumulated_country_emis_since_allocation_date()) %>%
        mutate(budget_left = total_country_budget_gt - cumulated_emis_since_ad_gt,
               budget_left_perc = budget_left / total_country_budget_gt * 100) %>%
        select(country, year, total_country_budget_gt, budget_left, budget_left_perc, data_id)
    })
    
  # VISUALIZATIONS OF CALCULATIONS ####
  
    # Barchart: How many years are left per country? ####
    
    cols_countries_years_left <- colorRampPalette(colors = brewer.pal(9, "Paired"))
    
    barchart_countries_years_left <- reactive({
      req(selected_countries(), just_emission_budgets_countries(), total_country_emis_since_allocation_date())
      just_emission_budgets_countries() %>%
        filter(country %in% c(selected_countries())) %>%
        left_join(total_country_emis_since_allocation_date()) %>%
        left_join(countries_emissions) %>%
        mutate(emissions = emissions / 1000000000) %>%
        mutate(budget_reach = base_year() + floor(total_country_budget_gt / emissions)) %>%
        createBarChartYearsLeft(
          data = .,
          base_year = base_year(),
          theme = ggplot_transparent_theme,
          cols = cols_countries_years_left(length(selected_countries()))
        )
    })
    
    output$years_left <- renderGirafe({
      girafe(ggobj = barchart_countries_years_left(), width_svg = 10, height_svg = 4.5)
    })
    
    # Heatmap: How many of the just budget is left? (for all years between base year and maximum year) ####
    
    heatmap_just_budget_scales <- reactive({
      req(selected_countries(), just_emission_budgets_countries_left())
      just_emission_budgets_countries_left() %>%
        filter(country %in% selected_countries()) %>%
        select(budget_left_perc) %>%
        createChartScales()
    })
    
    heatmap_budget_left_allyears <- reactive({
      req(just_emission_budgets_countries_left(), selected_countries())
      just_emission_budgets_countries_left() %>%
        filter(country %in% selected_countries()) %>%
        createHeatmapEmissionBudgetLeft(data = .,
                                        scales = heatmap_just_budget_scales(),
                                        from = base_year(),
                                        to = maximum_year,
                                        theme = ggplot_transparent_theme)
    })
    
    output$heatmap_budget_left_allyears <- renderGirafe(
      girafe(ggobj = heatmap_budget_left_allyears(), width_svg = 10, height_svg = 4.5)
    )
    
    # Barchart: Budget left (percent), leaders & laggards (facets) ####
    
    barchart_leaders_laggards_scales <- reactive({
      if(sum(just_emission_budgets_countries_left()$budget_left_perc[just_emission_budgets_countries_left()$year == maximum_year] < 0, na.rm = T) > 0) {
        scale_fill_gradientn(colours = c("red", "white", "green"),
                             values = scales::rescale(c(-2000, 0, 100)))
      } else {
        scale_fill_gradientn(colours = c("tan1", "lightgreen", "darkgreen"),
                             values = scales::rescale(c(0, 10, 100)))
      }
    })
    
    barchart_leaders_laggards <- reactive({
      just_emission_budgets_countries_left() %>%
        filter(year == maximum_year) %>%
        group_by(budget_left_perc < 0) %>%
        top_n(10, abs(budget_left_perc)) %>%
        ungroup() %>%
        head(20) %>% # If all countries have the same value, this ensures that not all countries are plotted, only the first 20
        mutate(country = reorder(country, budget_left_perc)) %>%
        createBarChartLeadersLaggards(
          data = .,
          theme = ggplot_transparent_theme,
          scales = barchart_leaders_laggards_scales()
        )
    })
    
    output$barchart_leaders_laggards <- renderGirafe(
      girafe(ggobj = barchart_leaders_laggards(), width_svg = 10, height_svg = 5)
    )
    
    
  # OBSERVERS ####
  
    # Trigger "Sources" ####
    
    onevent("click", "sources_temperature", toggle("sources_temperature_text"))
    onevent("click", "sources_consequences", toggle("sources_consequences_text"))
    onevent("click", "sources_scenarios", toggle("sources_scenarios_text"))
    onevent("click", "sources_emissions_timeseries", toggle("sources_emissions_timeseries_text"))
    onevent("click", "sources_continent_emissions", toggle("sources_continent_emissions_text"))
    onevent("click", "sources_chloropleth", toggle("sources_chloropleth_text"))
    onevent("click", "sources_rect", toggle("sources_rect_text"))
    onevent("click", "sources_gdp", toggle("sources_gdp_text"))
    onevent("click", "sources_ipcc", toggle("sources_ipcc_text"))
    onevent("click", "sources_justice_approaches", toggle("sources_justice_approaches_text"))
    onevent("click", "sources_years_left", toggle("sources_years_left_text"))
    onevent("click", "sources_budget_left", toggle("sources_budget_left_text"))
    onevent("click", "sources_laggards", toggle("sources_laggards_text"))
    
    # Sync all panels ####
    # In Shiny, you only can display an inputId once, but we need to have the same inputs (basie data for budget calculation) on several tabs
    # Thus, they have different inputIds, and we need this observer to synchronize them
    
    # Selected countries
    observe({
      updateSelectizeInput(
        session, inputId = "selected_countries", selected = c(selected_countries_2())
      )
    })
    observe({
      updateSelectizeInput(
        session, inputId = "selected_countries_2", selected = c(selected_countries())
      )
    })
    
    # Base year
    observe({
      updateSliderInput(
        session, inputId = "alloc_date", value = c(allocation_date_2())
      )
    })
    observe({
      updateSliderInput(
        session, inputId = "alloc_date", value = c(allocation_date_3())
      )
    })
    observe({
      updateSliderInput(
        session, inputId = "alloc_date_2", value = c(allocation_date())
      )
    })
    observe({
      updateSliderInput(
        session, inputId = "alloc_date_3", value = c(allocation_date_2())
      )
    })
    observe({
      updateSliderInput(
        session, inputId = "alloc_date_3", value = c(allocation_date())
      )
    })
    
    # Selected probability
    observe({
      updateSelectInput(
        session, inputId = "selected_probability", selected = c(input$selected_probability_2)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_probability", selected = c(input$selected_probability_3)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_probability_2", selected = c(input$selected_probability)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_probability_3", selected = c(input$selected_probability_2)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_probability__3", selected = c(input$selected_probability)
      )
    })
    
    # Warming degrees
    observe({
      updateSelectInput(
        session, inputId = "selected_warming_degrees", selected = c(input$selected_warming_degrees_2)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_warming_degrees", selected = c(input$selected_warming_degrees_3)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_warming_degrees_2", selected = c(input$selected_warming_degrees)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_warming_degrees_3", selected = c(input$selected_warming_degrees_2)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_warming_degrees_3", selected = c(input$selected_warming_degrees)
      )
    })
    
    # Calculation approach
    observe({
      updateSelectInput(
        session, inputId = "selected_calculation_approach", selected = c(input$selected_calculation_approach_2)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_calculation_approach", selected = c(input$selected_calculation_approach_3)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_calculation_approach_2", selected = c(input$selected_calculation_approach)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_calculation_approach_3", selected = c(input$selected_calculation_approach_2)
      )
    })
    observe({
      updateSelectInput(
        session, inputId = "selected_calculation_approach_3", selected = c(input$selected_calculation_approach)
      )
    })
    
    # Jump to next / previous page ####
    observeEvent(input$forwardToPage2,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Temperatures")
    })
    observeEvent(input$forwardToPage3,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Emissions")
    })
    observeEvent(input$forwardToPage4,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Consequences")
    })
    observeEvent(input$forwardToPage5,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Scenarios")
    })
    observeEvent(input$forwardToPage6,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Emitters")
    })
    observeEvent(input$forwardToPage7,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Per capita")
    })
    observeEvent(input$forwardToPage8,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "History")
    })
    observeEvent(input$forwardToPage9,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Affluence")
    })
    observeEvent(input$forwardToPage10,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Carbon budgets")
    })
    observeEvent(input$forwardToPage11,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Climate Justice")
    })
    observeEvent(input$forwardToPage12,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Years left")
    })
    observeEvent(input$forwardToPage13,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Budget left")
    })
    observeEvent(input$forwardToPage14,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Leaders")
    })
    observeEvent(input$forwardToPage15,{
      updateTabsetPanel(session, inputId = "tabset-panel", selected = "Conclusion")
    })
    
    # Create choice values for gdp scatterplot country selection ####
    
    observe({
      updateSelectizeInput(session, "selected_countries_gdp",
                           choices = country_list,
                           selected = c("China", "Mozambique", "Sweden"),
                           server = T
      )
    })
    
    
    # Create choice values for budget left country selection ####
    
    observe({
      updateSelectizeInput(session, "selected_countries",
                           choices = country_list,
                           selected = c("Australia", "Brazil", "Canada", "China", "Chile", "Ethiopia", "France",
                                        "Germany", "India", "Japan", "Qatar", "Russia", "Spain", "United Kingdom", "United States", "United Arab Emirates"),
                           server = T
      )
    })
    
    observe({
      updateSelectizeInput(session, "selected_countries_2",
                           choices = country_list,
                           selected = c("Australia", "Brazil", "Canada", "China", "Chile", "Ethiopia", "France",
                                        "Germany", "India", "Japan", "Qatar", "Russia", "Spain", "United Kingdom", "United States", "United Arab Emirates"),
                           server = T
      )
    })
    
    observe({
      updateSelectizeInput(session, "selected_countries_3",
                           choices = country_list,
                           selected = c("Australia", "Brazil", "Canada", "China", "Chile", "Ethiopia", "France",
                                        "Germany", "India", "Japan", "Qatar", "Russia", "Spain", "United Kingdom", "United States", "United Arab Emirates"),
                           server = T
      )
    })
    
    # Images on start page ####
    
    value_startpage <- reactiveVal(0)
    
    # Create endless chain of numbers 1 to 3
    observe({
      invalidateLater(10000, session)
      if(isolate(value_startpage() < 3)) {
        isolate(value_startpage(value_startpage() + 1))
      } else {
        isolate(value_startpage(1))
      }
    })
    
    # Depending on value, change the picture displayed
    observe({
      if(value_startpage() == 1) {
        shinyjs::show("startpage_image_1")
        shinyjs::hide("startpage_image_2")
        shinyjs::hide("startpage_image_3")
      } else if(value_startpage() == 2) {
        shinyjs::hide("startpage_image_1")
        shinyjs::show("startpage_image_2")
        shinyjs::hide("startpage_image_3")
      } else {
        shinyjs::hide("startpage_image_1")
        shinyjs::hide("startpage_image_2")
        shinyjs::show("startpage_image_3")
      }
    })
    
    # Images on consequences page ####
    
    value_consequences <- reactiveVal(0)
    
    # Create endless chain of numbers 1 to 5
    observe({
      invalidateLater(7000, session)
      if(isolate(value_consequences() < 5)) {
        isolate(value_consequences(value_consequences() + 1))
      } else {
        isolate(value_consequences(1))
      }
    })
    
    observe({
      if(value_consequences() == 1) {
        shinyjs::show("consequences_1")
        shinyjs::hide("consequences_2")
        shinyjs::hide("consequences_3")
        shinyjs::hide("consequences_4")
        shinyjs::hide("consequences_5")
      } else if(value_consequences() == 2) {
        shinyjs::hide("consequences_1")
        shinyjs::show("consequences_2")
        shinyjs::hide("consequences_3")
        shinyjs::hide("consequences_4")
        shinyjs::hide("consequences_5")
      } else if(value_consequences() == 3) {
        shinyjs::hide("consequences_1")
        shinyjs::hide("consequences_2")
        shinyjs::show("consequences_3")
        shinyjs::hide("consequences_4")
        shinyjs::hide("consequences_5")
      } else if(value_consequences() == 4) {
        shinyjs::hide("consequences_1")
        shinyjs::hide("consequences_2")
        shinyjs::hide("consequences_3")
        shinyjs::show("consequences_4")
        shinyjs::hide("consequences_5")
      } else {
        shinyjs::hide("consequences_1")
        shinyjs::hide("consequences_2")
        shinyjs::hide("consequences_3")
        shinyjs::hide("consequences_4")
        shinyjs::show("consequences_5")
      }
    })
    
    # Mobile phone users popup ####
    
    observeEvent(input$hide_notification_message, {
      shinyjs::hide("notification")  
    })
    
    
}