backup.r

#backup 

#Project initial
library(shiny)
library(shinythemes)
library(readxl)
library(rgdal)
library(tidyverse)
library(stringr)
library(debkeepr)
library(leaflet)

#source to function file
source('functions.R')

latLongZoom.original <- data.frame("Area" = c("World", "Europe", "Africa",
                                              "Middle East", "Pacfic Islands", "Asia"),
                                   "Lat" = c(30, 49.8, -6, 27, 0, 32),
                                   "Long" = c(53, 15.47, 30, 72.5, 116, 115),
                                   "Magnify" = c(2, 4.25, 2.5, 4, 4, 3.25))

latLongZoom <- latLongZoom.original

#Read in the data
joined.data.original <- read_csv("joined.csv")
map.data.original <- readOGR("filteredCountries.GeoJSON")

#make copies of original data
joined.data <- joined.data.original
map.data <- map.data.original

#Creating the UI
ui <- fluidPage(theme = shinytheme("darkly"),
                titlePanel("Interactive Textile Explorer"),
                sidebarPanel(#All inputs will go in this sidebarPanel
                  radioButtons(inputId = "dataSet",
                               label = "Choose company of interest",
                               choices = c("WIC", "VOC", "Both"),
                               selected = "Both"),
                  radioButtons(inputId = "dataType",
                               label = "Choose data type of interest",
                               choices = c("Quantity", "Value"),
                               selected = "Quantity"),
                  radioButtons(inputId = "regionChoice",
                               label = "Select one",
                               choices = c("Origin", "Destination"),
                               selected = "Origin"),
                  selectizeInput(inputId = "zoomTo",
                                 label = "Zoom to:",
                                 choices = levels(factor(latLongZoom$Area)),
                                 selected = "World"),
                  selectizeInput(inputId = "textileName",
                                 label = "Choose textile(s) of interest",
                                 choices = levels(factor(joined.data$textile_name)),
                                 multiple = TRUE),
                  # uiOutput(outputId = 'filtered_inputs'),
                  selectizeInput(inputId = "colors",
                                 label = "Choose color(s) of interest",
                                 choices = levels(factor(joined.data$colorGroup)),
                                 multiple = TRUE),
                  selectizeInput(inputId = "patterns",
                                 label = "Choose pattern(s) of interest",
                                 choices = levels(factor(joined.data$textile_pattern_arch)),
                                 multiple = TRUE),
                  selectizeInput(inputId = "process",
                                 label = "Choose process(es) of interest",
                                 choices = levels(factor(joined.data$textile_process_arch)),
                                 multiple = TRUE),
                  selectizeInput(inputId = "fibers",
                                 label = "Choose fiber(s) of interest",
                                 choices = levels(factor(joined.data$textile_fiber_arch)),
                                 multiple = TRUE),
                  selectizeInput(inputId = "geography",
                                 label = "Choose geography of interest",
                                 choices = levels(factor(joined.data$textile_geography_arch)),
                                 multiple = TRUE),
                  selectizeInput(inputId = "qualities",
                                 label = "Choose quality(s) of interest",
                                 choices = levels(factor(joined.data$textile_quality_arch)),
                                 multiple = TRUE),
                  selectizeInput(inputId = "inferredQualities",
                                 label = "Choose inferred quality(s) of interest",
                                 choices = levels(factor(joined.data$textile_quality_inferred)),
                                 multiple = TRUE),
                  actionButton(inputId = "updateBtn",
                               label = "Click to update map!"),
                  actionButton(inputId = 'table_updateBtn',
                               label = 'Click to update table!'),
                  br(), br(), #The inputs for the pie chart and bar chart
                  selectInput(inputId = "pieChart",
                              label = "Choose a modifier for the pie chart:",
                              choices = c(colnames(joined.data[,c(19:26)])),
                              selected = "textile_name"),
                  checkboxInput(inputId = "omitNAs",
                                label = "Omit NAs in charts"),
                  selectInput(inputId = "barChart",
                              label = "Choose a modifier for the bar chart:",
                              choices = c(colnames(joined.data[,c(19:26)])),
                              selected = "textile_name"),
                  checkboxInput(inputId = "facet",
                                label = "Facet by modifier")
                ),
                mainPanel(
                  tabsetPanel(#All of the outputs go here (introduction, map/graphs, data tables)
                    tabPanel(title= "Introduction",
                             h2("Dutch Textile Trade from 1710 to 1715"),
                             h5("Interact with Dutch West India Company (WIC) and East India Company (VOC) textile shipments from 1710 to 1715, with data compiled by Kehoe and Anderson. The Map Explorer allows the user to choose a company and data type of interest, while filtering by textile modifiers, and displays an interactive world map with a complementary pie chart and bar chart when a specific country is selected. The Table Explorer displays the compiled and cleaned dataset.")),
                    tabPanel(title = "Map Explorer",
                             leafletOutput(outputId = "countriesMap"),
                             plotOutput(outputId = "pieChart"),
                             plotOutput(outputId = "barChart")
                    ),
                    tabPanel(title = "Table Explorer",
                             dataTableOutput('update_inputs'))
                  )
                )
)

server <- function(input, output, session) {
  
  #can we filter by origin or dest yr?
  #and maybe we can put bubbles on the graph at lat/long for orig and dest, highlight the orig or dest 
  #button when the other is clicked
  
  #Render the data table based on the given search
  #let's modify this to allow hiding of inputs as well
  
  
  output$update_inputs <- renderDataTable(searchDelay = 1000,{
    input$table_updateBtn
    isolate(filter_by_inputs(joined.data.original,isolate(input)))}) #filters the data for what has been searched
  
  
  # output$filtered_inputs <- renderUI({
  #   
  #   filtered <- filter_by_inputs(joined.data.original,input)
  #   
  #   mainPanel(selectizeInput(inputId = "colors",
  #                  label = "Choose color(s) of interest",
  #                  choices = levels(factor(filtered$colorGroup)),
  #                  multiple = TRUE),
  #   selectizeInput(inputId = "patterns",
  #                  label = "Choose pattern(s) of interest",
  #                  choices = levels(factor(filtered$textile_pattern_arch)),
  #                  multiple = TRUE),
  #   selectizeInput(inputId = "process",
  #                  label = "Choose process(es) of interest",
  #                  choices = levels(factor(filtered$textile_process_arch)),
  #                  multiple = TRUE),
  #   selectizeInput(inputId = "fibers",
  #                  label = "Choose fiber(s) of interest",
  #                  choices = levels(factor(filtered$textile_fiber_arch)),
  #                  multiple = TRUE),
  #   selectizeInput(inputId = "geography",
  #                  label = "Choose geography of interest",
  #                  choices = levels(factor(filtered$textile_geography_arch)),
  #                  multiple = TRUE),
  #   selectizeInput(inputId = "qualities",
  #                  label = "Choose quality(s) of interest",
  #                  choices = levels(factor(filtered$textile_quality_arch)),
  #                  multiple = TRUE))
  # 
  # 
  # })
  
  #The map of countries to be rendered
  output$countriesMap <- renderLeaflet({
    #We only want it to update when the updateBtn is pushed
    input$updateBtn
    
    #reading in all of the inputs, isolating them
    dataSet <- isolate(input$dataSet)
    dataType <- isolate(input$dataType)
    regionChoice <- isolate(input$regionChoice)
    textileName <- isolate(input$textileName)
    colors <- isolate(input$colors)
    patterns <- isolate(input$patterns)
    process <- isolate(input$process)
    fibers <- isolate(input$fibers)
    geography <- isolate(input$geography)
    qualities <- isolate(input$qualities)
    inferredQualities <- isolate(input$inferredQualities)
    area <- isolate(input$zoomTo)
    table_update <- isolate(input$table_updateBtn)
    
    #Every time, we want to start with all of the data to filter through
    joined.data <- joined.data.original
    
    #Use the function to filter the inputs
    joined.data <- filter_by_inputs(joined.data,input)
    ####WHY IS THIS NOT ISOLATED?####
    
    
    if(regionChoice == "Destination"){ #Decide which to filter by
      joined.data <- isolate(filter_by_inputs(joined.data,isolate(input))) #filter by the chosen inputs
      if(dataSet != "Both"){ #if they are interested in a specific company
        totalValues <- joined.data %>% #Total values to graph things later on and color the map
          filter(company == dataSet) %>% filter_totalValue()
        # group_by(dest_country) %>%
        # select(dest_country, textile_quantity, deb_dec) %>%
        # na.omit() %>% #If it is missing these columns, it will cause issues
        # summarise(total_Quant = sum(textile_quantity),
        #           total_Dec = sum(deb_dec))
      }
      else{ #This will use data from both companies
        totalValues <- filter_totalValue(joined.data)
        #joined.data %>%
        #   group_by(dest_country) %>%
        #   select(dest_country, textile_quantity, deb_dec) %>%
        #   na.omit() %>%
        #   summarise(total_Quant = sum(textile_quantity),
        #             total_Dec = sum(deb_dec))
      }
      
      map.data@data <- left_join(map.data.original@data, #Join with the map data, using the original map data each time
                                 totalValues,
                                 by = c("ADMIN" = "dest_country"))
    }
    else{ #Redo everything from before, except now using orig_country
      if(dataSet != "Both"){ 
        totalValues <- joined.data %>%
          filter(company == dataSet) %>%
          group_by(orig_country) %>%
          select(orig_country, textile_quantity, deb_dec) %>%
          na.omit() %>%
          summarise(total_Quant = sum(textile_quantity),
                    total_Dec = sum(deb_dec))
      }
      else{
        totalValues <- joined.data %>%
          group_by(orig_country) %>%
          select(orig_country, textile_quantity, deb_dec) %>%
          na.omit() %>%
          summarise(total_Quant = sum(textile_quantity),
                    total_Dec = sum(deb_dec))
      }
      
      map.data@data <- left_join(map.data.original@data,
                                 totalValues,
                                 by = c("ADMIN" = "orig_country"))
    }
    
    #This will be used to zoom to a specific region on the map
    latLongZoom <- latLongZoom.original %>%
      filter(Area == area)
    
    viewLat <- latLongZoom[,"Lat"]
    viewLong <- latLongZoom[,"Long"]
    viewZoom <- latLongZoom[,"Magnify"]
    
    # if(dataType == "Quantity"){ #This will show the total quantity of textiles
    #   bins <- totalValues$total_Quant %>%
    #     auto_bin() #Custom function for determining the bins we will use
    #   
    #   country.colors <- colorBin(palette = "YlOrRd",
    #                              domain = totalValues$total_Quant,
    #                              bins = bins)
    #   #Mapping the data
    #   map.data %>%
    #     leaflet() %>%
    #     addTiles() %>%
    #     addPolygons(fillColor = ~country.colors(total_Quant), #We only use polygons for countries we may have actually used
    #                 fillOpacity = .7,
    #                 color = "black",
    #                 opacity = 1,
    #                 weight = 1,
    #                 label = ~ADMIN,
    #                 popup = ~paste("Total Quantity:", format(ifelse(is.na(total_Quant), 0, total_Quant), big.mark = ",", scientific = FALSE), sep = " "),
    #                 layerId = ~ADMIN) %>%
    #     setView(lat = viewLat, lng = viewLong, zoom = viewZoom) %>%
    #     addLegend(pal = country.colors,
    #               values = map.data@data$ADMIN,
    #               title = "Quantities of Textiles Shipped")
    # }
    # 
    
    if(dataType == "Quantity"){ #This will show the total quantity of textiles
      create_leaflet_map(map.data,totalValues,dataType,c(viewLat,viewLong,viewZoom))
      
      # country.colors <- get_binByDataType(totalValues,dataType)
      # 
      # 
      # #Mapping the data
      # map.data %>%
      #   leaflet() %>%
      #   addTiles() %>%
      #   addPolygons(fillColor = ~country.colors(return_colByDataType(map.data@data,dataType)), #We only use polygons for countries we may have actually used
      #               fillOpacity = .7,
      #               color = "black",
      #               opacity = 1,
      #               weight = 1,
      #               label = ~ADMIN,
      #               popup = ~return_popupByDataType(map.data@data,dataType),
      #               layerId = ~ADMIN) %>%
      #   setView(lat = viewLat, lng = viewLong, zoom = viewZoom) %>%
      #   addLegend(pal = country.colors,
      #             values = map.data@data$ADMIN,
      #             title = return_titleByDataType(dataType))
    }
    
    
    
    else if(dataType == "Value"){ #Redo everything the same except for total value
      bins <- totalValues$total_Dec %>%
        auto_bin()
      
      country.colors <- colorBin(palette = "YlOrRd",
                                 domain = totalValues$total_Quant,
                                 bins = bins)
      
      map.data %>%
        leaflet() %>%
        addTiles() %>%
        addPolygons(fillColor = ~country.colors(total_Dec),
                    fillOpacity = .7,
                    color = "black",
                    opacity = 1,
                    weight = 1,
                    label = ~ADMIN,
                    layerId = ~ADMIN,
                    popup = ~paste("Total Value:", format(ifelse(is.na(total_Dec), 0, total_Dec), big.mark = ",", scientific = FALSE), "guilders", sep = " ")) %>%
        setView(lat = viewLat, lng = viewLong, zoom = viewZoom) %>%
        addLegend(pal = country.colors,
                  values = map.data@data$ADMIN,
                  title = "Value of Textiles Shipped")
    }
  })
  
  #Used to render the plot for pie chart
  output$pieChart <- renderPlot({
    input$updateBtn
    name <- input$countriesMap_shape_click$id
    
    #only want to do this if they clicked on a country
    if(length(name) != 0){
      #Read in all of the inputs, but isolated
      modifier <- isolate(input$pieChart)
      dataSet <- isolate(input$dataSet)
      regionChoice <- isolate(input$regionChoice)
      textileName <- isolate(input$textileName)
      colors <- isolate(input$colors)
      patterns <- isolate(input$patterns)
      process <- isolate(input$process)
      fibers <- isolate(input$fibers)
      geography <- isolate(input$geography)
      qualities <- isolate(input$qualities)
      inferredQualities <- isolate(input$inferredQualities)
      
      #Again, reusing the original data
      joined.data <- joined.data.original
      
      #Filter all the inputs
      if(length(textileName) != 0){
        joined.data <- joined.data %>%
          filter(textile_name %in% textileName)
      }
      if(length(colors) != 0){
        joined.data <- joined.data %>%
          filter(colorGroup %in% colors)
      }
      if(length(patterns) != 0){
        joined.data <- joined.data %>%
          filter(textile_pattern_arch %in% patterns)
      }
      if(length(process) != 0){
        joined.data <- joined.data %>%
          filter(textile_process_arch %in% process)
      }
      if(length(fibers) != 0){
        joined.data <- joined.data %>%
          filter(textile_fiber_arch %in% fibers)
      }
      if(length(geography) != 0){
        joined.data <- joined.data %>%
          filter(textile_geography_arch %in% geography)
      }
      if(length(qualities) != 0){
        joined.data <- joined.data %>%
          filter(textile_quality_arch %in% qualities)
      }
      if(length(inferredQualities) != 0){
        joined.data <- joined.data %>%
          filter(textile_quality_inferred %in% inferredQualities)
      }
      
      #We care specifically about the destination here
      if(regionChoice == "Destination"){ #Only dest_country
        pie.data <- joined.data %>%
          filter(dest_country == name) %>%
          select(textile_quantity,
                 deb_dec,
                 all_of(modifier),
                 company)
      }
      else { #Only orig_country
        pie.data <- joined.data %>%
          filter(orig_country == name) %>%
          select(textile_quantity,
                 deb_dec,
                 all_of(modifier),
                 company)
      }
      
      #Omit na of the selected columns to avoid errors
      if(input$omitNAs){
        pie.data <- pie.data %>%
          na.omit()
      }
      else{ #Fix a problem for if NA is the only data point
        pie.data[3][is.na(pie.data[3])] <- "None indicated"
      }
      
      if(dataSet != "Both"){ #Controlling for company selection
        pie.data <- pie.data %>%
          filter(company == dataSet)
      }
      
      if(isolate(input$dataType) == "Quantity"){ #If they're interested in quantity
        if(nrow(pie.data) != 0){ #check to see if there are values left to publish
          pie.data %>% 
            ggplot(aes(x="",
                       y = textile_quantity)) +
            geom_bar(stat="identity",
                     width=1,
                     aes_string(fill=modifier))+
            coord_polar("y", start=0) + #This line in particular changes the bar chart to a pie chart
            labs(x = NULL,
                 y = NULL,
                 fill = NULL) +
            scale_fill_discrete(name = paste(modifier)) +
            theme_void() +
            ggtitle(label = paste(modifier, "distribution for", name, "with these filters."))
        }
        else{ #No rows were found
          ggplot() +
            ggtitle(label = paste(name, " has no data for these filters and ", modifier, ".", sep = ""))
        }
      }
      else{ #This will do total value the same way, except graphing deb_dec
        if(nrow(pie.data) != 0){
          pie.data %>%
            ggplot(aes(x="",
                       y = deb_dec)) +
            geom_bar(stat="identity",
                     width=1,
                     aes_string(fill=modifier))+
            coord_polar("y", start=0) +
            labs(x = NULL,
                 y = NULL,
                 fill = NULL) +
            scale_fill_discrete(name = paste(modifier)) +
            theme_void() +
            ggtitle(label = paste(modifier, "monetary distribution for", name, "with these filters."))
        }
        else{
          ggplot() +
            ggtitle(label = paste(name, " has no data for these filters and ", modifier, ".", sep = ""))
        }
      }
    }
    else{ #This comes up if they have not clicked any countries
      ggplot() +
        ggtitle(label = "Select a country with data for these textiles in order to display a pie chart here.")
    }
  })
  
  #Rendering the bar chart - this works nearly the exact same way as the pie chart
  #except when it is graphing the outputs, it is doing so with a bar chart instead of a pie chart
  output$barChart <- renderPlot({
    input$updateBtn
    name <- input$countriesMap_shape_click$id
    
    if(length(name) != 0){
      modifier <- isolate(input$barChart)
      modifierObj <- paste("`", modifier, "`", sep = "")
      dataSet <- isolate(input$dataSet)
      regionChoice <- isolate(input$regionChoice)
      textileName <- isolate(input$textileName)
      colors <- isolate(input$colors)
      patterns <- isolate(input$patterns)
      process <- isolate(input$process)
      fibers <- isolate(input$fibers)
      geography <- isolate(input$geography)
      qualities <- isolate(input$qualities)
      inferredQualities <- isolate(input$inferredQualities)
      orig_yr <- isolate(input$orig_yr)
      
      joined.data <- joined.data.original
      
      if(length(textileName) != 0){
        joined.data <- joined.data %>%
          filter(textile_name %in% textileName)
      }
      if(length(colors) != 0){
        joined.data <- joined.data %>%
          filter(colorGroup %in% colors)
      }
      if(length(patterns) != 0){
        joined.data <- joined.data %>%
          filter(textile_pattern_arch %in% patterns)
      }
      if(length(process) != 0){
        joined.data <- joined.data %>%
          filter(textile_process_arch %in% process)
      }
      if(length(fibers) != 0){
        joined.data <- joined.data %>%
          filter(textile_fiber_arch %in% fibers)
      }
      if(length(geography) != 0){
        joined.data <- joined.data %>%
          filter(textile_geography_arch %in% geography)
      }
      if(length(qualities) != 0){
        joined.data <- joined.data %>%
          filter(textile_quality_arch %in% qualities)
      }
      if(length(inferredQualities) != 0){
        joined.data <- joined.data %>%
          filter(textile_quality_inferred %in% inferredQualities)
      }
      if(regionChoice == "Destination"){
        
        bar.data <- joined.data %>%
          
          filter(dest_country == name) %>%
          select(textile_quantity,
                 deb_dec,
                 orig_yr,
                 all_of(modifier),
                 company)
      }
      else{
        bar.data <- joined.data %>%
          filter(orig_country == name) %>%
          select(textile_quantity,
                 deb_dec,
                 orig_yr,
                 all_of(modifier),
                 company)
      }
      
      if(input$omitNAs){
        bar.data <- bar.data %>%
          na.omit()
      }
      else{
        bar.data[4][is.na(bar.data[4])] <- "None indicated"
      }
      
      if(dataSet != "Both"){
        bar.data <- bar.data %>%
          filter(company == dataSet)
      }
      
      if(isolate(input$dataType) == "Quantity"){
        if(nrow(bar.data) != 0){
          if(isolate(input$facet)){
            bar.data %>%
              ggplot(aes(x = orig_yr, y = textile_quantity)) +
              geom_bar(stat="identity",
                       aes_string(fill=modifier)) +
              labs(x = "Original Year",
                   y = "Textile Quantity",
                   fill = NULL) +
              scale_fill_discrete(name = paste(modifier)) +
              theme_bw() +
              ggtitle(label = paste(modifier, "distribution for", name, "with these filters.")) +
              facet_wrap(~get(modifier))
          }
          else{
            bar.data %>%
              ggplot(aes(x = factor(orig_yr), y = textile_quantity)) +
              geom_bar(stat="identity",
                       aes_string(fill=modifier)) +
              labs(x = "Original Year",
                   y = "Textile Quantity",
                   fill = NULL) +
              scale_fill_discrete(name = paste(modifier)) +
              theme_bw() +
              ggtitle(label = paste(modifier, "distribution for", name, "with these filters."))
          }}
        else{
          ggplot() +
            ggtitle(label = paste(name, " has no data for these filters and ", modifier, ".", sep = ""))
        }}
      else{
        if(nrow(bar.data) != 0){
          if(isolate(input$facet)){
            bar.data %>%
              ggplot(aes(x = factor(orig_yr), y = textile_quantity)) +
              geom_bar(stat="identity",
                       aes_string(fill=modifier)) +
              labs(x = "Original Year",
                   y = "Textile Quantity",
                   fill = NULL) +
              scale_fill_discrete(name = paste(modifier)) +
              theme_bw() +
              ggtitle(label = paste(modifier, "distribution for", name, "with these filters.")) +
              facet_wrap(~get(modifier))
          }
          else{
            bar.data %>%
              ggplot(aes(x = orig_yr, y = deb_dec)) +
              geom_bar(stat="identity",
                       aes_string(fill=modifier)) +
              labs(x = "Original Year",
                   y = "Textile Value (guilders)",
                   fill = NULL) +
              scale_fill_discrete(name = paste(modifier)) +
              theme_bw() +
              ggtitle(label = paste(modifier, "monetary distribution for", name, "with these filters."))
          }}
        else{
          ggplot() +
            ggtitle(label = paste(name, " has no data for these filters and ", modifier, ".", sep = ""))
        }}
    }
    else{
      ggplot() +
        ggtitle(label = "Select a country with data for these textiles in order to display a bar chart here.")
    }
    
  })
}

shinyApp(ui, server)