server.R

shinyServer(function(input, output) {
    
    # =========================================================================
    # Reactive resources
    # =========================================================================
    resource.trend <- reactive({
        df <- dataframes$trend %>%  # subset/filter df_base based on user selections
            filter(population == input$trend_population,
                   year >= input$trend_years[1],
                   year <= input$trend_years[2])
        return(df)
    })
    
    resource.occupation <- reactive({
        df <- dataframes$occupation %>%  # subset/filter df_base based on user selections
            filter(age_group %in% input$occupation_age_group,
                   race %in% input$occupation_race,
                   occupation %in% input$occupation_occupation)
        return(df)
    })
    
    resource.education <- reactive({
        if (input$education_category == "Gender") {
            df <- dataframes$education %>% 
                filter(category %in% c("Men", "Women"))
        }
        if (input$education_category == "Race") {
            df <- dataframes$education %>%
                filter(category %in% c("White", "Black", "Asian", "Hispanic"))
        }
        df <- df %>%  # subset/filter df_base based on user selections
            filter(metric == input$education_metric,
                   education %in% input$education_education)
        return(df)
    })
    
    
    
    # =========================================================================
    # Server outputs : Datatables
    # =========================================================================
    output$trend_datatable <- renderDataTable({
        return(resource.trend())
    }, options=list(pageLength=10, autoWidth=FALSE))
    
    output$occupation_datatable <- renderDataTable({
        return(resource.occupation())
    }, options=list(pageLength=10, autoWidth=FALSE))
    
    output$education_datatable <- renderDataTable({
        return(resource.education())
    }, options=list(pageLength=10, autoWidth=FALSE))
    
    
    
    # =========================================================================
    # Server outputs : Plots
    # =========================================================================
    output$trend_barchart <- renderPlot({
        # get data from dataframe
        df <- resource.trend()
        df_men <- df %>% filter(gender == "Men")
        df_women <- df %>% filter(gender == "Women")
        population <- input$trend_population
        
        # plotting
        plot <- ggplot(df, aes(x=year, y=value, group=gender, fill=gender)) +
            geom_bar(data=df_men, aes(y=value), stat="identity", color="black") + 
            geom_text(data=df_men, aes(y=value, label=value, color=gender), angle=90, size=3.5, fontface="italic", hjust=-0.25) + 
            geom_bar(data=df_women, aes(y=-value), stat="identity", color="black") + 
            geom_text(data=df_women, aes(y=-value, label=value, color=gender), angle=90, size=3.5, fontface="italic", hjust=1.25) + 
            geom_text(aes(y=0, label=year), angle=90, size=4, hjust=0.5, color="white") + 
            scale_x_continuous(breaks=seq(min(df$year), max(df$year), by=1)) + 
            scale_y_continuous(labels=abs, expand=c(0.4, 0.4)) + 
            scale_fill_manual(values=CATEGORYCOLORS) + 
            scale_color_manual(values=CATEGORYCOLORS) +
            labs(title=sprintf("%s (%s - %s)", input$trend_population, input$trend_years[1], input$trend_years[2]),
                 x="",
                 y=population) + 
            theme(panel.background=element_blank(),
                  axis.text.x=element_blank(),
                  axis.ticks=element_blank())
        return(plot)
    })
    
    
    output$occupation_facetplot <- renderPlot({
        # get data from dataframe
        df <- resource.occupation()
        df_men <- df %>% filter(gender == "Men")
        df_women <- df %>% filter(gender == "Women")
        percentage_flag <- input$occupation_percentage
        
        # plotting
        plot <- ggplot(df, aes(x=occupation, y=value, group=gender, color=gender, fill=gender)) +
            facet_grid(age_group ~ race) + 
            scale_y_continuous(labels=abs, expand=c(0.4, 0.4)) + 
            scale_fill_manual(values=CATEGORYCOLORS) + 
            scale_color_manual(values=CATEGORYCOLORS) +
            labs(title="Employment by Occupations (2013)",
                 x="Occupation",
                 y="Employment") + 
            theme(panel.background=element_blank(),
                  axis.text.x=element_blank(),
                  axis.ticks=element_blank())
        if (percentage_flag) {  # conditional geom_text label
            plot <- plot + 
                geom_text(data=df_men, aes(y=percentage, label=sprintf("%1.1f%%", percentage)), size=3, fontface="italic", hjust=-0.25) + 
                geom_text(data=df_women, aes(y=-percentage, label=sprintf("%1.1f%%", percentage)), size=3, fontface="italic", hjust=1.25) + 
                geom_bar(data=df_men, aes(y=percentage), stat="identity", color="black") + 
                geom_bar(data=df_women, aes(y=-percentage), stat="identity", color="black")
        } else {
            plot <- plot + 
                geom_text(data=df_men, aes(y=value, label=value), size=3, fontface="italic", hjust=-0.25) + 
                geom_text(data=df_women, aes(y=-value, label=value), size=3, fontface="italic", hjust=1.25) +
                geom_bar(data=df_men, aes(y=value), stat="identity", color="black") + 
                geom_bar(data=df_women, aes(y=-value), stat="identity", color="black")
        }
        plot <- plot + coord_flip()  # flip coordinates
        return(plot)
    })
    
    
    output$education_barchart <- renderPlot({
        # get data from dataframe
        df <- resource.education()
        category <- input$education_category
        metric <- input$education_metric
        
        # plotting
        plot <- ggplot(df, aes(x=education, y=value, group=category, color=category, fill=category, ymax=max(value)*1.1)) +
            geom_bar(stat="identity", color="black", position=position_dodge(width=0.6), width=0.5) + 
            geom_text(aes(label=value), size=3.5, fontface="italic", hjust=-0.5, position=position_dodge(width=0.6)) + 
            scale_y_continuous(labels=comma, expand=c(0.4, 0.4)) + 
            scale_color_manual(values=CATEGORYCOLORS) +
            scale_fill_manual(values=CATEGORYCOLORS) +
            labs(title=sprintf("%s (by %s)", metric, category),
                 x="Education Level",
                 y="") + 
            theme(panel.background=element_blank(),
                  axis.ticks.y = element_blank())
        plot <- plot + coord_flip()  # flip coordinates
        return(plot)
    })
    
    
    
    
    # =========================================================================
    # Server outputs : Downloads
    # =========================================================================
    output$trend_download <- downloadHandler(
        filename <- function() {
            return("employment_trend.csv")
        },
        content <- function(filename) {
            df <- dataframes$trend
            write.csv(df, file=filename, row.names=FALSE)
        }
    )
    
    output$occupation_download <- downloadHandler(
        filename <- function() {
            return("employment_occupation.csv")
        },
        content <- function(filename) {
            df <- dataframes$occupation
            write.csv(df, file=filename, row.names=FALSE)
        }
    )
    
    output$education_download <- downloadHandler(
        filename <- function() {
            return("employment_education.csv")
        },
        content <- function(filename) {
            df <- dataframes$education
            write.csv(df, file=filename, row.names=FALSE)
        }
    )
    
})