shiny_app.R

library(tidyr)
library(ggplot2)
library(dplyr)
library(shiny)
library(shinythemes)
library(readr)
library(plotly)

ui <- fluidPage(theme = shinytheme("united"),
                titlePanel("Evaluation Review Application"),
                sidebarLayout(
                  sidebarPanel(
                    selectInput("teacher", "Select Teacher:", choices = teachers, selected = "Teacher A"),
                    conditionalPanel(condition = "input.tabs==1",
                                     sliderInput("slider", "Binwidth:", min = 1, max = 20, value = 10, step = 5),
                                     checkboxInput("groupdatahist", "Group Data", FALSE)
                                     ),
                    conditionalPanel(condition = "input.tabs==2",
                                     checkboxInput("groupdatadrops", "Group Data", FALSE)
                                     ),
                    conditionalPanel(condition = "input.tabs==3",
                                     tags$p("The selected teacher is the green dot. Teacher Rating is calculated from the evaluation responses. \nIf every student marked 'Strongly Agree', the teacher would have a score of 5. \n If every student marked 'Strongly Disagree, the teacher would have a score of 1."))
                    ),
                  mainPanel(
                    fluidRow(
                      column(width = 9,
                             tabsetPanel(id = "tabs",
                               tabPanel("Grades", value = 1, plotlyOutput("grades_hist")),
                               tabPanel("Drop Rates", value = 2, dataTableOutput("droprates")),
                               tabPanel("Teacher Ratings", value = 3, plotOutput("dotplot")),
                               tabPanel("Which Activities Enhanced Your Learning?", value = 4, tableOutput("enhanced")),
                               tabPanel("If you Could Change One Thing?", value = 5, tableOutput("change"))
                               )
                             )
                      ),
                    fluidRow(h3("Course Evaluation Responses"),
                             column(width = 9, 
                                    plotOutput("evalplots")
                                    )
                             )
                    )
                  )
)

server <- function(input, output) {
  
  output$evalplots <- renderPlot({
    
      eval_response_data %>%
        group_by(Teacher, Question, Response) %>%
        tally() %>%
        filter(Teacher == input$teacher) %>%
        mutate(Response = factor(Response, 
                                 levels = c("Strongly Agree", "Somewhat Agree", "Neither Agree nor Disagree", "Somewhat Disagree", "Strongly Disagree"))) %>%
        ggplot(aes(Response, n)) +
        geom_bar(stat = "identity", fill = "lightgreen", color = "black") +
        facet_grid(~Question) +
        labs(x = "", y = "") +
        theme_minimal() +
        theme(axis.text.x = element_text(angle = 45, hjust = 1, size = 12))
  })
  
  output$droprates <- renderDataTable({
    
    if (input$groupdatadrops == TRUE) {
    
      all_data %>%
        group_by(Teacher) %>%
        filter(Option == "Premium", RegDate > "2015-07-01") %>%
        summarize(`Drop Rate` = round(mean(GRADE %in% drop_filters), 2),
                  Enrolls = n(),
                  Drops = sum(GRADE %in% drop_filters)) %>%
        arrange(desc(`Drop Rate`))
      
    } else {
      
      all_data %>%
        group_by(Teacher, Course) %>%
        filter(Option == "Premium", RegDate > "2015-07-01") %>%
        summarize(`Drop Rate` = round(mean(GRADE %in% drop_filters), 2),
                  Enrolls = n(),
                  Drops = sum(GRADE %in% drop_filters)) %>%
        arrange(desc(`Drop Rate`))
      
    }
  })
  
  output$grades_hist <- renderPlotly({
    
    if (input$groupdatahist == TRUE) {
      ggplotly(
      all_data %>%
        filter(Option == "Premium", 
               RegDate > "2015-07-01" & RegDate < "2015-11-01", 
               Teacher == input$teacher, 
               !(GRADE %in% drop_filters)) %>%
        ggplot(aes(as.numeric(GRADE))) +
        geom_histogram(binwidth = input$slider, fill = "pink", color = "black") +
        ggtitle("Grade Distribution") +
        labs(x = "Grade") +
        theme_minimal()
      )
      
    } else {
      all_data %>%
        filter(Option == "Premium", 
               RegDate > "2015-07-01" & RegDate < "2015-11-01", 
               Teacher == input$teacher, 
               !(GRADE %in% drop_filters)) %>%
        ggplot(aes(as.numeric(GRADE))) +
        geom_histogram(binwidth = input$slider, fill = "pink", color = "black") +
        ggtitle("Grade Distribution") +
        facet_grid(. ~ Course) +
        labs(x = "Grade") +
        theme_minimal()
    }
  })
  
  output$enhanced <- renderTable({
    eval_text_data  %>% 
      filter(Teacher == input$teacher) %>%
      select(2) %>%
      filter(!is.na(.[,1]))
  })
  
  output$change <- renderTable({
    eval_text_data  %>% 
      filter(Teacher == input$teacher) %>%
      select(3) %>%
      filter(!is.na(.[,1]))
  })
  
  output$dotplot <- renderPlot({

    eval_response_data %>%
      mutate(Response = factor(Response, 
                               levels = c("Strongly Disagree", "Somewhat Disagree", "Neither Agree nor Disagree", "Somewhat Agree", "Strongly Agree"))) %>%
      group_by(Teacher) %>%
      summarize(teacher_rating = sum(mean(as.numeric(Response)))) %>%
      mutate(fillcolor = ifelse(Teacher == input$teacher, "fill", "nofill")) %>%
      arrange(desc(teacher_rating)) %>%
      ggplot(aes(teacher_rating, fill = fillcolor)) +
      geom_dotplot(binwidth = 0.15, aes(size = 1)) +
      scale_y_continuous(NULL, breaks = NULL) +
      labs(x = "Teacher Rating") +
      scale_fill_manual(values = c("lightgreen", "black")) +
      theme_minimal() + 
      theme(legend.position="none")
  })
}

shinyApp(ui = ui, server = server)