Shiny Fuel Consumption App

This presentation forms part of the final assignment of the Developing Data Products course offered by Johns Hopkins University at Coursera.org.

I have used the R dataset named mtcars to illustrate:

• a Shiny user interface that uses a slider to obtain car weight as input - see code on third slide.

• a Shiny server that predicts fuel consumption based on the selected car weight - see code on fourth slide.

The Shiny Fuel Consumption app has duly been deployed here:

• https://merciacarolinawentzel.shinyapps.io/ddp_week_4_assignment/

library(shiny)

shinyUI(
  fluidPage(
    headerPanel(
      "Shiny fuel consumption app"
    ),
    sidebarPanel(
      sliderInput(
        inputId = 'selected_wt', 
        label = 'Slide to a car weight (1000 pounds) to predict fuel consumption (miles per gallon)',
        value = 1.513, min = 1.513, max = 5.424, step = 0.001
      ),
      submitButton('Submit')
    ),
    mainPanel(
      h3("Current car weight selection"), verbatimTextOutput("selected_wt"), h5("1000 pounds"), h3(""),
      h3("Predicted fuel consumption"), verbatimTextOutput("predicted_mpg"), h5("miles per gallon"), h3(""),
      plotOutput('plot')
    )
  )
)

library(shiny)
library(datasets)

shinyServer(
  function(input, output) {

    output$selected_wt <- renderPrint({
      input$selected_wt
    })

    output$predicted_mpg <- renderPrint({
      predict_mpg(input$selected_wt)      
    })

    output$plot <- renderPlot({
      plot(
        mtcars$wt, mtcars$mpg, type="p", 
        main = "Car Weight and Fuel Consumption", 
        xlab = "Car Weight (1000 pounds)", 
        ylab = "Fuel Consumption (miles per gallon)") +
        abline(lm(formula = "mpg ~ wt", data = mtcars))
    })
  }
)

predict_mpg <- function(carweight_lbs) {
  fit <- lm(mpg ~ wt, data = mtcars)
  est <- predict(fit, data.frame(wt = carweight_lbs), interval = "predict")
  est[1,"fit"]
}

This presentation was created using RStudio Presenter.

Thank you