Learning Goals

*At the end of this exercise, you will be able to: 1. Understand the basic building blocks of a shiny app. 2. Understand how shiny works with user inputs to build outputs.

Shiny

Shiny is an R package which makes it easy to design interactive web applications without knowing any html, css, or java script. First we need to make sure we have shiny installed and loaded.

We will also need the tidyverse and palmerpenguins to run the examples.

## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr     1.1.4     ✔ readr     2.1.5
## ✔ forcats   1.0.0     ✔ stringr   1.5.1
## ✔ ggplot2   3.5.1     ✔ tibble    3.2.1
## ✔ lubridate 1.9.4     ✔ tidyr     1.3.1
## ✔ purrr     1.0.4     
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

Shiny Example

Shiny apps can be designed to do almost anything you can imagine. They are very useful for making specialized tools for specific needs in business and research, or just for fun! Let’s run a basic example on our computer. Type the following into the R console. You should have the file plot_iris.R in your working directory. Use the drop down menus to change the x and y axes on the plot.

runApp("plot_iris.R")

Built-in Examples

There are many built-in examples in shiny. You can run them by typing the following into the R console. Examples: “01_hello”, “02_text”, “03_reactivity”, “04_mpg”, “05_sliders”, “06_tabsets”, “07_widgets”, “08_html”, “09_upload”, “10_download”, “11_timer”

#runExample("04_mpg")

Other Examples

Shiny apps can be very powerful and complex. Here are some examples of shiny apps that have been built by the life science community.

Shiny App Skeleton

A shiny app consists of a user interface (ui) function, server function, and run function.

  1. The ui function controls the user inputs and the way the app will be displayed; i.e. this controls how your app looks.
  2. The server function is the part of the app which takes the values of the user inputs, performs calculations and/or makes plots, and prepares the outputs for display; i.e. this controls how your app works.
  3. The run function combines the ui and server function to run the app.

Below is a bare bones shiny app. You can get this code snippet by typing shinyapp + shift + tab. It doesn’t do anything because we haven’t created any inputs with the ui or any outputs with the server.

library(shiny)

ui <- fluidPage(
  
)

server <- function(input, output, session) {
  
}

shinyApp(ui, server)

For the rest of the lab we will work on building an example app using palmerpenguins.

Shiny Inputs

There are many types of inputs in shiny, see the shiny cheat sheet. Our example app uses selectInput(), which takes a set of predefined choices and supplies them to the user in a drop down menu format. We are working with the palmerpenguins data set, so we want the selections to be the column names of the data frame.

Let’s get the names of the palmerpenguins data frame and add them as a selectInput() to our ui. selectInput() takes several arguments including the label that the server function will use, the label that will display to the user, the selection choices, and the initially selected choice. I always start by making a simple example plot so I know where I am going.

penguins %>% 
  ggplot(aes(x=bill_length_mm, y=body_mass_g))+
  geom_point(na.rm=T)

Since we will make a scatterplot, let’s focus on the continuous variables. The first step is to setup the x variable input.

ui <- fluidPage(
    selectInput("x", # we are naming the select input "x"
                "Select X Variable", # this is the label that will display to the user
                choices = c("bill_length_mm", "bill_depth_mm", "flipper_length_mm", "body_mass_g"), # these are the choices the user will have
                selected = "bill_length_mm") # this is the initially selected choice
)

server <- function(input, output) {
  
}

shinyApp(ui, server)

Let’s repeat the process for the y variable input.

ui <- fluidPage(
  
    selectInput("x", 
                "Select X Variable", 
                choices = c("bill_length_mm", "bill_depth_mm", "flipper_length_mm", "body_mass_g"), 
                selected = "bill_length_mm"),
    
    selectInput("y", 
                "Select Y Variable", 
                choices = c("bill_length_mm", "bill_depth_mm", "flipper_length_mm", "body_mass_g"), 
                selected = "body_mass_g")

    ) #closes the fluidPage

server <- function(input, output) {
  
}

shinyApp(ui, server)

Server Side

Great, we now have our user inputs set up! Next we need to work with those inputs so R can make the plot. This can be the tricky part. The processing is done in the server function. The ui passes the names of our choices via an object called inputs to the server function. We need to access our inputs by their names with the $ symbol. To get the x input we type input$x and y is input$y.

All reactive inputs need to be wrapped in a reactive environment called a reactive function. Reactivity is the core of shiny. Let’s wrap our reactive input values in a reactive function on the server side and run our app as a test. What happens? Are we doing anything with the inputs yet?

ui <- fluidPage(
  
    selectInput("x", 
                "Select X Variable", 
                choices = c("bill_length_mm", "bill_depth_mm", "flipper_length_mm", "body_mass_g"), 
                selected = "bill_length_mm"),
    
    selectInput("y", 
                "Select Y Variable", 
                choices = c("bill_length_mm", "bill_depth_mm", "flipper_length_mm", "body_mass_g"), 
                selected = "body_mass_g")
)

server <- function(input, output) {
  reactive({input$x
            input$y
            })
}

shinyApp(ui, server)

Shiny Outputs

So we have our inputs in a reactive environment, but we want to actually use those inputs to make a plot and display it on the ui. To make and display the plot, we need to save it to a named output object that the ui can use. To do this we use the reactive expression renderPlot() and access the plot on the ui side with plotOutput(). The inputs from selectInput() are character strings, so we need to use aes_string() in ggplot.

penguins %>% 
  ggplot(aes(x=bill_length_mm, y=body_mass_g, color=species))+
  geom_point(na.rm=T)+
  theme_light(base_size=14)

ui <- fluidPage(
  
    selectInput("x", 
                "Select X Variable", 
                choices = c("bill_length_mm", "bill_depth_mm", "flipper_length_mm", "body_mass_g"), 
                selected = "bill_length_mm"),
    
    selectInput("y", 
                "Select Y Variable", 
                choices = c("bill_length_mm", "bill_depth_mm", "flipper_length_mm", "body_mass_g"), 
                selected = "body_mass_g"),
    
    plotOutput("plot", width="500px", height="400px")
)

server <- function(input, output, session) {
  
  output$plot <- renderPlot({
    
    ggplot(data=penguins, 
           aes_string(x=input$x, y=input$y, color="species"))+
      geom_point(na.rm=T)+
      theme_light(base_size=14)
  })
  
}

shinyApp(ui, server)

Congratulations, you made your first shiny app!

Practice

  1. Let’s work with the homerange data again, just to keep things consistent. Please load the data as a new object homerange and have a look at the column names.
## Rows: 569 Columns: 24
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (16): taxon, common.name, class, order, family, genus, species, primarym...
## dbl  (8): mean.mass.g, log10.mass, mean.hra.m2, log10.hra, dimension, preyma...
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
homerange %>% 
  ggplot(aes(x=locomotion, fill=trophic.guild))+
  geom_bar(position="dodge")

homerange %>% 
  ggplot(aes(x=locomotion, fill=thermoregulation))+
  geom_bar(position="dodge")

  1. Build a shiny app that produces a barplot that explores locomotion type with a fill reaction for the variables trophic.guild and thermoregulation.
library(shiny)

ui <- fluidPage(
  
  radioButtons("x", 
               "Select Fill Variable", 
               choices=c("trophic.guild", "thermoregulation"),
               selected="trophic.guild"),
  
  plotOutput("plot")
)

server <- function(input, output, session) {
  
  output$plot <- renderPlot({
    
    ggplot(data=homerange, 
           aes_string(x="locomotion", fill=input$x))+
      geom_bar(position="dodge", alpha=0.8, color="black")+
      labs(x=NULL, y=NULL, fill="Fill Variable")
})
}

shinyApp(ui, server)

Packages for next time

Please install shinydashboard below for part 2!

#install.packages("shinydashboard")

That’s it, let’s take a break!

–>Home