R Workshop

Module 6: “To Infinity and Beyond” (1)

2018-04-25
Bobae Kang
(Bobae.Kang@illinois.gov)

Source: R Markdown

“R Markdown is a file format for making dynamic documents with R. An R Markdown document is written in markdown (an easy-to-write plain text format) and contains chunks of embedded R code.”

- Garret Grolemund from R Studio

• Embedding R code chunks (and the outputs)
• Multiple document formats
• Reproducibility
• Automation

• Install rmarkdown
  • not necessary if using RStudio IDE
• Create a R Markdown document
  • File > New File > R Markdown
• “Knit” the R Markdown file into a document in the desired format

• YAML header
• Knit and preview outputs
• Markdown
• Code chucks

• Title, author, date
• output
  • output format
  • output options
  • different output format has different options

• Headers
• Font types (italic, bold, strikethrough, superscript)
• Lists (ordered, unordered)
• Hyperlinks
• Images
• Blackquotes
• Horizontal line/page break
• Math equations (using LaTeX)
• Code/R code (chunk, in-line)

# use the following to control global options
knitr::opts_chunk$set(eval = TRUE, echo = TRUE, ...)

• Global options
  • Act as the default settings for the whole document

_```{r eval = FALSE, echo = FALSE, ...}

• Local options (per code chunk)
  • Only applied to the specific code chunck
  • Overide the global options if relevant
• See here for full documentation of knitr chunk options

• Printing basic data frame object
• Proper (HTML) tables with knitr::kable()
• Interactive tables with DT::datatable()

Source: Wikimedia Commons

• R Markdown supports various output formats
• Output format can be specified using output in the YAML header
  • The same document can be rendered in multiple formats
• Different output format has different YAML render options
  • Refer the page 2 of R Markdown cheat sheet on RStudio's Cheat Sheets page for a full list

“An R Notebook is an R Markdown document with chunks that can be executed independently and interactively, with output visible immediately beneath the input.”
- “R Notebooks”, RStudio

• While all other R Markdown formats requires “knitting” to see the final output, an R notebook document is automatically re-rendered each time the source file (.Rmd) is saved
• Offers a more interactive workflow

• There are many R packages (90+) for taking full advantage of the interactivity that web can offer.
• With these packages, we can easily incorporate interactive widgets into HTML documents generated using R Markdown
• Examples of htmlwidgets include:
  • plotly and highcharter for interactive visualizations
  • leaflet for interactive maps
  • DT for interactive data tables
• Visit htmlwidgets for R website to find out more

“RPubs is a quick and easy way to disseminate data analysis and R code and do ad-hoc collaboration with peers.”
- RStudio Team

• Publishing an R Markdown output on RPubs is as easy as clicking the “publish” button on RStudio
• Comparable to Tableau Public's Gallery website(?)

Source: Wikimedia Commons

• RStudio supports many options for creating presentation slides that are both modern-looking and highly customizable
• Popular ways to create slides using R include:
  • R Markdown formats:
    • ioslides_presentation (HTML)
    • revealjs::revealjs_presentation (HTML)
    • slidy_presentation (HTML)
    • beamer_presentation (PDF)
  • R Presentation (HTML)

---
title: "My first ioslide presentation"
output: ioslides_presentation
---

• ioslides output format is built into RStudio
• Fully integrates R Markdown syntax
  • Creating new slides is as easy as using # and ## headings
• To learn more about ioslides, read the “Presentations with ioslides” article on RStudio

---
title: "My first revealjs presentation"
output: revealjs::revealjs_presentation
---

• With revealjs R package, it is possible to use R Markdown (and its syntax) to generate slides using reveal.js, a JavaScript library for interactive slides in HTML
  • Try this demo slides generated using reveal.js
• To learn more about revealjs, read the “Presentations with reveal.js” article on RStudio

• HTML presentations with slidy
  • Use output: slidy_presentation in the YAML header
  • See the “Presentations with Slidy” article on RStudio
• PDF presentations with beamer
  • Use output: beamer_presentation
  • See the “Presentations with Beamer” article on RStudio
• HTML presentations with xaringan
  • Use output: xaringan::moon_reader
  • Visit the package GitHub repository for more

g

• R Presentation is not a R Markdown document
• Similar to R Notebook, R Presentation offers an interactive “Preview” which gets automatically re-rendered each time the source file (.RPres) is saved
• R Presentation slides are built on reveal.js
• All Workshop slides are generated using R Presentation
• Unfortunately, it is no longer in active development
• To learn more about R Presentation, visit relevant pages on RStudio Support website

Source: R Studio

“Shiny is an open source R package that provides an elegant and powerful web framework for building web applications using R. Shiny helps you turn your analyses into interactive web applications without requiring HTML, CSS, or JavaScript knowledge.”
- RStudio.com

• Try my ICJIA Uniform Crime Report Data dashboard app here
• Check out more examples on RStudio's “Shiny User Showcase” page

• A data dashboard is a visual interface to data to allow its viewers for gain key insights
• It is often an interactive application that provides viewers with options to explore data from multiple angles interactively
• A Shiny application can make a great dashboard

• Shiny can be used to generate interactive documents with Shiny app widgets
• An interactive document is an R Markdown file with the YAML head including runtime: shiny and output: html_document or output: ioslids_presentation
• Embedded Shiny R code will render a mini application

install.pacakges("shiny")
library(shiny)

runExample("01_hello")

• Shiny is a package: install and import it
• Try a simple built-in example for Shiny application!

• A Shiny app consists of two parts, server and ui
• They can be separated into two files, server.R and ui.R
  • Both files must be located in the same directory
• Alternatively, both parts can live in a single file, app.R

function(input, output, session) {
  ## R code for server logic
}

• server.R contains the server-side logic of the application
• input object is an environment for storing user inputs modified by user's interaction with the application ui
• output object is an environment for storing elements (plots, tables, texts, etc) to be rendered and shown in the ui
• session object is an environment that can be used to access information relating to the session
  • session is optional

fluidPage(
  ## R code for user interface
)

• ui.R defines the user interface elements:
  • layouts, panels, inputs and outputs
• Available layouts include: the sidebar layout and grid layout
• Available panels include: the title panel, sidebar panel, main panel, tabset panel, navigation list panel, and more
• shiny offers many functions to take user inputs
• shiny offers many functions to render outputs (as defined in server.R)

library(shiny)

server <- function(input, output) {
  # server-side logic
}

ui <- fluidPage(
  # user interface
)

shinyApp(ui, server)

• app.R must include shinyApp(ui, server) at the end
• server is an object containing code for server-side logic
• ui is an object containing user interface elements

• Developing a Shiny app takes much practice, planning and pain experiment
  • However, the reward is YUGE!
• Start with these tutorials by RStudio
• Follow along these articles by RStudio
• Refer to Shiny Gallery for examples and inspirations

• Shiny applications, including the interactive documents, can be deployed for use in the following ways:
  • Via shinyapps.io
  • Via Shiny Server
  • Via ShinyProxy

• The easiest way to deploy/publish a Shiny application
  • Takes only a few clicks on the RStudio IDE
• Requires signing up to shinyapps.io
  • Can log in with Google or GitHub account
• See the pricing for hosting Shiny apps on the website
  • Free hosting is available but limited to 5 applications and 25 active hours per month

Source: Wikimedia Commons

• A collection R Markdown documents in HTML format can be made into a website
  • To do so, we need a _site.yml file in the same folder as individual documents to be put together
  • _site.yml specifies the name and the routing structure of the resulting website
  • With _site.yml and all HTML documents ready, run rstudio::render_site() to generate the website
• For details, see the “R Markdown Websites” article on RStudio website

“GitHub Pages is a static site hosting service designed to host your personal, organization, or project pages directly from a GitHub repository.”
-“What is GitHub Pages?”, GitHub Help

• GitHub is a free and commercial online repository service for storing and sharing projects using Git
• The ICJIA R Workshop website is hosted on GitHub Pages
• Read more about GitHub Pages and hosting static sites on GitHub Pages on GitHub Help here

• bookdown package, built on R Markdown, facilitates writing books and long articles/reports.
  • A bookdown publication is made downloadable in PDF, EPUB and MOBI formats
• Check out bookdown package website to find out more
• Also, read Xie, Y. (2018). bookdown: Authoring Books and Technical Documents with R Markdown for a comprehensive guide for bookdown
  • The book itself is generated using bookdown

• blogdown is a package to generate static websites using R Markdown and the Hugo, a popular open-source static website generator
• Visit Awesome Blogdown website for a curated list of blogdown examples
• Also, read Xie, Y. et al. (2018). blogdown: Creating Websites with R Markdown for a comprehensive guide for blogdown

Source: Giphy