Workshop

1. Short intro to Markdown (this!)
2. A document you can work through to get started
3. Exercises along the way where we’ll highlight how to make useful documents (e.g. APA manuscript)
4. No need to stay until end if you feel you’ve got what you wanted

Learning goals

By the end of this workshop you should:

• Know what Markdown is and how the syntax works,
• See how to integrate code and data with RMarkdown,
• Understand the different output formats from RMarkdown and how to generate them
• Know about generating APA format files with papaja and bibtex.

RMarkdown = R + Markdown

1. It combines the magic of R with the simplicity of Markdown documents
2. The purpose: write your manuscripts and automatically insert results!
3. Use cases (amongst others):

• prevent rounding errors
• promote computational reproducibility
• encourage error correction by minimizing manual work

The Markdown

• Markdown is the simplest document language!
• Open standard, convertable into .tex, .docx, .html, .pdf, etc
• You can learn Markdown in five (!) minutes
• Despite being simple, very powerful and used on many web platforms

Basic Markdown: headings

• # Header 1 creates a heading of the first level
• ## Header 2 creates a heading of the second level
• ### Header 3 …
• etc

Basic Markdown: letter styles

• **bold** creates bold
• *italic* creates italic
• ***bold and italic!*** creates bold and italic!

The R

• If you’re here, you probably know R
• Programming language for statistics
• Packages to add functionality
• Everything open-source!

Combining R and Rmarkdown

• Can write text, include code
• We can include both code and its outputs, or just one of the two
• “Literate programming”: mix code and explanation so that it’s easier to figure out what’s going on.

Example code chunk

x <- rnorm(100, 50, 20)
y <- rnorm(143, 51, 20)

t.test(x, y)

## 
##  Welch Two Sample t-test
## 
## data:  x and y
## t = 0.067369, df = 204.01, p-value = 0.9464
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -5.266346  5.638966
## sample estimates:
## mean of x mean of y 
##  49.07422  48.88791

Example in line

1. However, we can also evaluate code in text, and replace text with output
2. For example, let’s round \(\pi\) to four decimal places, using the code round(pi, 4)

• 3.1416

Output formats

• HTML - easy to share results on the web
• PDF - for manuscripts/reports (requires installing latex)
• Word - for editing with non-R collaborators, journals

Templates

• Nicely formatted handouts / reports
• APA manuscripts using papaja (we’ll get into this in depth)
• Specialized styles like PNAS or Cognitive Science
• Don’t waste time reformatting - just change the template

RMarkdown presentations

• If you want to go overboard, you can render slides from code too
• This is not always perfect, but it’s great for course notes, especially teaching stats.
• This presentation also written in Rmarkdown!

Workshop

Work through the handout accompanying these slides, which you can find at http://bit.ly/rmarkdown-pdf and http://bit.ly/rmarkdown-html