An Introduction to knitr

• badge does not flip (a sign of intelligence of organizers!)
• live code!

• motivation (a real one “better homework” and a correct one “reproducible research”)
• features
• discussion of remaining problems
• see the lab session by Jeff Johnston on Friday for practical applications

• literate programming
  • designed for software development
  • heaven for reporting (Sweave, ingenious!)
  • boring for software (LaTeX package/doc)
• reproducible research
  • sounds like “more work to do”?
  • I do not want to learn anything new; let me copy and paste
• one day they will be doing reproducible research unconciously

• the problem of “when you type a backslash, God kills a kitten”
  • how many kittens were killed after I taught them LaTeX?
• the problem of “disappearing summer interns” (coined by Carlos Scheidegger)
• one of the best comments in source code

# When I wrote this, only God and I understood what I was doing
# Now, God only knows

• any computing language (not yet)
  • R, Python, Perl, C++, Shell scripts, awk, sed, CoffeeScript, …
  • even SAS
• any authoring language
  • LaTeX, HTML, Markdown, reSructuredText, AsciiDoc

• see 002-minimal.Rnw and 001-minimal.Rmd

• Documentation: https://yihui.org/knitr (video)
• Development: https://github.com/yihui/knitr
• Examples: https://github.com/yihui/knitr-examples
• Questions: http://stackoverflow.com/questions/tagged/knitr

• R + Markdown, R + LaTeX, R + HTML
• numeric output, graphics (also animations), syntax highlighting, code formatting, …

• if you do not change anything, just load the old results
• lazyLoad()

• use the engine option, e.g. engine = 'python', engine = 'awk'
• see knitr-examples/.*engine.*.Rmd (e.g. 023)

• current design: for each chunk, use a new process, e.g. system('python -c "print(1+1)"')
• Doug Bates: engine = Julia?
• how do I open a persistent sub-process in R, pipe code into it, and get the results back?
  • socket?
  • fifo?

• http://rpubs.com
• tons of reports (R Markdown + knitr + RStudio)

Run this locally:

# please use the latest deveopment version on Github
if (!require('knitr') || packageVersion('knitr') <= '1.3') install.packages('knitr', repos = c('http://rforge.net', 'http://cran.rstudio.org'), type = 'source')
demo('notebook', package = 'knitr')

• reproducible statistical graphics
• http://vis.supstat.com
• R Markdown + Jekyll + Github

• i.e. what you see here
• RStudio 0.98 (not released yet)

• Gentleman and Temple Lang (2004)
• one R package to rule them all!
• data, R, man, tests, demo, src, vignettes
• VignetteBuilder: knitr in DESCRIPTION
• \VignetteEngine{knitr::knitr} in vignettes
• see details