An introduction to data analysis, using R. Experimental.
R CSS
Switch branches/tags
Nothing to show
Clone or download
Pull request Compare This branch is 6 commits behind briatte:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
R-2013-Lyon
code
data
images
.Rprofile
001_intro.Rmd
002_readings.Rmd
010_setup.Rmd
011_r.Rmd
012_rstudio.Rmd
013_practice.Rmd
020_objects.Rmd
021_vectors.Rmd
022_variables.Rmd
023_practice.Rmd
030_functions.Rmd
031_control.Rmd
032_iteration.Rmd
033_practice.Rmd
040_data.Rmd
041_dataio.Rmd
042_reshaping.Rmd
043_practice.Rmd
080_models.Rmd
081_correlation.Rmd
082_ols.Rmd
083_practice.Rmd
090_ts.Rmd
091_lags.Rmd
092_smoothing.Rmd
093_practice.Rmd
100_maps.Rmd
101_geocoding.Rmd
102_choropleths.Rmd
110_networks.Rmd
111_influence.Rmd
112_networkds.Rmd
120_data.Rmd
121_sources.Rmd
122_figures.Rmd
README.md
index.Rmd
style.css
syllabus.pdf

README.md

README

Ivaylo Petev and myself use this repository to teach an undergraduate introduction to data analysis. The course is online.

If you are reading the course on its online pages, just replace the .html extension of a page by .R to download the underlying code.

HOWTO

The course pages are formatted in R Markdown syntax and were converted to HTML with knitr 1.4:

install.packages("knitr")
citation("knitr")

The knitting routine is in the .Rprofile. To compile the whole course, set the IDA folder as your working directory and then type ida.build() (takes a bit more than five minutes on optic fiber).

Other files are called from the code/ and data/ folders. Most datasets are downloaded on the fly if they are missing from the data/ folder, so make sure that you are online while running the scripts.

The whole course was coded and taught with RStudio. The code was ran on R 2.15.2, 2.15.3, 3.0.0 and 3.0.1, on a MacBook Air running OS X 10.8 and Mac OS X 10.9. Most plots use ggplot2 version 0.9.3.1 (just in case compatibility breaks at some point).

CREDITS

Thanks to the Sciences Po Reims staff, who offered invaluable support, and to the small group of students who enrolled in (and survived to) the course. The R-2013-Lyon slides have a bit more detail on the practicals.

Bits and pieces of the code were posted to Gist, RPubs and Stack Overflow during development. Thanks to the great R developer and user communities that live online, and which we are now proud to count ourselves in.

If you share the spirit of all this, you should consider joining the Foundation for Open Access Statistics and check out places like OpenCPU, the Open Knowledge Foundation and other initiatives in open access, open data, open source and open science.

HISTORY

Aug 2013: better data management, with large or multiple-file datasets read from ZIP archives. Switched datasets to .csv thanks to GitHub.

Jul 2013: typos and broken links. Removed some functions in .Rprofile that are now part of the questionr package.

Jun-2013: first draft. Everything kind of works, Sessions 5--7 are unlisted, the code/ folder contains a few more exercises. That's it for now!

May-2013: added more course content and better resolution (100dpi) for all plots.

Apr-2013: added a lot of course content and cleaner plots. Also adding the R-2013-Lyon folder for a conference presentation on the course.

Mar-2013: reviewed course structure: less files, more code, tons of new examples and exercises.

Feb-2013: more efficient .Rprofile functions and improved knitr routine, tidier code on the early sessions.

Jan-2013: first release.

First release: January 2013.
Last revised: August 2013.