cms-opendata-education/cms-rmaterial-multiple-languages

This repository contains exercises in R programming language using CMS Open Data.

CMS Open Data: Plotting an Invariant Mass Histogram in R

In this repository, you will find introductions to both the R programming language and to a simple analysis with R using CMS Open Data from the CERN Open Data portal. The tutorial is available in interactive form as a Jupyter notebook using the IR Kernel (in English only, for the moment) and as a PDF (in English and Spanish). You can run the Jupyter notebook in your browser without installing any software on your machine, through the power of Binder. However, installation instructions are available below if you wish to play with the notebook files offline. Note that the data will require at least a temporary connection to the CERN Open Data portal.

Tutorial files

Language	Jupyter notebook	For print
English (EN)		View PDF
Spanish (ES)	Coming soon!	View PDF

Instructions for offline use

1. Install Jupyter.
2. Install R.
3. Make the IR kernel available to Jupyter.
4. In the terminal, clone this repo onto your machine, and change directories so you're inside the repo.
5. Still in the terminal, type $ jupyter notebook. Your default browser will open with the Jupyter interface.
6. Click on the file ending with .ipynb to begin the interactive tutorial.

Credits

• Original text and code: Edith Villegas Garcia
• Additions and suggestions (the tidyverse example): Andrew John Lowe
• Editing and conversion to Jupyter notebook by Achintya Rao

Technical information

You can use Binder to interact with any such notebook via a browser without installing Jupyter or R, by hosting your notebooks on GitHub. To do so, make sure that your GitHub repository contains the Dockerfile shown here: https://github.com/binder-examples/dockerfile-r. Install any additional packages you want to interact with (e.g. tidyverse) by including them in the Dockerfile.

Licence

Coming soon!

cms-rmaterial-multiple-languages's Issues

Use Tidyverse instead of base R

Hi Achintya,

0. The zeroth reason for opening an issue is that I want to encourage reproducible research, and a notable benefit of making the code, paper, and/or data for analysis available on GitHub is that readers of a published work can address their questions to the author(s) directly by opening an issue rather than sending them an email. The consequence of this is that the knowledge exchange happens in the open and is coupled to the work in question, rather then a private email thread in someone's inbox. This means that others can benefit from the knowledge exchange. The FAQ writes itself!

1. I like that you gave a quick introduction to R, but with time I've become increasingly convinced that it is better start teaching the tidyverse before base R. In my own teaching materials I started with base R before introducing the tidyverse, but I now think that this approach is backwards. It's like learning C before learning C++, which is how I was taught C++ (it could be argued that this approach is actually counterproductive). The tidyverse enables students to do powerful things quickly, and there is the added "wow factor" that will encourage adoption. See here for more discussion on this: Teach the tidyverse to beginners

2. I've found knitr and rmarkdown incredible useful for producing elegantly formatted and cross-referenced documents that are immediately fit for journal submission. Can you provide any recipes for doing this from a Jupyter notebook? Is there anything as nice as knitr for Jupyter notebooks? Ideally, I want the Jupyter notebook to be the source of the published document; readers can then visit the Binder-powered repo to reproduce the work themselves. This is simpler than providing the Rmd or ipynb and instructing the reader to install a bunch of software, and doesn't ensure replication of the execution environment.