Reproducibility data analysis with R
Dates
9-12 October 2023
To foster international participation, this course will be held online
Course overview
With considerable effort, you wrote R code to analyze your data and generate a final document or report to present the results. You then give the code to a colleague for them to evaluate it, they
don’t know which packages they need to install, they don’t have the necessary data and they don’t know if the correct script is report_final.Rmd or report_final_final.Rmd. Even after your
guidance, they find that the code does not run on their machine. Worse, when you try to run it yourself a few weeks later, you realize that a package update "broke" your code and now it doesn't
run.
This course will help you avoid those issues. You will learn how to organize a project to speed up collaboration and maximize its reproducibility by leveraging existing tools in the R ecosystem
--such as RMarkdown, renv, and others--, version control and working environments.
Target audience and assumed background
This course is intended for researchers, data scientists, and anyone who uses R to generate documents and who wants to collaborate with other people (or themselves in the future) with the minimum
amount of pain possible.
Basic prior experience with R is recommended. If you have ever read data and generated a graph or table based on it, you have everything you need to participate.
Learning outcomes
By the end of this course, participants will be able to:
- Create an R project that outputs a reproducible document.
- Create and manage a reproducible environment that specifies packages and their versions.
- Track changes with git.
- Collaborate with others and themselves with GitHub.
- Create and publish containers.
Program
Daily schedule:
3-6 pm (Berlin time): live lectures, live coding, live exercises.
Asynchronous homework support via Slack.
- Introduction to reproducibility
- RStudio projects
-
- Folder structure
- R Package structure
- R markdown.
-
- Sintaxis.
- Using templates from rticles
- Using LaTeX templates
Tuesday– Classes from 3-6 PM Berlin time
- here package
- Git and GitHub
-
- Setup and basic ideas
- Basic workflow (add, commit)
- Collaboration (forks, pull requests)
- Repo documentation (README, Licence, code of conduct).
- Sharing data
-
- Data repositories (DOI and details)
- Access to data from code
- Automate project structure with rrtools
-
- Code documentation
- Managing dependencies with renv
- Introduction to Containers
- Docker
-
- Create a container with a Dockerfile
- Docker + renv
- Publish container on dockerhub
Instructors
Paola Corrales: Ph.D. student at the University of Buenos Aires. She studies atmospheric sciences applying data assimilation techniques to improve short-term forecasts of severe events in Argentina. Trainer and instructor for The Carpentries and an RStudio certified instructor. She also develops openly licensed materials to teach and learn R from scratch.
More information about Paola: https://paocorrales.github.io/.
Elio Campitelli: Ph.D. student at the
University of Buenos Aires in atmospheric sciences and an R package developer. They apply open science principles with a strong emphasis on reproducibility by publicly making all the code and
data available. They are a founding member of the R User Group in Buenos Aires and an outstanding contributor to the LatinR conference (as recognized by LatinR founding chairs). They maintain
several open-source R packages (e.g., ggnewscale; metR) and contribute to other packages, such as data.table and ggplot2.
More information about Elio: https://eliocamp.github.io
Cost overview
Should you have any further questions, please send an email to info@physalia-courses.org
Cancellation Policy:
> 30 days before the start date = 30% cancellation fee
< 30 days before the start date= No Refund.
Physalia-courses cannot be held responsible for any travel fees, accommodation or other expenses incurred to you as a result of the cancellation.