This webpage gathers information about my activities from the lectures given at the University of Grenoble to the Master students in Computer Science (MOSIG) from October 2024 to January 2025.
26/09/24 [ AL ] Presentation of the lecture | Processing data with the Tidyverse, beautiful viz with ggplot
- Indicate your name on the Pad. You will use to collaborate and fill in all the information you can.
- Register on the Mattermost through this invitation link. This is the preferred communication mode.
- Set up a public github or gitlab project for this lecture. You will take notes on this lecture and turn your homework and computational documents in this project.
- Register to the MOOC on Reproducible Research.
- Follow modules 1 + 2 of the MOOC with as much exercises as possible (except the last one of module2, on Challenger; watching interviews is optional)
- Set up a computational document system (e.g., Rstudio or Jupyter on your laptop or through the UGA JupyterHub).
- Report the URL of your git project, your mattermost ID on the Pad.
- Start learning R by reading this R crash course for computer scientists (Rmd sources).
03/10/24 [ JMV ] Introduction to the scientific method and computer science epistemology | Publications | Visualization and Exploratory Data Analysis
- Read Popper’s text and write a short summary in your GitHub repository
- [riticize figures of Jean-Marc’s slides by:
- Applying the checklist for good graphics;
- Proposing a better representation (hand-drawing is fine) that passes the checklist.
- Report this work for the figures on the github project.
- MOOC: Complete exercise 5 of module 2 (Challenger). Write a short text explaining what is good and wrong about this document (you may want to provide an updated version of the notebook) and upload on the github space.
- [] Hands-on, improve the experiment design and the analysis.
- Continue the hands-on by improving the experiment design and the analysis. Share your findings on the Pad and/or on your public fork of the project.
- Compute confidence intervals for the data in https://github.com/alegrand/M2R-ParallelQuicksort
- Fit a linear model for the data in https://github.com/alegrand/M2R-ParallelQuicksort
- Try to complete the peer-evaluation of the MOOC
- Play with the DoE Shiny Application (https://arnaud-legrand.shinyapps.io/design_of_experiments/?user_a7710).
- All eleven variables are in [0,1]. The goal is to find the combination of variables where the output is the higher. This may require to identify which variables are significant, guessing a model for the system, etc.
- The website will record the combinations you try and you should write a small report on how you proceed. You’ll find your login in front of your name in the pad and you should replace user_a7710 by this login.