Synopis

Key Features of the Group Project:

  • students propose topics of their choosing
    • working in small groups of two, three, four, five, …) is strongly encouraged and preferred
    • there is no perfect group size but three or four worked well for many teams
    • we allow single-person teams but recommend teams of three or four
  • all entries for the project are to be submitted and documented via GitHub
    • you should also use git (via your team’s GitHub repository) for your intermediate steps
    • i.e. do not just upload finished work but all steps
  • submission must ‘run’ i.e. code must do something sensible with data:
    • possible tasks for data are gathering, cleaning, aggregating, modeling, estimating, visualizing, summarizing …
    • and any combination thereof
  • provide a write-up in markdown
    • which should rendered nicely at GitHub, inclusion of figures or charts for visualisation is encouraged
    • you can choose html or pdf as the final format (note that pdf requires latex at your end)
  • you think of final deliverables as
    • the actual project plan and idea manifested in actual code
    • the actual project description in a write-up in markdown, i.e. a short paper
    • the actual project presentation recorded via Zoom or alike (more details forthcoming)
  • you may use multiple course components:
    • shell, sql, markdown, R, …
    • but your work should focus and center around R and different R packages
    • “anything that runs on ‘morrow’, our RStudio Server”, goes
    • please avoid technologies we cannot deploy there
    • i.e. GPU-heavy deep learners may be out; cpu-based ones may work if installable
    • so in a nutshell if you can install it from CRAN and deploy it on RStudio, you can use it