a-reproducible-research-workflow

a-reproducible-research-workflow copied to clipboard

#+TITLE: A Reproducible Research Workflow

• Introduction

1. Run experiments, monitor process and save results
2. Collect results and display
3. Communicating results

• Run experiments and save results

• Check out demo scripts
  • Folder structure and each file
  • The main script with a CLI
  • [[https://wandb.ai/][W&B]] short intro
  • A master script for batch execution
  • Parallel execution
• It's a better practice to have off-line evaluation.
  • Easy to re-do evaluation
  • Time saving

• Collect results and display

• Check out =notebooks/report-exp_1.ipynb=
  • $\LaTeX$ tables made easy

• Generate an informal report

• [[https://en.wikipedia.org/wiki/Literate_programming][Literate programming]] via Org mode
  • A [[https://www.jstatsoft.org/article/view/v046i03][paper]] on how literate programming is good for reproducible research #+begin_src jupyter-python def fib(n): if n <= 2: return 1 else: return fib(n - 1) + fib(n - 2) #fib(4) #[fib(i) for i in range(5)] [["i", "fib(i)"]] + [None] + [[i, fib(i)] for i in range(5)] #+end_src
• Check out =docs/meeting-27-june-2021.org=
• Export to different formats
  • Formatting reveal.js #+begin_src #+OPTIONS: toc:nil reveal_width:1200 reveal_height:900 #+REVEAL_MIN_SCALE: 0.6 #+REVEAL_MAX_SCALE: 0.8 #+REVEAL_TRANS: slide #+REVEAL_THEME: simple #+REVEAL_PLUGINS: (markdown highlight search) #+REVEAL_HLEVEL: 2 #+REVEAL_HEAD_PREAMBLE: #+REVEAL_EXTRA_CSS: /Users/kai/org/revealjs.css #+end_src
• Limitations of Org mode or Emacs
  • Learning curve
  • Collaborative editing