Reproducible research: A demonstration
This is a repository which stores a presentation on reproducible research with a case study of Cantonese intonation. The materials were presented in Lab of Phonetics and Speech Science, Institute of Linguistics, CASS, on 7th May, 2019.
The file rr.md is the presentation file. The presentation was written in markdown, and the presentation was generated using pandoc. The folder figs contains two files, workflow.mmd and fda.mmd, which generates flow charts of the workflows of phonetic research and FDA, respectively. The two flow charts were written in mermaid. The scrips folder contains three data preparation scripts written in Python.
The presentation file rr.md can be compiled using Pandoc. Type the following commands in the terminal (or cmd/PowerShell in Windows):
git clone https://github.com/ge-chunyu/RR-Cantonese
cd RR-Cantonese
pandoc -t beamer --pdf-engine=xelatex rr.md -o rr.pdf You will receive the following warning:
[WARNING] Could not fetch resource './workflow.png': replacing image with description
[WARNING] Could not fetch resource './fda.png': replacing image with description
[WARNING] Could not fetch resource './final_q.png': replacing image with description
[WARNING] Could not fetch resource './PC1.png': replacing image with description
[WARNING] Could not fetch resource './PC2.png': replacing image with description
[WARNING] Could not fetch resource './pc1-pc2.png': replacing image with descriptionThe two files included in the figs folder can be used to generate the first two figures. You may try to generate these figures by yourself using R with the DiagrammeR package or Mermaid. You can download all the figures on my Dropbox. Put the figures in the RR-Cantonese folder and rerun the code will get the same presentation file as I did. Note that LaTeX is also required to generate beamer files.
The analyses of Cantonese intonation using FDA relies heavily on the work by Michele Gubian. Thus, the codes for data analyses are not uploaded to the repo.