This course is automatically staged and deployed using github actions.
- Staging site: updated when a change is made on the master or staging branches.
- Production site: updated when a release is made.
- Commit your changes to master or staging.
- Once the github action runs, look for the entry called
🐣: Check content diff
, showing which files changed after content rebuild. - For changed files, look on http://staging-learn.siuba.org/ to see that the changes make sense.
- Create a release, which will trigger a rebuild and deployment!
⚠️ : Every time you change the master branch, when someone clicks "run" binder will need to rebuild! This process can take over a minute. If you make changes on staging, binder will not need to rebuild after every change.
In order to build this course locally, you can run the following (for MacOSX).
# install this font, which supports many asian characters
# you have to clear matplotlib's cache, or it won't see the new font.
brew install font-noto-sans-cjk-jp
rm ~/.matplotlib/fontlist*
# install requirements
pip install -r tutorial/requirements-dev.txt
# regenerate everything
make notebooks
# note that if you want to force everything to regenerate, use this flag
make notebook -B
- Formatting cleanup
- Exercise headings
- Clearly marking the slides
- Some typos / markdown getting mangled
- Few places where details element is used instead of a question box
- Better slide transitions when user clicks down (i.e. appears immediately)
- Run some people through chapters 2 and 3
- Intro page? (could punt to down the road)
- A couple of screencasts showing 10 minute analyses, using skills from course (could also use verbs like count(), and refer to docs).
- Suggested Tidy Tuesdays to try
- A "next steps" outro lesson (could put at end of chapter 3 for now)
- Figure out and add license
- Remove /build/ from lesson urls
- Record video for slides
- Clicking up key terms brings up definitions
- A big cheatsheet showing everything taught
- Chapter 4
- first 2 lessons on plots, then 1 on bridge material (e.g. simple pandas read_csv)?