Data Visualization Project
Groups constitution
Most of the students will be working in groups of 3. If the number of students is not a multiple of 3, we will also form groups of 4 (as few as possible).
Project data requirements
The dataset must be relevant in order to use the concepts and the different data visualization techniques we will learn during this course.
The teacher will validate the dataset
Grading criteria
- 5 points for the substance of the story you will tell
- 9 points for the usage of the different concepts and data visualization techniques
- 3 points for your proficiency with
matplotlib
- 3 points for your proficiency with
bokeh
Deliverable
- A bokeh server application (this is the recommended option)
- A notebook with the code and the visualizations
- A one page report of the project
Bokeh server applications seem to be the easiest to implement. Indeed, interactions with JupyterLab (or any browser based environment for execution or rendering) implies implementing by hand the logic of update when using widgets. Therefore, without this very low level logic implementation, bokeh looses some very interesting features.
Step by step guide
Step | Session | Milestone | Deliverable |
---|---|---|---|
1 | Session 1 | Group formation | Groups of 3-4 students formed |
2 | Session 2 | Dataset selection | Initial dataset chosen by group |
3 | Session 2 | Project planning | Project planning (ask teacher's opinion) |
4 | Session 3 | Data validation | Dataset approved by teacher |
5 | Session 3 | Story scripting | Preliminary story outline & key insights identified |
6 | Session 3-4 | First graphs | Initial visualizations with matplotlib |
7 | Session 4 | Advanced visualizations | Interactive graphs with bokeh |
8 | Session 5 | Final polish | Complete notebook + one-page report |
9 | Session 5 | Presentation | Project presentation to class |