There are additional TODO notes inside the jupyter notebook.
- Make the readers and plotting functions generic and able to process any kind of input data.
- Stack all plots together with a shared timeline, to ease comparison.
- Customize colors and legends in datashader plots: https://discourse.holoviz.org/t/how-to-customize-histogram-for-linked-large-time-series-curve-plots-with-full-code-example-for-newbies-like-me/1581/4 https://stackoverflow.com/questions/64886000/timeseries-with-multiple-colors-and-legend-using-datashader-and-holoviews?noredirect=1#comment114745140_64886000 https://stackoverflow.com/questions/5854515/interactive-large-plot-with-20-million-sample-points-and-gigabytes-of-data/55967461?noredirect=1#comment114646563_55967461
- Add some stats, such as rolling average/min/high points per 24h (bokeh supports this feature).