This project was conducted in 2020, during the second significant wave of COVID-19 in Victoria, Australia. Its website is located at https://the2ndwave.amanjitgill.com.
-
Wrote Jupyter notebooks to scrape daily COVID-19 cases by LGA (local government area) and generate spatial heatmaps for Victoria and Melbourne.
-
Once the notebooks were working, I turned them into Python scripts and invoked them once a day, using a bash script. This script was itself automatically triggered using a cron job.
-
To the bash script, I added calls to
ffmpegto turn the generated spatial heatmaps into time-lapse videos showing the progression of COVID-19 through Victoria. -
Created an interactive Dash visualisation allowing users to enter the name of any Victorian LGA and see a time-series graph of its COVID-19 cases. This app also allows users to compare multiple LGAs on the same graph.
-
Initially published the Dash app on Heroku, but I'm now using DigitalOcean to serve it, because it offers fast load times (no 'sleep' mode) at a low cost.
-
Wrote a website in
html/cssand used this to publish both the time-lapse videos and the Dash app. This is published through GitHub Pages and available at the URL above.