COVID19 Figures

[HagenFritz]

More COVID-19 Figures

The main authors weren't huge fans of our original figures so we are going to fix them up and include new ones.

Line Plots

For these time series plots -- which include both mean and percent change -- let's only include historic data as far back as 2018. I know this limits our comparison, but Zoltan thinks that it is best to be consistent.

Heatmaps

I think heatmaps will be more enlightening than line plots. The simplest way will be to create heatmaps for each pollutant and each year. We can eventually create a 3x3 subplot that includes all data but this might get cumbersome.

• For each year (2018, 2019, and 2020) we create a heatmap for a single pollutant (ozone, pm2.5, nox)
• Rows correspond to weeks in the year and the columns are days of the week
• Let's start by looking at the average daily concentration for each pollutant as a single cell in the heatmap
• Make sure that the scale of the colobar is equal for all years of a particular pollutant. If you are using seaborn (which I suggest), you can specify this with the vmin and the vmax.
• Check out the heatmap from seaborn

[HagenFritz]

Nicely done! See reports dir for figures.