Read on Twitter
1/ Back down to 11ln% increase in Alberta as the testing backlog clears. Looks like Quebec's good day yesterday was an outlier, and back to 18ln% growth today. That's better than 25%, though, and it looks like the curve might be sustainably, slowly bending.
2/ Two additional deaths in Alberta for 13 total. Implied mortality still hovering around 7%. There will be another bad day or two, and then I hope that the impact of physical distancing will start reducing the growth rate of hospital admissions and other bad outcomes.
3/ BC with four days in a row of sub 5% growth. Lotus land is the promised land for all of us... we will all get there eventually if we keep up with the work.
4/ Great, great thread here about what I call the fog of war, the tendency for data to lag and those lags to be uncertain (ie, stochastic). https://twitter.com/DFisman/status/1245689823949729792
Especially happy to see exponential percentages for time series. e^rt = easy percentage math. Much love for e.
Especially happy to see exponential percentages for time series. e^rt = easy percentage math. Much love for e.
5/ For my purposes, fog of war doesn't matter too much I think, because I rely on daily reported data and don't true up to the "correct" date of symptom reporting. This makes for a noisier time series, which is why I'm also sticking to cumulative counts instead of daily changes.
