Jonathan Spring

Cool! I took a look comparing not just to past calendar years but to the trajectory starting every month. The top matches start mid 1981. 3/4ths of them resulted in lower rates within the next 12mo. gist.github.com/jonspring/1b…

I used something like this for looking at aspect ratios of artworks in SFMOMA’s collection.

Here's an example using particles. Haven't fiddled much, different parameters might improve the aesthetics. gist.github.com/jonspring/b0…

Legion of Honor is also (sort of) from the expo.

#Tidytuesday first look at permit locations in SF, by proposed building type. Neat data set, lots to look at. gist.github.com/jonspring/49…

People who understand forex, why do we see this daily pattern? Ruble stronger (~115/USD) during UTC workday, weaker (~130/USD) otherwise? Related to recent easing of yuan controls? Or is this an artifact of data collection, e.g. based on unrestricted vs restricted forex markets?

Looks like hospitalization from Omicron continues to recede. Looks like only about 15% of COVID hospitalizations warranted the ICU, down from prior waves when more like 30-40% did.

Omicron has now sent more people to SF ICU's than Delta, and is catching up with last winter's wave.

Another COVID spiral riff, now with more even more objectionable chartjunk! #rstats #gganimate #seismogram

It was starting to look like Omicron wasn't sending people to the ICU in San Francisco (just "acute care" hospitalizations), but uptick in recent days. Compared to Delta peak, "acute" now higher and ICU usage >1/2 and still rising.

So far, Omicron has sent has sent some San Franciscans to acute care, but little/no ICU impact. Different than prior waves. Crossing my fingers this dynamic holds.

Crossing my fingers the SF ICU #s stay low. Omicron wave is charting new path, sending some San Franciscans to acute care but very few to ICU so far.

igraph FTW! I'm framing this one. Well done @grayheel! @AdventOfCode

Is there a variation of this that would work with `facet_wrap(~grp, ncol = 2, scales = "free_x")` and varied ranges? We'd want a per-facet or per-chart value for the nudge that depends on the range of the underlying data...

@Bob_Wachter, seems the next few days are critical for understanding the immediate mortality impact of the recent delta spike in the bay area. Is there reason to expect the linkage has improved recently?

Aha! Plenty of high hits didn't make it as far as the simplified ballistics would suggest. Looks like need to add a drag factor. Any suggestions?