Lemmie help on this data please

[_Jersey Girl]

Lemmie you gave us a link to this website to watch the progress of CV-19. https://coronavirus.1point3acres.com/

It's one of the sources I've been using the entire time. When you look at the US and the first two graphs, you can see cumulative growth rate percentages stated on the first two graphs using a kind of shadow text. Same thing when you click on a state.

I noticed that the percentages are no long being posted for my own state while they are for most any other state, and they haven't been posted for several days now. I don't understand why that would happen. Can you explain, please?

Anyone else know?

Thanks!

[_Lemmie]

It may just be an artifact of the sizing. After clicking on your state, instead of looking at the two small graphs side by side at the top, scroll down to the larger graph under trends. On one graph it plots both your state and the u.s., your state’s growth rate %age in shadow text shows there as 3% today.

[_Jersey Girl]

It may just be an artifact of the sizing. After clicking on your state, instead of looking at the two small graphs side by side at the top, scroll down to the larger graph under trends. On one graph it plots both your state and the u.s., your state’s growth rate %age in shadow text shows there as 3% today.

Thanks! Yes, that's what I've been doing. On the two graphs in question, the percentages were different almost each day. I was interested in observing the lag between new cases and deaths. Do you think this is something that could return?

[_Lemmie]

It may just be an artifact of the sizing. After clicking on your state, instead of looking at the two small graphs side by side at the top, scroll down to the larger graph under trends. On one graph it plots both your state and the u.s., your state’s growth rate %age in shadow text shows there as 3% today.

Thanks! Yes, that's what I've been doing. On the two graphs in question, the percentages were different almost each day. I was interested in observing the lag between new cases and deaths. Do you think this is something that could return?

Not sure what you are referring to when you say the lag between new cases and deaths?

[_Jersey Girl]

Thanks! Yes, that's what I've been doing. On the two graphs in question, the percentages were different almost each day. I was interested in observing the lag between new cases and deaths. Do you think this is something that could return?

Not sure what you are referring to when you say the lag between new cases and deaths?

When the new cases shoot up...death spikes are soon to follow. There's a lag between the two.

Does that make sense?

[_Lemmie]

Not sure what you are referring to when you say the lag between new cases and deaths?

When the new cases shoot up...death spikes are soon to follow. There's a lag between the two.

Does that make sense?

Yes, but just eyeballing your state’s graph isn’t enough to establish that spikes in one will cause spikes in the other, beyond the obvious correlation that death follows sickness. The three overt spikes in deaths are preceded by 1) no spike before that day but a spike in new cases on same day, 2) a small increase but not spike day before, and 3) an increase but not spike two days before.

All of that just serves to emphasize that you can only casually get trends out of a visual of a graph, you have to dig into the math to see if there is an actual correlation, let along causal effect, specifically between single data points of new case local optima and single data points of death local optima.

Look at the chart at the very bottom, tracking the median time and the range of exposure, symptoms, and admission to acute care. The pattern you are seeing comes more out of that overall analysis of averages and ranges than of any correlation between specific single day points or spikes. Think of illness and death as rolling averages, not single day’s data.

ETA: it occurs to me I am being unnecessarily technical. So yes, in general, more cases, more deaths.

[_Res Ipsa]

Lemmie, is a seven-day rolling average an appropriate way to visualize the trend, or is something likes a Loess smooth better?

[_Jersey Girl]

Thanks for explaining, Lemmie. I've been obsessively recording these for over 2 months now.

[_Lemmie]

Thanks for explaining, Lemmie. I've been obsessively recording these for over 2 months now. :eek:

No, I understand. And that site really pulls stuff together very well and very reliably, right? And looking for relationships is normal. It helps me to put even some tentative human context to the data. What I am really comforted to see is the slow down out here, especially NY, New Jersey. It has been brutal.

[_Jersey Girl]

Thanks for explaining, Lemmie. I've been obsessively recording these for over 2 months now.

No, I understand. And that site really pulls stuff together very well and very reliably, right? And looking for relationships is normal. It helps me to put even some tentative human context to the data. What I am really comforted to see is the slow down out here, especially NY, New Jersey. It has been brutal.

It's been somewhat of a comforting ritual for me to record the national and local numbers each day. I follow two other websites as well. I've kept an eye on Hudson, Essex, Monmouth and Ocean Counties the entire time. Followed Cuomo's daily briefings as well.

I've found that when I record the numbers I can more easily put them in their place so to speak, and set them aside to get on with life such as it is. So, it's become part of self care routines for me.