Updated 2/12 PM – RO on Trend jumped to 1.3 – Seems like more data coming out about infected rate.
Updated 2/11 PM – RO normalized, recovery # seem to be accelerating [see footer of site] but for sure we cannot trust the data coming out of China right now – Just the logistics alone of capturing that data – WHO has been authorized to be in country, so we might see more coming out now.
Updated 2/10 AM – Below RO 1.2 – Numbers out look like China is claiming minimal new case rate. Interesting to see where that number goes in the next couple of days — see “FOOTER” of site for latest numbers:
Updated 2/8 AM – Falling below RO 1.3 (on trend/data). Adding logscale chart as well, which looks like a tapering off. However the death rate is alarmingly consistent – indicating the Corona Virus is a mathematical genius! or just a good Fund manager, i.e. the data pattern is extremely questionable at this time — what we can expect is a 1.7-1.8% Death rate tomorrow.
1/30 – 213 – 11,900 – 1.79%
1/31 – 259 – 14,533 – 1.78%
2/1 – 304 – 17,483 – 1.74%
2/2 – 361 – 20,707 – 1.74%
2/3 – 429 – 23,731 – 1.81%
2/4 – 493 – 27,431 – 1.80%
2/5 – 564 – 31,481 – 1.79%
2/6 – 636 – 34957 – 1.82%
Updated 2/6 – Below RO1.3 – trending to 1.1-2
Updated: 2/4 – Solid on RO 1.3, refactored trend line as cases have exceeded 2k/daily (infection rate). Data is still questionable.
Updated: 2/3 – still trending on R0 .13
Updated: 2/2 – Calc error, puts RO at 1.3 on current data.
The graph/chart describes the tracking of the confirmed nCoV cases and mapping it to the nearest Replication Factor (R0). The data is coming in at various times and it is appropriate for us to recognize the logistical challenges of even getting accurate data from these locations.
In summary there was a clear path to a RO of 1.5+ as of 1/30, however that seems to have taken a course downward. The current RO is closer to 1.1-1.3. In order to achieve a 1.5 RO rate at this time, the confirmed cases would need to change by 8%. We will continue to track this more programmatically in time.
How I got to this data
1. Normalized the data up to 1/27 (too noisy) –
Rate of change was >60% d/d
Which by the way that is what appears others are starting with – IMO not reasonable: see below –Assumption made was ground data (initial) which was sporadic and not organized. If someone took the first 5-7 days 1/16-1/23, they would have been R0 2.5-3.0, and they were — 3.8 dropped to 2.6 post that week: https://twitter.com/DrEricDing/status/1221132573340061697
2. I calculated standard trend 1/26+ (utilizing (daily) the last 2 data points)
3. Took a stdDev from trend on data points, which revealed trend was stable enough (2 days)
Note: We are only 6 days into a reasonable trend/pattern.
4. Plotted the actual Confirmed cases against Trend, then against R0 estimations (confirmed*1.x)
5. Current data [today] says R0 1.3 is high side of trend and <2% variance of Actual Confirmed.
I am not forecasting R0, just mapping the data based upon what we know today (and I know these are low numbers). That said, even if the data was off 20% (likely) the R0 would read out at 1.5 actual. Not the 3.8, 2.6 estimations based upon initial data/chaos. Again, i can factor in a 20% error and stay within 1.5.
Opinion: The Actual Confirmed data would have to be 50% off to hit 2.0 — That’s not to say it won’t hit 2.x, but I can’t subscribe yet to a R0 of 2.x, as the following is occurring: Contact rate is slowing, Infection Period by 2/14 (if not sooner) should go down (assuming we have real numbers), causing stabilization/reduction of trend @ confirmed cases, additionally transportation (bans and closures) will assist in this reduction greatly.