The top graph shows the growth in total number of (positive) cases in New York State between 03/23/2020 and 04/08/2020. The slope was clearly exponential until 04/05/2020 (14th day), when the curve started to tilt forward, indicating flattening of the steep growth. The three data points on the outer end of the curve, i.e., 15th, 16th and 17th day, are not only below the trendline, but are also diverging with a widening angle. Obviously, the aforesaid three diverging data points have lowered the r-squared below the perfect level.
The second graph shows the number of deaths in NYS during the same period. The trendline remains exponential with a slightly steeper slope, moving the r-squared higher than the prior one. Nonetheless, the three outer most data points remain below the exponential trendline, thus closely aligning with the total cases. Of course, since the slope is slightly steeper here, the resulting r-squared is higher than the prior case. Again, the scatter remains almost identical.
The third graph (bottom) shows the regression between the two data sets, keeping the period constant. As expected, the scatter shows a similar exponential slope with an r-squared that remains in close proximity. Considering this divergence has been continuing for three consecutive days -- not just one-off -- it is pointing to an emerging trend.
Now, the devil is frightened. By the time we re-run this study next week with newer data points, the devil will be totally cornered.
Data Source: https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_New_York_(state)
Stay safe.
-Sid Som
homequant@gmail.com
No comments:
Post a Comment