There are 2 basic ways the COVID-19 data plotter. can help see this.

• Use the Search feature to identify all the states meeting that criteria
• Look at each state individually and apply a Linear fit to the data

• Selected everything
• Manually filter out the main hot spot cities
• Then use the above 2 methods to see how the less infected (affected) parts of the states are doing

On the Map Data tab, select the US States - Confirmed dataset.

Other datasets could be used, but this one provides a few extra options

In the world map, zoom into an area near Georgia

Use the mouse wheel to zoom - drag to center

Click on the Georgia marker

This will display the associated data in the Individual Plots window
Notice that the line is increasing in almost a straight line

At the bottom of the Individual Plots window, click Display Daily Data

• There are still a lot of new cases, or
• There is just a lot of additional testing

What is really needed is - the number of hospital admissions - but we have to use the data we have.

At any rate, this is the basic configuration for the following examples.

• Adjust the map (zoom and position) to show the continental US
• Access the Filters tab in the Controls window.
• On Filters tab
  • Select Daily / Slope / is <=
  • Set the Less Than or Equals value to zero
  • Set the range to cover 2 weeks - for this example, use 79 to 93 (4/10/20 to 4/24/20)
  • Click Clear All - all the markers will turn red
  • Click Select - all the states meeting the criteria will turn green

If the range beginning is changed to 83 (see the next section), then Georgia is added to the list.

• Select a series and
• Specify the range - start and end points

In the Individual Plots window, right click the series name - Georgia

This opens the Chart Properties dialog

Drag the Ranges tab to create a new window

In row 2, set the min/max to 79 and 93

Position these 3 windows so that they can be seen at the same time

Individual Plots, Chart Properties, Ranges

In the Chart Properties dialog, click on the Series / Trends tab

In this, select Least squares and Range 2

Or does it?

It should be obvious that the result suffers from an end-point bias and, thus, may not be valid.

Move the mouse cursor over the row 2 minimum value (currently 80) in the Ranges window (don't bother clicking on it) and move the mouse wheel. You can see the trend line change as the end-point changes.

If the beginning date is 76 or 83, (18 days or 11 days) the slope is strongly negative. But for exactly 14 days, it is positive.

So, does Georgia meet the criteria or not?

Don't forget - we don't know if the number of new cases is increasing or if we are just testing more people.

Note: When the 04-25-20 data became available, Georgia passed the 14-day test.

For this we will use the US cities and counties dataset (option on the Map Data tab) and the Aggregate Plots chart.

Select the dataset and zoom the map such that the Georgia markers are centered and not overlapping

Just so that we can see them

Access the Filters tab in the Controls window

Select Limit actions to a single state and your state of interest

Click Select All

Every county in Georgia should have a green marker

On the Color Keys tab, select Total Count

This indicates the cumulative number of cases on the selected date
This date is controlled via the slider and other animation controls

In the Individual Plots window you can see the data for the counties you just removed. When in single mode, each time you click on a marker its data is added to this plot. The checkboxes can be used to hide individual series. The fact that they are so similar (the peaks and troughs line up) indicates, to me, that these are caused by the way the data is reported and not by some underlying way that people are getting sick. This is a very strong indication that end-point-bias should produce questionable results.

In the Aggregate Plots window, click New Plot

At the bottom of this window, click Display Daily Data

• Right click the series
• Open Ranges in a separate window and set the range
• Select Least Squares and Range 2 on the Trends tab

This is 04-24-20 data - it changes every day 11670 / 21128 = 55.2% are not in the 6 selected counties

So, at about 800 new cases a day and a fairly flat change over the selected time period, is Georgia ready to reopen?

Based on this - probably not, but hospital admissions would probably be a better indicator.

Deaths are the next best indicator since improved treatments will be automatically incorporated. The statewide results are similar.

Using the filters this time, only 9 Georgia counties have total deaths over 20. (One is over a hundred.)

• First, select all the Georgia sites
• Use the filter the deselect sites with a cumulative maximum greater than 20.
• Produce an Aggregate plot
• Do an analysis as before

Note: 50 counties have zero deaths, but only 2 have zero cases.

URL: http:// mc-computing.com / Science_Facts / COVID-19 / 14-day_Decline.html