Table 2:

Average growth rates and the number of counties in each cluster for K = 2. Model R corresponds to Algorithm 1 where we use the sample correlation matrix. Model A₁ corresponds to Algorithm 4 where we use the k-nearest neighbors graph (k = 7). Model A₂ corresponds to Algorithm 4 where we use the ϵ-neighborhood graph (ϵ = 0.007). Groups 1 and 2 are the obtained partitions ${\widehat{G}}_1$ and ${\widehat{G}}_2$, respectively. Growth Rate is the approximated exponential growth rate, calculated as in subsection 2.5. Presented are the averages of these growth rates and their associated SEs for the counties in two groups, clustered by different methods. R, second phase is for the clusters obtained for the period 05/10/2020 – 07/10/2020.

	Group 1			Group 2
Model	No. of Counties	Growth Rate	SE	No. of Counties	Growth Rate	SE
R	467	0.1589	0.0020	483	0.1704	0.0019
A1	462	0.1583	0.0020	488	0.1677	0.0019
A2	470	0.1605	0.0020	470	0.1664	0.0020
R, second phase	487	0.0207	0.0005	463	0.0233	0.0005