raid
Dad Photographer
This may be really off=topic, but then again, we are limited in our photography adventures by the spread of COVID-19.
https://www.linkedin.com/posts/raid...permutation-activity-6721489199700156416-VsUv
https://www.linkedin.com/posts/raid...permutation-activity-6721489199700156416-VsUv
raid
Dad Photographer
Each cluster identifies an outbreak in COVID-19 in recent weeks. The colors of the counties indicate how low or high the long term rates are from January 22-September 22.
raid
Dad Photographer
Let me know if you have any questions.
I can see that many green counties (with low long term rates) now have outbreaks.
I can see that many green counties (with low long term rates) now have outbreaks.
raid
Dad Photographer
I posted these results also on LinkedIn.com.
raid
Dad Photographer
If you have any questions about the maps, just ask me. I want to provide useful information to all on the outbreaks and the long-term rates of COVID-19, It is a very accurate disease surveillance software (SaTScan) that was developed at the National Cancer Institute mainly by Professor Martin Kulldorff (Harvard University).
It can be seen in the maps that the outbreaks are now happening in States that had a low rate of COVID-19.
It can be seen in the maps that the outbreaks are now happening in States that had a low rate of COVID-19.
raid
Dad Photographer
This part includes a spatial analysis of COVID-19 data for the first two weeks of September 2020.
Six of my research students in STA6912 at the University of West Florida have completed their analysis of COVID-19 data from usafacts.org [Bionca Shields, Jon Stevenson, Misty Uher, Conor Voreis, Luke Hodgkins, Kyle Stein]. The time period covered is for September 1-14. 2020, showing cumulative confirmed cases and deaths only for these 14 days. Like any data sets, there is always room for inaccuracy. Each map has a matching table on which the Top 20 significant clusters are summarized. The Discrete Poisson model in satscan (see www.satscan.org) was used for a purely spatial analysis. ArcMap was used to create the maps. The first map is for clusters of confirmed cases.
1. A heatmap is created from the data with ArcMap.
2. SaTScan is used with the discrete Poisson model to identify significant purely spatial clusters.
3. The clusters are shown as circles or spots on top of the heatmap.
4. The tables show information on the Top 20 clusters in each map. The relative risk (RR) value for a cluster shows how much higher COVID-19 the counties in each cluster have relative to the rest of 3108 counties in the contiguous US.
https://www.linkedin.com/posts/raid...al-analysis-activity-6714546149086248960-45IG
Six of my research students in STA6912 at the University of West Florida have completed their analysis of COVID-19 data from usafacts.org [Bionca Shields, Jon Stevenson, Misty Uher, Conor Voreis, Luke Hodgkins, Kyle Stein]. The time period covered is for September 1-14. 2020, showing cumulative confirmed cases and deaths only for these 14 days. Like any data sets, there is always room for inaccuracy. Each map has a matching table on which the Top 20 significant clusters are summarized. The Discrete Poisson model in satscan (see www.satscan.org) was used for a purely spatial analysis. ArcMap was used to create the maps. The first map is for clusters of confirmed cases.
1. A heatmap is created from the data with ArcMap.
2. SaTScan is used with the discrete Poisson model to identify significant purely spatial clusters.
3. The clusters are shown as circles or spots on top of the heatmap.
4. The tables show information on the Top 20 clusters in each map. The relative risk (RR) value for a cluster shows how much higher COVID-19 the counties in each cluster have relative to the rest of 3108 counties in the contiguous US.
https://www.linkedin.com/posts/raid...al-analysis-activity-6714546149086248960-45IG
raid
Dad Photographer
confirmed cases
