Select your preferences - country and forecasting horizon

Connect with the author on:

Links for my twitter, linkedin and github accounts.

Table of countries

Table is sorted by total Active cases. New cases = cases in the past day (24h). ActCases/MilPop = Active cases per 1 million population

Information about this app and forecasting COVID-19

Statistics

Forecasted total cumulative cases for the selected country + 90% upper prediction interval

Select multiple countries for comparison

Select one statistic for comparison

Set parameters for comparison of "since first" trajectories graph

Comparison of countries' trajectories since their first total N-th case/ N-th death

Comparison of countries

Information about analysis capabilities in this dashboards tab

Select one statistic for clustering trajectories

Set parameters for comparison of "since first" trajectories data

Select parameters for analysis/ clustering

Clustered countries' trajectories based on selected statistics and parameters (see above)

DTW distance measure and hierarchical clustering with Ward criterion are used. Colored dashed lines are DTW barycenters.

Focus plot for detailed analysis of countries in the selected cluster

Interactive plot - zoom in/out etc.

Dendrogram of clustered countries' trajectories based on similarities of selected statistic

DTW distance measure and hierarchical clustering with Ward criterion are used.

2D MDS Scatter plot of clustered countries' trajectories for selected statistic

MDS - Multidimensional scaling - parametric.

Information about analysis capabilities in this dashboards tab

Select multiple statistics for clustering countries based on these data

Select parameters for analysis/ clustering

2D/MDS Scatter plot of countries for selected statistics

If a number of statistics is equal to 2, then scatter plot is used without MDS. If n > 2 then MDS is always used. MDS - Multidimensional scaling - parametric.

Dendrogram of clustered countries based on similarities of selected statistics

Euclidean distance measure and hierarchical clustering with Ward criterion are used.

Clusters' members averages

World statistics

Forecasted total cumulative cases for the World + 90% upper prediction interval

Forecasted total cumulative deaths for the World + 90% upper prediction interval