Timeline Models

Data source: Excess Mortality, Our World in Data, based on the economist data set.
https://github.com/owid/covid-19-data/tree/master/public/data/excess_mortality

Showing excess mortality for all groups as age groups as a dimensionless p_score (percentage over normal, see documentation). Data range from 2020-01-05 to 2021-12-05. Positive values are flipped, producing valleys in the model.

The US curve is used in the spiral models, both as negative and positive values.

Model oriented to match the sequence of countries in the chart below. Top to bottom: “Chile”,“Russia”, “South Korea”, “United Kingdom”, “United States”.

library(tidyverse)
library(lubridate)
excess <- read_csv("excessMortalityOWD_2020-2021.csv")

excess %>% filter(location%in%c("Chile","Russia", "South Korea", "United Kingdom", "United States")) %>% 
                    ggplot(aes(x=date, y=p_scores_all_ages)) + 
                    geom_line() + 
                    facet_grid(rows=vars(location)) + theme_void()

Spirals

The US data from the last row of the chart above, inspired by a popular NYT infographic.

US Excess mortality 2020-Dec. 2021, data starts in the center, moves outward over time. Left: positive values as peaks; Right: positive values as valleys.

Smoothed age groups by country

Showing the excess mortality for different age groups in 9 countries as a single snapshot (Oct. 17, 2021). The mortality is expressed as dimensionless p_scores (see data documentation). The age groups are 1=baseline, 2=15-64, 3=65-74, 4=75-85, 5=over 85.

The line graphs are smoothed as splines and rotated around the over 85 group, which forms the center of the resulting rings. The outer ring is the age group 15-64. Values are flipped (positive values go into the base, negative ones protrude).

Model oriented to match the sequence of countries in the chart below. Top left to bottom right: “Iceland”,“Israel”,“Latvia”,“New Zealand”,“Norway”,“Poland”,“Slovenia”,“Spain”,“Switzerland”

five <- c("Austria","Belgium","Bulgaria","Chile","Denmark",
          "Estonia","Finland","France","Germany","Hungary",
          "Iceland","Israel","Latvia","Lithuania","Netherlands",
          "New Zealand","Norway","Poland","Portugal","Slovakia",
          "Slovenia","Spain","Switzerland","United Kingdom","United States")

three <- c("Iceland","Israel","Latvia",
           "New Zealand","Norway","Poland",
           "Slovenia","Spain","Switzerland")

excess %>% filter(location%in%three&ymd(date)=="2021-10-17") %>%
  group_by(location) %>%
  summarise(p_scores_0=0,
            p_scores_15_64=mean(p_scores_15_64), 
            p_scores_65_74=mean(p_scores_65_74), 
            p_scores_75_84=mean(p_scores_75_84), 
            p_scores_85plus=mean(p_scores_85plus)) -> age

age %>% pivot_longer(cols=2:6, values_to = "p_scores") %>%
  mutate(agegroup=as.numeric(factor(name))) -> age_long

age_long %>% 
  ggplot(aes(x=agegroup, y=p_scores))+
  geom_line() + 
  facet_wrap(~location) +
  ggtitle("agegroups: 1=baseline, 2=15-64, 3=65-74, 4=75-85, 5=over 85.")

Rings - stepwise age groups

Showing the average excess mortality for different age groups in 9 countries for the year 2021. The mortality is expressed as dimensionless p_scores (percentage over normal, see documentation). The age groups are 0-14, 15-64, 65-74, 75-85, over 85.

The values are flipped (positive values go into the base, negative ones protrude), and rotated around the over 85 group, which forms the center of the resulting rings. The outer ring is the age group 0-14.

Model oriented to match the sequence of countries in the chart below. Top left to bottom right: “Hungary”, “Iceland”, “Israel”,“New Zealand”, “Norway”, “Poland”,“Slovenia”, “South Korea”, “Switzerland”

## Warning: Removed 1 rows containing missing values (position_stack).