inzidenz über impfquote

impf.R

library(tidyverse)
library(ggrepel)
library(readxl)
download.file("https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/Daten/Impfquotenmonitoring.xlsx?__blob=publicationFile",
              "Impfquotenmonitoring.xlsx")
# https://de.statista.com/statistik/daten/studie/1178874/umfrage/bundeslaender-mit-den-meisten-coronainfektionen-in-der-letzten-woche/
# no automatic download.
read_excel("statistic_id1178874_bundeslaender-mit-den-meisten-coronainfektionen-in-der-letzten-woche.xlsx",
           sheet = 2, skip = 4, col_names = F) %>% 
  set_names(c("Bundesland", "Inzidenz")) %>% 
  filter(Bundesland != "Bundesweit") %>% 
  mutate(Bundesland = recode(Bundesland, "Schleswig Holstein"="Schleswig-Holstein"),
         Inzidenz = as.numeric(Inzidenz))-> inz
read_excel("Impfquotenmonitoring.xlsx",
           sheet = "Impfquote_bundesweit_Alter",
           skip = 3,
           col_names = F) %>% 
  select(c(2,7)) %>% 
  set_names(c("Bundesland","Impfquote")) %>% 
  filter(!is.na(Impfquote)) %>% 
  filter(!Bundesland %in% c("Bundesressorts***",
                            "Gesamt")) %>% 
  mutate(Impfquote = as.numeric(Impfquote)) %>% 
  full_join(inz) %>% 
  ggplot(aes(Impfquote, Inzidenz, label = Bundesland)) +
  geom_point()+
  geom_text_repel()+
  theme_bw()+
  geom_smooth(col = "red", span = .9)
ggsave("impf.jpg")