UK Cases

Tintin’s Covid-19 Charts

Quick Snowy …

Source:

https://ourworldindata.org/coronavirus

‘R is below one in the community’ - Patrick Valence 16th April 2020.

Data from:

Total and daily confirmed COVI-19 cases: total-and-daily-cases-covid-19,csv

Merge these two by column: daily-deaths-covid-19.csv total-covid-deaths-region.csv

buildDf <- function(df)
{
  #
  # Build data frame for an entity
  #
  df$cases <- as.numeric(trim(df$cases))
  df$deaths <- as.numeric(trim(df$deaths))
  
  df['date.int'] <- rep(NA,nrow(df))
  df$date.int <- to.date.int( df$date, 2 )
  #
  # This is day number with 1 for first (by date)
  # entry for this entity.
  #
  df <- arrange(df,date.int)
  df['day.number'] <- rep(NA,nrow(df))
  df$day.number <- c(1:nrow(df))
  
  # Deaths

  df['deaths.ma'] <- rep(NA,nrow(df))
  df$deaths.ma <- ma(df$deaths,3)
  
  df['deaths.na.to.zero'] <- rep(0,nrow(df))
  df$deaths.na.to.zero <- df$deaths
  
  df['deaths.ma.na.to.zero'] <- rep(0,nrow(df))
  df$deaths.ma.na.to.zero <- df$deaths.ma
  
  # Cases
  
  df['cases.ma'] <- rep(NA,nrow(df))
  df$cases.ma <- ma(df$cases,3)
  
  # Estimate of R
  
  df['r.est'] <- rep(NA,nrow(df))
  df$r.est <- r.vec(df$cases.ma)
  
  df['r.est.ma'] <- rep(NA,nrow(df))
  df$r.est.ma <- ma(df$r.est,3)
  
  # Time to double
  
  # Derive time-to-double on ma cases
  df['time.to.double'] <- rep(NA,nrow(df))
  df$time.to.double <- time.to.double(df$cases.ma)
  
  # Now do ma on time to double
  df['time.to.double.ma'] <- rep(NA,nrow(df))
  df$time.to.double.ma <- ma(df$time.to.double,5)
  
  df['time.to.double.from.r'] <- rep(NA,nrow(df))
  df$time.to.double.from.r <- time.to.double.from.r(df$r.est.ma)
  
  # Up/down marker
  
  df['decreased.cases'] <- rep(NA,nrow(df))
  df$decreased.cases <- upDown(df$cases)
  df['decreased.cases.ma'] <- rep(NA,nrow(df))
  df$decreased.cases.ma <- upDown(df$cases.ma)
  
  return(df)
}

#df <- buildDf(df)

# Assign day numbers by country by doing something like:

prt <- TRUE
buildForEntity <- function(a1,a2){
  
  if ( prt ) {
    print('class of a1:')
    print(class(a1));
    print('class of a2:')
    print(class(a2));
    print('names of a1:')
    print(names(a1));
    print('names of a2:')
    print(names(a2));
    print(a2$entity[1])
    prt <<- FALSE
  }
  #
  # This will work but it will set day number arbitrarily.
  #
  return(buildDf(a1))
}

df.new <- df.new %>% group_by(entity) %>% group_modify(buildForEntity)

## [1] "class of a1:"
## [1] "tbl_df"     "tbl"        "data.frame"
## [1] "class of a2:"
## [1] "tbl_df"     "tbl"        "data.frame"
## [1] "names of a1:"
## [1] "code"       "date"       "cum.cases"  "cases"      "cum.deaths"
## [6] "deaths"    
## [1] "names of a2:"
## [1] "entity"
## [1] Afghanistan
## 206 Levels: Afghanistan Albania Algeria Andorra Angola ... Zimbabwe

## Warning: package 'EpiEstim' was built under R version 3.6.3

## Default config will estimate R on weekly sliding windows.
##     To change this change the t_start and t_end arguments.

## TableGrob (3 x 1) "arrange": 3 grobs
##       z     cells    name           grob
## incid 1 (1-1,1-1) arrange gtable[layout]
## R     2 (2-2,1-1) arrange gtable[layout]
## SI    3 (3-3,1-1) arrange gtable[layout]

df.from.march <- filter(df.new, date >= '2020-03-01')
ggplot(
  filter(df.from.march,
         (entity=='France')|
         (entity=='Italy')|
         (entity=='Germany')|
         (entity=='Spain')|
         (entity=='United Kingdom')),
  aes(x=date,y=cum.cases,color=entity)
  ) +
geom_line() +
geom_point() +
theme_minimal()