index.Rmd

---
title: "Covid-19 County Case Tracker"
runtime: shiny
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## About
While many sources are providing national data on the UK coronavirus outbreak, these figures are largely driven by the huge amount of cases in major cities. In order to see the real impact of Covid-19 on your local community, **this application provides visualisations of the coronavirus data that has been made available for counties in England, and health boards in Scotland and Wales.** Over the coming months, these visualisations will help to track how your community's efforts are helping to flatten your county's specific Covid-19 curve.

All Covid-19 data for English counties has been sourced from the UK Department of Health and Social Care, all data for Scottish counties has been sourced from the Scottish Government, and all data for Welsh counties has been sourced from Public Health Wales. 

## England County Covid-19 Data (Scotland, Wales Below)

As of June 7th 2020, this data is up-to-date with the official figures released by the UK Department of Health and Social Care.

**On May 30th, the Department of Health and Social Care removed a considerable number of cases from the data. The reason for this has been given on their website as: ** 

***'... it has been identified that a group of test results from the previous 2 months, from a single NHS laboratory, were incorrectly reported to the national data-set as positive. Only a small proportion (<1%) of cases has been removed from the cumulative England total, but for some areas in the West Midlands (the region affected by this data correction), the proportions of cases removed will be greater.'***

**This may explain a large drop in cases for West Midlands counties seen in the plots below (for example, Birmingham) on May 30th.**

**On May 31st, it seems these figures were revised downwards again. It is currently not been made clear why.**

Select your county from the drop-down menu:
```{r eruptions, echo=FALSE}
library(dplyr, warn.conflict = FALSE)
library(ggplot2, warn.conflict = FALSE)
library(bbplot, warn.conflict = FALSE)
library(tidyr, warn.conflict = FALSE)
path_eng <- "https://coronavirus.data.gov.uk/downloads/csv/coronavirus-cases_latest.csv"
eng_cases <- suppressMessages(readr::read_csv(file = path_eng)) %>%
filter(`Area type` == "Upper tier local authority") %>%
mutate(date = lubridate::ymd(`Specimen date`)) %>%
select(date, region = "Area name", cases = "Cumulative lab-confirmed cases") %>%
arrange(date)

counties <- sort(unique(eng_cases$region))
selectInput(
  'county', label = 'County:',
  choices = counties, selected = "Oxfordshire"
)

renderPlot({
cases <- eng_cases %>%
  rename(local = cases) %>%
  mutate(imported = 0) 

cases <- data.table::as.data.table(cases)

cases <- cases[region==input$county]

full_cases <- cases %>%
  complete(date = seq.Date(min(date), max(date), by="day")) %>%
  fill(`region`) %>%
  fill(`local`) %>% 
  filter(date >= "2020-03-09")

cases <- full_cases$local

dates <- seq.Date(min(full_cases$date), max(full_cases$date), by = "day")

df <- data.frame(dates,cases) 
last_point <- df %>% # most recent value
  slice(which.max(dates))
ypos = max(cases) - 0.005*diff(range(cases))
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$county,  "Covid-19 Cases"), title = paste(input$county,  "Total Covid-19 Cases (Linear Scale)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_point(data = last_point, aes(x = dates, y = cases), col = "#007f7f", shape = 21, fill = "white", size = 2, stroke = 1.7) +
geom_text(data = last_point, aes(x = dates, y = cases, label = cases), size = 6, vjust = 2.0, nudge_x = 0.5) +
geom_vline(data=df, xintercept=as.numeric(dates[dates=="2020-03-23"]), linetype=2) +
annotate(geom="label", x=dates[dates=="2020-03-23"], label="UK lockdown begins", y=ypos, fontface="italic", colour="grey10", hjust=1.03, vjust=1, size=5) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})

renderPlot({
cases <- eng_cases %>%
  rename(local = cases) %>%
  mutate(imported = 0) 

cases <- data.table::as.data.table(cases)

cases <- cases[region==input$county]

full_cases <- cases %>%
  complete(date = seq.Date(min(date), max(date), by="day")) %>%
  fill(`region`) %>%
  fill(`local`) %>% 
  filter(date >= "2020-03-09")

cases <- full_cases$local

dates <- seq.Date(min(full_cases$date), max(full_cases$date), by = "day")

df <- data.frame(dates,cases) 
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
labs(x = "Date", y = paste(input$county,  "Covid-19 Cases"), title = paste(input$county,  "Total Covid-19 Cases (Logarithmic Scale)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
scale_y_continuous(trans = 'log2') +
geom_hline(yintercept = 0, size = 2, colour="#333333") +
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_point(data = last_point, aes(x = dates, y = cases), col = "#007f7f", shape = 21, fill = "white", size = 2, stroke = 1.7) +
geom_text(data = last_point, aes(x = dates, y = cases, label = cases), size = 6, vjust = 2.0, nudge_x = 0.5) +
geom_vline(data=df, xintercept=as.numeric(dates[dates=="2020-03-23"]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})

renderPlot({
cases <- eng_cases %>%
  rename(local = cases) %>%
  mutate(imported = 0) 

cases <- data.table::as.data.table(cases)

cases <- cases[region==input$county]

full_cases <- cases %>%
  complete(date = seq.Date(min(date), max(date), by="day")) %>%
  fill(`region`) %>%
  fill(`local`) %>% 
  filter(date >= "2020-03-09")

cases <- full_cases$local

dates <- seq.Date(min(full_cases$date), max(full_cases$date), by = "day")

A <- cases[-1]
B <- cases[-length(cases)]
diff <- c(0, A-B)
diff[diff==-1] <- 0
diff[1] <- NA

cases <- diff

df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_bar(stat="identity",colour = "white",fill="steelblue") +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$county,  "Covid-19 Cases"), title = paste(input$county,  "Daily Covid-19 Cases")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_vline(data=df, xintercept=as.numeric(dates[dates=="2020-03-23"]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})

renderPlot({
cases <- eng_cases %>%
  rename(local = cases) %>%
  mutate(imported = 0) 

cases <- data.table::as.data.table(cases)

cases <- cases[region==input$county]

full_cases <- cases %>%
  complete(date = seq.Date(min(date), max(date), by="day")) %>%
  fill(`region`) %>%
  fill(`local`) %>% 
  filter(date >= "2020-03-09")

cases <- full_cases$local

dates <- seq.Date(min(full_cases$date), max(full_cases$date), by = "day")

A <- cases[-1]
B <- cases[-length(cases)]
diff <- c(0, A/B)
diff[is.na(diff) | diff=="Inf"] <- 0
diff <- diff - 1
diff[diff==-1] <- 0
diff[1] <- NA
cases <- diff
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$county,  "Covid-19 Cases"), title = paste(input$county,  "Daily % Change (On Previous Day's Total)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
scale_y_continuous(labels = scales::percent) +
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_vline(data=df, xintercept=as.numeric(dates[dates=="2020-03-23"]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})
```

## Scotland Health Board Covid-19 Data

As of June 7th 2020, this data is up-to-date with the official figures released by the Scottish Government.

Select your health board from the drop-down menu:
```{r scottish_eruptions, echo=FALSE}
scottish_data <- read.csv("scotland_covid_data.csv", row.names=1)
scottish_counties <- sort(row.names(scottish_data))
selectInput(
  'region', label = 'Health Board:',
  choices = scottish_counties, selected = "Greater Glasgow and Clyde"
)

renderPlot({
cases <- as.matrix(scottish_data)[input$region,]
dates <- seq(as.Date("2020/03/09"), by = "days", length.out = length(scottish_data[input$region,]))
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ypos = max(cases) - 0.005*diff(range(cases)) 
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$region,  "Covid-19 Cases"), title = paste(input$region,  "Total Covid-19 Cases (Linear Scale)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_point(data = last_point, aes(x = dates, y = cases), col = "#007f7f", shape = 21, fill = "white", size = 2, stroke = 1.7) +
geom_text(data = last_point, aes(x = dates, y = cases, label = cases), size = 6, vjust = 2.0, nudge_x = 0.5) +
geom_vline(data=df, xintercept=as.numeric(dates[15]), linetype=2) +
annotate(geom="label", x=dates[15], label="UK lockdown begins", y=ypos, fontface="italic", colour="grey10", hjust=1.03, vjust=1, size=5) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})
renderPlot({
cases <- as.matrix(scottish_data)[input$region,]
dates <- seq(as.Date("2020/03/09"), by = "days", length.out = length(scottish_data[input$region,]))
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
labs(x = "Date", y = paste(input$region,  "Covid-19 Cases"), title = paste(input$region,  "Total Covid-19 Cases (Logarithmic Scale)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
scale_y_continuous(trans = 'log2') +
geom_hline(yintercept = 0, size = 2, colour="#333333") +
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_point(data = last_point, aes(x = dates, y = cases), col = "#007f7f", shape = 21, fill = "white", size = 2, stroke = 1.7) +
geom_text(data = last_point, aes(x = dates, y = cases, label = cases), size = 6, vjust = 2.0, nudge_x = 0.5) +
geom_vline(data=df, xintercept=as.numeric(dates[15]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})

renderPlot({
A <- scottish_data[,-1]
B <- scottish_data[,-length(scottish_data)]
zeros <- integer(length(scottish_counties))
diff <- cbind(zeros, A-B)
cases <- as.matrix(diff)[input$region,]
dates <- seq(as.Date("2020/03/09"), by = "days", length.out = length(cases))
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_bar(stat="identity",colour = "white",fill="steelblue") +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$region,  "Covid-19 Cases"), title = paste(input$region,  "Daily Covid-19 Cases")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_vline(data=df, xintercept=as.numeric(dates[15]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})

renderPlot({
A <- scottish_data[,-1]
B <- scottish_data[,-length(scottish_data)]
zeros <- integer(length(scottish_counties))
diff <- cbind(zeros, A/B)
diff[is.na(diff) | diff=="Inf"] <- 0
diff <- diff - 1
diff[diff==-1] <- 0
diff[,1] <- NA
cases <- as.matrix(diff)[input$region,]
dates <- seq(as.Date("2020/03/09"), by = "days", length.out = length(cases))
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$region,  "Covid-19 Cases"), title = paste(input$region,  "Daily % Change (On Previous Day's Total)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
scale_y_continuous(labels = scales::percent) +
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_vline(data=df, xintercept=as.numeric(dates[15]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})
```

## Wales Health Board Covid-19 Data

As of June 7th 2020, this data is up-to-date with the official figures released by Public Health Wales.

Select your health board from the drop-down menu:
```{r welsh_eruptions, echo=FALSE}
welsh_data <- read.csv("wales_covid_data.csv", row.names=1)
welsh_counties <- sort(row.names(welsh_data))
selectInput(
  'welsh_region', label = 'Health Board:',
  choices = welsh_counties, selected = "Aneurin Bevan"
)

renderPlot({
cases <- as.matrix(welsh_data)[input$welsh_region,]
dates <- seq(as.Date("2020/03/20"), by = "days", length.out = length(welsh_data[input$welsh_region,]))
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ypos = max(cases) - 0.005*diff(range(cases))
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$welsh_region,  "Covid-19 Cases"), title = paste(input$welsh_region,  "Total Covid-19 Cases (Linear Scale)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_point(data = last_point, aes(x = dates, y = cases), col = "#007f7f", shape = 21, fill = "white", size = 2, stroke = 1.7) +
geom_text(data = last_point, aes(x = dates, y = cases, label = cases), size = 6, vjust = 2.0, nudge_x = 0.5) +
geom_vline(data=df, xintercept=as.numeric(dates[4]), linetype=2) +
annotate(geom="label", x=dates[4], label="UK lockdown begins", y=ypos, fontface="italic", colour="grey10", hjust=-0.03, vjust=1, size=5) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})

renderPlot({
cases <- as.matrix(welsh_data)[input$welsh_region,]
dates <- seq(as.Date("2020/03/20"), by = "days", length.out = length(welsh_data[input$welsh_region,]))
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
labs(x = "Date", y = paste(input$welsh_region,  "Covid-19 Cases"), title = paste(input$welsh_region,  "Total Covid-19 Cases (Logarithmic Scale)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
scale_y_continuous(trans = 'log2') +
geom_hline(yintercept = 0, size = 2, colour="#333333") +
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_point(data = last_point, aes(x = dates, y = cases), col = "#007f7f", shape = 21, fill = "white", size = 2, stroke = 1.7) +
geom_text(data = last_point, aes(x = dates, y = cases, label = cases), size = 6, vjust = 2.0, nudge_x = 0.5) +
geom_vline(data=df, xintercept=as.numeric(dates[4]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})
renderPlot({
A <- welsh_data[,-1]
B <- welsh_data[,-length(welsh_data)]
zeros <- integer(length(welsh_counties))
diff <- cbind(zeros, A-B)
cases <- as.matrix(diff)[input$welsh_region,]
dates <- seq(as.Date("2020/03/20"), by = "days", length.out = length(cases))
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_bar(stat="identity",colour = "white",fill="steelblue") +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$welsh_region,  "Covid-19 Cases"), title = paste(input$welsh_region,  "Daily Covid-19 Cases")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_vline(data=df, xintercept=as.numeric(dates[4]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})

renderPlot({
A <- welsh_data[,-1]
B <- welsh_data[,-length(welsh_data)]
zeros <- integer(length(welsh_counties))
diff <- cbind(zeros, A/B)
diff[is.na(diff) | diff=="Inf"] <- 0
diff <- diff - 1
diff[diff==-1] <- 0
diff[,1] <- NA
cases <- as.matrix(diff)[input$welsh_region,]
dates <- seq(as.Date("2020/03/20"), by = "days", length.out = length(cases))
df <- data.frame(dates,cases)
last_point <- df %>% # most recent value
  slice(which.max(dates))
ggplot(df, aes(x = dates, y = cases)) +
geom_line(colour = "#007f7f", size = 1) +
geom_hline(yintercept = 0, size = 1, colour="#333333") +
labs(x = "Date", y = paste(input$welsh_region,  "Covid-19 Cases"), title = paste(input$welsh_region,  "Daily % Change (On Previous Day's Total)")) +
scale_x_date(date_breaks = "4 days", date_labels = "%d/%m") + 
scale_y_continuous(labels = scales::percent) +
theme(axis.text.x = element_text(angle = 90, vjust=0.5, hjust=1)) +
geom_vline(data=df, xintercept=as.numeric(dates[4]), linetype=2) +
bbc_style() +
theme(plot.title = element_text(hjust = 0.5, size=20))
})
```