extract_from_demographics.R

library(rvest) # process web pages
library(dplyr) # general data pipeleine tools
library(lubridate) # date functions

# saving demographic pages as html into a folder
# called case_demografics_html_asof
# with a datestamp name for the contents
Demographs <- list.files("../case_demografics_html_asof", pattern=".*html$")

extract_vax <- function(x) {
  file_w._path <- paste0("../case_demografics_html_asof/",x)
  report_date <- ymd(x)
  # extract the files with rvest
  tbls <- file_w._path %>%
    read_html() %>%
    html_nodes("table") %>%
    html_table(header=TRUE)
  # in the entire period the vaccine data is table 4
  newHos <- as.data.frame(tbls[[4]])
  newHos$asOf <- report_date
  return(newHos)
}

# extract the table for each webpage
list_of_reports <- lapply(Demographs, extract_vax)

# turn the list of tables into one table
df_reports <- bind_rows(list_of_reports) %>%
  rename(Vax=1,under12cases=2,under12hosptalised=3, ReportDate=4) %>%
  filter(Vax %in% c("Not eligible for vaccination at the time they were reported as a case",
                    "Not eligible for vaccine due to being less than 12 years old")) %>%
  select(ReportDate, under12cases) %>%
  arrange(ReportDate) %>%
  mutate(new_cases = under12cases - lag(under12cases))
write.csv(df_reports, file="~/Desktop/under12.csv", row.names=FALSE)  

under12.csv