Air quality.R

###   Update ScotPHO Care and Wellbeing indicator: 
#   99132: Air quality

# Data source is DEFRA Population-weighted annual mean PM2.5 data:
# https://uk-air.defra.gov.uk/data/pcm-data#population_weighted_annual_mean_pm25_data


### functions/packages -----

source("1.indicator_analysis.R") 
library(purrr)


### 1. Read in data ----

# Identify data files
data_files <- paste0(data_folder, "Received Data/Air quality/", 
                     list.files(path = paste0(data_folder,"Received Data/Air quality"), pattern = ".csv"))

# Define column names
column_names <- c("areacode", "total", "non_anthropogenic", "anthropogenic", "areaname")

# Read in data
data_list <- lapply(data_files, function(x) read_csv(data_files, col_names = column_names, skip = 3, id = "file_name")) 

# Save as single data frame
dat <- reduce(data_list, full_join)

# Read in geography lookup
dictionary <- readRDS(paste0(lookups, "Geography/opt_geo_lookup.rds")) %>%
  filter(areatype %in% c("Scotland", "Council area")) %>% 
  select(c(code, areaname))



### 2. Prepare data  -----

data <- dat %>% 
  
  # Create year variable by substracting characters from file name
  mutate(year = as.numeric(substr(file_name, start = 68, stop = 71)),
         
         # Tidy LA names
         areaname = str_replace_all(areaname, c(" and " = " & ",
                                                "Edinburgh, City of" = "City of Edinburgh",
                                                "Eilean Siar" = "Na h-Eileanan Siar")),
         
         # Create new columns
         trend_axis = year,
         def_period = paste0(year, " calendar year"),
         lowci = NA, upci = NA, rate = total,
         ind_id = 99132,
         numerator = total) %>%

  # Join area codes by area name
  left_join(dictionary, by = "areaname") %>%
  #air quality data only available for scotland and council area
  droplevels() %>% # remove all traces of non-matched geographies

  # Filter for Scottish data
  filter(str_detect(code, "^S")) %>% 
  
  # Select relevant columns
  select(ind_id, code, year, trend_axis, def_period, numerator, rate, lowci, upci)


### 3. Prepare final files -----

# Save files in folder to be checked
write.csv(data, paste0(data_folder, "Data to be checked/air_quality_shiny.csv"), row.names = FALSE)
write_rds(data, paste0(data_folder, "Data to be checked/air_quality_shiny.rds"))



#END