/R/data/get_popn_nhs.R

https://gitlab.com/pjbirrell/real-time-mcmc · R · 58 lines · 45 code · 6 blank · 7 comment · 2 complexity · 53e48fc597280dd5fefaf77903765b8e MD5 · raw file 

    

  1. library(readr)
    2. 

  2. library(tidyr)
    3. 

  3. library(dplyr)
    4. 

  4. 

  5. source("utils.R")
    6. 

  6. dir.data <- ("../../data/population")
    7. 

  7. ## Load in the population data in long format
    8. 

  8. pop <- read_csv(build.data.filepath(subdir = "", "pop_long_nhsrlo_2018_syoa.csv"))
    9. 

  9. ## Load in a lookup table to get the region names, rather than codes.
    10. 

  10. lookup <- read_csv(build.data.filepath(subdir = "", "nhs_region_lookup.csv"))
    11. 

  11. lookup <- unique(lookup[, c(3, 4)])
    12. 

  12. ## If we're aggregating by age, get the aggregation being used or set to default.
    13. 

  13. if(!exists("age.agg")){
    14. 

  14.     age.agg <- c(0, 1, 5, 15, 25, 45, 65, 91)
    15. 

  15.     age.labs <- c("<1yr","1-4","5-14","15-24","25-44","45-64","65+")
    16. 

  16.     }
    17. 

  17. 

  18. ## Put data into wide format - so it resembles the data sent by the ONS.
    19. 

  19. pop <- pop %>%
    20. 

  20.     filter(Sex == 4) %>%
    21. 

  21.     pivot_wider(id_cols = NHSERApr19CD, names_from = Age, values_from = Population)
    22. 

  22. pop <- pop %>%
    23. 

  23.     mutate("All ages" = apply(pop[, -1], 1, sum)) %>%
    24. 

  24.     inner_join(.,lookup) %>%
    25. 

  25.     rename(Code = NHSERApr19CD,
    26. 

  26.            Name = NHSERApr19NM) %>%
    27. 

  27.     mutate(Geography = "NHS Region") %>%
    28. 

  28.     select(Code,
    29. 

  29.            Name,
    30. 

  30.            Geography,
    31. 

  31.            "All ages",
    32. 

  32.            everything())
    33. 

  33. 

  34. ## Put back into long format with age aggregation.
    35. 

  35. ## Long format...
    36. 

  36. regions <- pop %>%
    37. 

  37.     pivot_longer(-(1:4), names_to = "Age", values_to = "popn")
    38. 

  38. regions$Age <- as.integer(regions$Age)
    39. 

  39. regions$Grp <- cut(regions$Age, age.agg, age.labs,right=FALSE,ordered_result=T)
    40. 

  40. regions.short <- aggregate(regions$popn, by = list(Region=regions$Name,Age=regions$Grp), FUN = sum, drop=TRUE)
    41. 

  41. regions.short$Region <- gsub(" ", "_", regions.short$Region)
    42. 

  42. 

  43. 

  44. ## Define some aggregate regions that may get used...
    45. 

  45. nhs.regions <- list("London" = "London",
    46. 

  46.                     "Outside_London" = c("North East and Yorkshire", "North West", "Midlands", "East of England", "South East", "South West"),
    47. 

  47.                     "UNITED_KINGDOM" = "UNITED KINGDOM",
    48. 

  48.                     "ENGLAND" = unlist(lookup[, 2]),
    49. 

  49. 					"East_of_England" =	"East of England",
    50. 

  50. 					"London" = "London",
    51. 

  51. 					"Midlands" = "Midlands",
    52. 

  52. 					"North_East_and_Yorkshire" = "North East and Yorkshire",
    53. 

  53. 					"North_West" = "North West",
    54. 

  54. 					"South_East" = "South East",
    55. 

  55. 					"South_West" = "South West"
    56. 

  56.                     )
    57. 

  57. 

  58. save(pop, nhs.regions, file = build.data.filepath(subdir = "", "pop_nhs.RData"))
    59.