kjhealy · March 24, 2021 15:19
diff --git a/uk_map_example.r b/uk_map_example.r
 # UK map v. quick example


 ## Libraries
 library(tidyverse)
 library(sf)
 #> Linking to GEOS 3.8.1, GDAL 3.1.4, PROJ 6.3.1


 ## Get the map data

 ## An authoritative source for UK map files is the [ONS Open Geography Portal](https://geoportal.statistics.gov.uk).
 # https://geoportal.statistics.gov.uk/search?collection=Dataset&sort=name&tags=all(BDY_LAD)
 # Look under the `Data` tab for the link to the geojson file. We are going to directly grab a boundary file of UK 
 # local authority areas:
 uk_lads <- read_sf("https://opendata.arcgis.com/datasets/69cd46d7d2664e02b30c2f8dcc2bfaf7_0.geojson")

 uk_lads

 #> Simple feature collection with 382 features and 10 fields
 #> geometry type:  MULTIPOLYGON
 #> dimension:      XY
 #> bbox:           xmin: -8.649996 ymin: 49.88234 xmax: 1.763571 ymax: 60.84575
 #> geographic CRS: WGS 84
 #> # A tibble: 382 x 11
 #>    OBJECTID LAD19CD  LAD19NM     LAD19NMW  BNG_E  BNG_N   LONG   LAT Shape__Area
 #>       <int> <chr>    <chr>       <chr>     <int>  <int>  <dbl> <dbl>       <dbl>
 #>  1        1 E060000… Hartlepool  " "      447157 531476 -1.27   54.7   96512311.
 #>  2        2 E060000… Middlesbro… " "      451141 516887 -1.21   54.5   55229150.
 #>  3        3 E060000… Redcar and… " "      464359 519597 -1.01   54.6  248409004.
 #>  4        4 E060000… Stockton-o… " "      444937 518183 -1.31   54.6  205231500.
 #>  5        5 E060000… Darlington  " "      428029 515648 -1.57   54.5  198812771.
 #>  6        6 E060000… Halton      " "      354246 382146 -2.69   53.3   82869023.
 #>  7        7 E060000… Warrington  " "      362744 388456 -2.56   53.4  178742907.
 #>  8        8 E060000… Blackburn … " "      369490 422806 -2.46   53.7  139386373.
 #>  9        9 E060000… Blackpool   " "      332763 436633 -3.02   53.8   33677587.
 #> 10       10 E060000… Kingston u… " "      511894 431716 -0.304  53.8   70882300.
 #> # … with 372 more rows, and 2 more variables: Shape__Length <dbl>,
 #> #   geometry <MULTIPOLYGON [°]>


 ## Get some sample COVID data from https://coronavirus.data.gov.uk
 # https://api.coronavirus.data.gov.uk/v2/data?areaType=utla&metric=cumDeaths60DaysByDeathDate&format=csv

 covid <- read_csv("https://api.coronavirus.data.gov.uk/v2/data?areaType=utla&metric=cumDeaths60DaysByDeathDate&format=csv")

 #> ── Column specification ────────────────────────────────────────────────────────
 #> cols(
 #>   date = col_date(format = ""),
 #>   areaType = col_character(),
 #>   areaCode = col_character(),
 #>   areaName = col_character(),
 #>   cumDeaths60DaysByDeathDate = col_double()
 #> )

 covid
 #> # A tibble: 55,239 x 5
 #>    date       areaType areaCode  areaName   cumDeaths60DaysByDeathDate
 #>    <date>     <chr>    <chr>     <chr>                           <dbl>
 #>  1 2021-03-22 utla     E10000007 Derbyshire                       2084
 #>  2 2021-03-21 utla     E10000007 Derbyshire                       2084
 #>  3 2021-03-20 utla     E10000007 Derbyshire                       2083
 #>  4 2021-03-19 utla     E10000007 Derbyshire                       2080
 #>  5 2021-03-18 utla     E10000007 Derbyshire                       2078
 #>  6 2021-03-17 utla     E10000007 Derbyshire                       2075
 #>  7 2021-03-16 utla     E10000007 Derbyshire                       2074
 #>  8 2021-03-15 utla     E10000007 Derbyshire                       2072
 #>  9 2021-03-14 utla     E10000007 Derbyshire                       2068
 #> 10 2021-03-22 utla     E08000002 Bury                              587
 #> # … with 55,229 more rows

 ### We rename the `areaCode` to `LAD19CD` so we can join the data to the map file in a moment.
 ### Then we filter the table so we just look at a single day; i.e. just one observation per LAD.

 covid_sm <- covid %>%
  rename(LAD19CD = areaCode) %>%
  filter(date == "2021-03-17")

 covid_sm
 #> # A tibble: 149 x 5
 #>    date       areaType LAD19CD   areaName        cumDeaths60DaysByDeathDate
 #>    <date>     <chr>    <chr>     <chr>                                <dbl>
 #>  1 2021-03-17 utla     E10000007 Derbyshire                            2075
 #>  2 2021-03-17 utla     E08000002 Bury                                   585
 #>  3 2021-03-17 utla     E10000019 Lincolnshire                          1865
 #>  4 2021-03-17 utla     E08000016 Barnsley                               903
 #>  5 2021-03-17 utla     E08000008 Tameside                               757
 #>  6 2021-03-17 utla     E08000019 Sheffield                             1288
 #>  7 2021-03-17 utla     E06000030 Swindon                                303
 #>  8 2021-03-17 utla     E06000033 Southend-on-Sea                        694
 #>  9 2021-03-17 utla     E10000014 Hampshire                             2695
 #> 10 2021-03-17 utla     E10000031 Warwickshire                          1230
 #> # … with 139 more rows


 ### Looks like we have 149 unique places in this data. So there will be a bunch of missing LADs
 ### in the map.

 ### We merge this with our map data:
 uk_lads_covid <- uk_lads %>%
  left_join(covid_sm, by = "LAD19CD")

 # looks OK
 uk_lads_covid %>%
  select(LAD19CD, LAD19NM, geometry, cumDeaths60DaysByDeathDate)

 #> Simple feature collection with 382 features and 3 fields
 #> geometry type:  MULTIPOLYGON
 #> dimension:      XY
 #> bbox:           xmin: -8.649996 ymin: 49.88234 xmax: 1.763571 ymax: 60.84575
 #> geographic CRS: WGS 84
 #> # A tibble: 382 x 4
 #>    LAD19CD  LAD19NM                                  geometry cumDeaths60DaysBy…
 #>    <chr>    <chr>                          <MULTIPOLYGON [°]>              <dbl>
 #>  1 E060000… Hartlepool   (((-1.177633 54.69919, -1.173981 54…                287
 #>  2 E060000… Middlesbrou… (((-1.282626 54.56528, -1.262559 54…                393
 #>  3 E060000… Redcar and … (((-1.149131 54.61433, -1.154624 54…                328
 #>  4 E060000… Stockton-on… (((-1.282626 54.56528, -1.270612 54…                512
 #>  5 E060000… Darlington   (((-1.696926 54.53601, -1.705274 54…                296
 #>  6 E060000… Halton       (((-2.674641 53.35366, -2.630622 53…                304
 #>  7 E060000… Warrington   (((-2.576743 53.44606, -2.57039 53.…                548
 #>  8 E060000… Blackburn w… (((-2.551298 53.75639, -2.465808 53…                455
 #>  9 E060000… Blackpool    (((-3.04795 53.87573, -3.01975 53.8…                515
 #> 10 E060000… Kingston up… (((-0.2414035 53.75491, -0.2516817 …                714
 #> # … with 372 more rows


 ### Now we can draw a rough and ready map
 uk_lads_covid %>%
  ggplot(mapping = aes(fill = cumDeaths60DaysByDeathDate)) +
  geom_sf() +
  labs(fill = "60-Day COVID") +
  theme_void() +
  theme(legend.position = "top")


 ## The basic sequence will be the same always:
 ## 1. Find the boundaries you want and get the map file for them. Import that with read_sf()
 ## 2. Find the data you want. Make sure it is at the same level of observation as your map.
 ## 3. Make sure there is a column you can merge/join on---usually the unique official id of the spatial unit.
 ## 4. Join the tables. Usually this will be a left_join(), with the map data on the left and the the data of interest joined to it
 ## 5. Now you have a table with spatial data and your measures of interest as columns.
 ## 5. Draw your map with the `fill` aesthetic assigned to your statistic of interest.
	# UK map v. quick example


	## Libraries
	library(tidyverse)
	library(sf)
	#> Linking to GEOS 3.8.1, GDAL 3.1.4, PROJ 6.3.1


	## Get the map data

	## An authoritative source for UK map files is the [ONS Open Geography Portal](https://geoportal.statistics.gov.uk).
	# https://geoportal.statistics.gov.uk/search?collection=Dataset&sort=name&tags=all(BDY_LAD)
	# Look under the `Data` tab for the link to the geojson file. We are going to directly grab a boundary file of UK
	# local authority areas:
	uk_lads <- read_sf("https://opendata.arcgis.com/datasets/69cd46d7d2664e02b30c2f8dcc2bfaf7_0.geojson")

	uk_lads

	#> Simple feature collection with 382 features and 10 fields
	#> geometry type: MULTIPOLYGON
	#> dimension: XY
	#> bbox: xmin: -8.649996 ymin: 49.88234 xmax: 1.763571 ymax: 60.84575
	#> geographic CRS: WGS 84
	#> # A tibble: 382 x 11
	#> OBJECTID LAD19CD LAD19NM LAD19NMW BNG_E BNG_N LONG LAT Shape__Area
	#> <int> <chr> <chr> <chr> <int> <int> <dbl> <dbl> <dbl>
	#> 1 1 E060000… Hartlepool " " 447157 531476 -1.27 54.7 96512311.
	#> 2 2 E060000… Middlesbro… " " 451141 516887 -1.21 54.5 55229150.
	#> 3 3 E060000… Redcar and… " " 464359 519597 -1.01 54.6 248409004.
	#> 4 4 E060000… Stockton-o… " " 444937 518183 -1.31 54.6 205231500.
	#> 5 5 E060000… Darlington " " 428029 515648 -1.57 54.5 198812771.
	#> 6 6 E060000… Halton " " 354246 382146 -2.69 53.3 82869023.
	#> 7 7 E060000… Warrington " " 362744 388456 -2.56 53.4 178742907.
	#> 8 8 E060000… Blackburn … " " 369490 422806 -2.46 53.7 139386373.
	#> 9 9 E060000… Blackpool " " 332763 436633 -3.02 53.8 33677587.
	#> 10 10 E060000… Kingston u… " " 511894 431716 -0.304 53.8 70882300.
	#> # … with 372 more rows, and 2 more variables: Shape__Length <dbl>,
	#> # geometry <MULTIPOLYGON [°]>


	## Get some sample COVID data from https://coronavirus.data.gov.uk
	# https://api.coronavirus.data.gov.uk/v2/data?areaType=utla&metric=cumDeaths60DaysByDeathDate&format=csv

	covid <- read_csv("https://api.coronavirus.data.gov.uk/v2/data?areaType=utla&metric=cumDeaths60DaysByDeathDate&format=csv")

	#> ── Column specification ────────────────────────────────────────────────────────
	#> cols(
	#> date = col_date(format = ""),
	#> areaType = col_character(),
	#> areaCode = col_character(),
	#> areaName = col_character(),
	#> cumDeaths60DaysByDeathDate = col_double()
	#> )

	covid
	#> # A tibble: 55,239 x 5
	#> date areaType areaCode areaName cumDeaths60DaysByDeathDate
	#> <date> <chr> <chr> <chr> <dbl>
	#> 1 2021-03-22 utla E10000007 Derbyshire 2084
	#> 2 2021-03-21 utla E10000007 Derbyshire 2084
	#> 3 2021-03-20 utla E10000007 Derbyshire 2083
	#> 4 2021-03-19 utla E10000007 Derbyshire 2080
	#> 5 2021-03-18 utla E10000007 Derbyshire 2078
	#> 6 2021-03-17 utla E10000007 Derbyshire 2075
	#> 7 2021-03-16 utla E10000007 Derbyshire 2074
	#> 8 2021-03-15 utla E10000007 Derbyshire 2072
	#> 9 2021-03-14 utla E10000007 Derbyshire 2068
	#> 10 2021-03-22 utla E08000002 Bury 587
	#> # … with 55,229 more rows

	### We rename the `areaCode` to `LAD19CD` so we can join the data to the map file in a moment.
	### Then we filter the table so we just look at a single day; i.e. just one observation per LAD.

	covid_sm <- covid %>%
	rename(LAD19CD = areaCode) %>%
	filter(date == "2021-03-17")

	covid_sm
	#> # A tibble: 149 x 5
	#> date areaType LAD19CD areaName cumDeaths60DaysByDeathDate
	#> <date> <chr> <chr> <chr> <dbl>
	#> 1 2021-03-17 utla E10000007 Derbyshire 2075
	#> 2 2021-03-17 utla E08000002 Bury 585
	#> 3 2021-03-17 utla E10000019 Lincolnshire 1865
	#> 4 2021-03-17 utla E08000016 Barnsley 903
	#> 5 2021-03-17 utla E08000008 Tameside 757
	#> 6 2021-03-17 utla E08000019 Sheffield 1288
	#> 7 2021-03-17 utla E06000030 Swindon 303
	#> 8 2021-03-17 utla E06000033 Southend-on-Sea 694
	#> 9 2021-03-17 utla E10000014 Hampshire 2695
	#> 10 2021-03-17 utla E10000031 Warwickshire 1230
	#> # … with 139 more rows


	### Looks like we have 149 unique places in this data. So there will be a bunch of missing LADs
	### in the map.

	### We merge this with our map data:
	uk_lads_covid <- uk_lads %>%
	left_join(covid_sm, by = "LAD19CD")

	# looks OK
	uk_lads_covid %>%
	select(LAD19CD, LAD19NM, geometry, cumDeaths60DaysByDeathDate)

	#> Simple feature collection with 382 features and 3 fields
	#> geometry type: MULTIPOLYGON
	#> dimension: XY
	#> bbox: xmin: -8.649996 ymin: 49.88234 xmax: 1.763571 ymax: 60.84575
	#> geographic CRS: WGS 84
	#> # A tibble: 382 x 4
	#> LAD19CD LAD19NM geometry cumDeaths60DaysBy…
	#> <chr> <chr> <MULTIPOLYGON [°]> <dbl>
	#> 1 E060000… Hartlepool (((-1.177633 54.69919, -1.173981 54… 287
	#> 2 E060000… Middlesbrou… (((-1.282626 54.56528, -1.262559 54… 393
	#> 3 E060000… Redcar and … (((-1.149131 54.61433, -1.154624 54… 328
	#> 4 E060000… Stockton-on… (((-1.282626 54.56528, -1.270612 54… 512
	#> 5 E060000… Darlington (((-1.696926 54.53601, -1.705274 54… 296
	#> 6 E060000… Halton (((-2.674641 53.35366, -2.630622 53… 304
	#> 7 E060000… Warrington (((-2.576743 53.44606, -2.57039 53.… 548
	#> 8 E060000… Blackburn w… (((-2.551298 53.75639, -2.465808 53… 455
	#> 9 E060000… Blackpool (((-3.04795 53.87573, -3.01975 53.8… 515
	#> 10 E060000… Kingston up… (((-0.2414035 53.75491, -0.2516817 … 714
	#> # … with 372 more rows


	### Now we can draw a rough and ready map
	uk_lads_covid %>%
	ggplot(mapping = aes(fill = cumDeaths60DaysByDeathDate)) +
	geom_sf() +
	labs(fill = "60-Day COVID") +
	theme_void() +
	theme(legend.position = "top")


	## The basic sequence will be the same always:
	## 1. Find the boundaries you want and get the map file for them. Import that with read_sf()
	## 2. Find the data you want. Make sure it is at the same level of observation as your map.
	## 3. Make sure there is a column you can merge/join on---usually the unique official id of the spatial unit.
	## 4. Join the tables. Usually this will be a left_join(), with the map data on the left and the the data of interest joined to it
	## 5. Now you have a table with spatial data and your measures of interest as columns.
	## 5. Draw your map with the `fill` aesthetic assigned to your statistic of interest.