map_postcode.r

#Required libraries
library(tidyverse)
library(rgdal)
library(rgeos)
library(broom)
library(maps)
#Postcode spatial data
#You need postcode map here https://www.opendoorlogistics.com/wp-content/uploads/Data/UK-postcode-boundaries-Jan-2015.zip
england <- readOGR( 
  dsn= "./osm/postcode/Distribution/" , 
  layer="Districts",
  verbose=FALSE
)
#Get aggregated data as shown by Nabil
samples <- read_delim("sample_postcode.tsv", delim = " ", col_names = F) %>% 
  select(pc = X2, count = X1)

england@data<- left_join(england@data, samples, by = c("name" = "pc"))
england@data$pid  <-  getSpPPolygonsIDSlots(england) 

has_data <- subset(england, !is.na(count))

bbox_has_data <- rgeos::bbox2SP(bbox = bbox(has_data), proj4string = CRS(proj4string(england)))

#Clip area of interest
aoi <- gIntersection(england,bbox_has_data, byid = T)

#Get polygon ID
aoi_id <- getSpPPolygonsIDSlots(aoi)
#Create polygon dataframe
aoi_df <- data.frame(id = aoi_id, row.names = aoi_id)
#Create SpatialPolygonDataFrame
aoi_spdf <- SpatialPolygonsDataFrame(aoi, aoi_df)
aoi_spdf@data <- aoi_spdf@data %>% 
  mutate(id = gsub(" 1", "", id))
#Tidy Spatial data for ggplot
new_data_fortify <- broom::tidy(aoi_spdf, region = "id") %>% arrange()

#Tidy spatial data has samples
tidy_has_data <- fortify(has_data, region="name") %>% 
  arrange()

#Categorise data
has_data@data$samples <- cut(has_data@data$count, c(0,1,5,10,20,30,40,50,60,70,80,90,10000))

#Create centroid label
has_data_map <-  map(has_data, plot=FALSE, fill = TRUE)
has_data_centroids <- maps:::apply.polygon(has_data_map, maps:::centroid.polygon)
has_data_centroids_matrix <- Reduce(rbind, has_data_centroids)
dimnames(has_data_centroids_matrix) <- list(gsub("[^,]*,", "", names(has_data_centroids)),
                                 c("long", "lat"))
postcode_label <- as.data.frame(has_data_centroids_matrix)
postcode_label$label <- row.names(postcode_label) %>% gsub(":[0-9]","",.)

#Plot map
ggplot(tidy_has_data) +
  geom_map(aes(long, lat, map_id = order), map = new_data_fortify, fill = "grey70", color = "#000000", size = 0.15) +
  geom_map(aes(long, lat, map_id = id), map = tidy_has_data, fill = "#ffffff", color = "#000000", size = 0.15) +
  geom_map(data = has_data@data, aes(fill = samples, map_id = name), map = tidy_has_data, color = "white", size = 0.15) +
  geom_text(data=postcode_label, aes(x = long, y =  lat, label=label), size = 2.5, color  = "black") +
  scale_fill_viridis(na.value='#404040', discrete = TRUE, option="viridis") +
  ggtitle("Samples by post code") +
  theme(legend.position="right") +
  theme_void()

ggsave("sample_postcode.png", width = 15, height = 15, units = "cm")