[R] Labelling a fortified GADM map plotted with ggplot and geom_map
kenneth at kidscodejeunesse.org
Mon Mar 19 14:18:57 CET 2018
I am having trouble getting data labels to display over the provinces in a GADM map of Canada.
Specifically, I need the variable "Number" from the data set "by_province", grouped by "region", to appear on the corresponding regions of the map.
The data set "by_province" looks like this:
long lat order hole piece region group Number
-110.37074 60.00006 1 FALSE 1 Alberta Alberta.1 132
-110.36250 60.00006 2 FALSE 1 Alberta Alberta.1 132
-110.35103 60.00006 3 FALSE 1 Alberta Alberta.1 132
and the data set "tract" is the map data without the "Number" variable.
I looked into working this out from by researching the geom_map function here: http://ggplot2.tidyverse.org/reference/geom_map.html
my code looks like this:
# get the raw map data
can_map <- getData('GADM', country = "CAN", level = 1) # download map level with provinces
tract <- fortify(can_map, region = "NAME_1") # transforms data from shapefiles into a dataframe that ggplot can understand, from http://www.kevjohnson.org/making-maps-in-r/
# create subsets of the kcj data if needed, for example by year
kids_data_2017 <- subset(kids_data, year == 2017) # data for the year 2017
kids_data_2018 <- subset(kids_data, year == 2018) # data for the year 2018
# extract the needed data
kids_province_data <- data.table::data.table(kids_data_2017$Province_Territory__c, kids_data_2017$Number_of_kids__c)
names(kids_province_data) <- "Province"
names(kids_province_data) <- "Number"
# sum the data by province
kids_province_sums <- aggregate(.~Province, data = kids_province_data, sum)
# join the data to the map
names(tract) <- "region"
names(kids_province_sums) <- "region"
by_province <- left_join(tract, kids_province_sums)
# create the data map
kids_map <- ggplot(by_province, aes(map_id = region)) + #plots the map in by_province separating by region
geom_map(aes(fill = Number), #generates aestheticsa for the plot
map = tract, #takes the data from tract
color = "#ffffff", #makes the color of the borders between regions white
size = 0.15) + #sets the thickness of the boarder lines
coord_map("polyconic") + #sets the coordinates to polyconic (rounded lat and long)
scale_fill_gradient(name = "Children Reached", #sets the gradient of the value scale: names the scale
low = grey_2, #color of the low end
high = orange_1) + #color of the high end
expand_limits(x = tract$long, #ensure limits include all values for all plots
y = tract$lat) +
labs(x = NULL, #add labels, no x title
y = NULL, #no y title
title = "Number of Children Reached by Province", #map title
subtitle = "2017") + #map subtitle
geom_text(data = by_province, #add a text layer
aes(long, #aethetics for the text, x axis
lat, #y axis
label = Number, #the value to display
size=3)) + #size of the text
theme(axis.ticks = element_blank(), #theme of the graph, no axis ticks
axis.text = element_blank(), #no axis text
panel.background = element_blank(), #blank background
panel.border = element_blank()) #blank border
# save as png
ggsave(kids_map, file = "kids_map.png", width = 6, height = 4.5, type = "cairo-png”)
I have asked this question on stack overflow, and was refered to this answer:
The solution there did not fix my problem, though it did get me closer.
The solution on that post is using a single vector of labels.
My post on stackoverflow has images of my output:
[[alternative HTML version deleted]]
More information about the R-help