How to overlay lat/lon points on a spatial map in ggplot?

Question

I have managed to create a map of Europe below which works.

1. How can I add a dataset (which is in a .csv format) which has lat and lons. I made a dummy dataset below and I need to overlay the lat/lon for each Place and mark it with an x for metricA and o for metricB. Is there a way I can do this?

2. The map comes up really small and the margins are really big. I tried to add theme(plot.margin=grid::unit(c(0,0,0,0), "mm")) based on this example from stack overflow R ggplot, remove white margins in ggsave/ggplot but when I add it before coord_sf(xlim = c(-40, 60), ylim = c(30, 90)) + I get an Error

Error in grid.Call(C_textBounds, as.graphicsAnnot(x$label), x$x, x$y, : polygon edge not found and when I add it after, it doesn't change the margins to make the plot bigger. What can I do to make the figure bigger and the margins smaller?

 library("ggplot2")
 theme_set(theme_bw())
 library("sf")
 library("rnaturalearth") 
 library("rnaturalearthdata")

 world <- ne_countries(scale = "medium", returnclass = "sf") 
 class(world)
             
ggplot(data = world) +
geom_sf(color = "black", fill = "grey") +  
coord_sf(xlim = c(-40, 60), ylim = c(30, 90)) +
ggtitle("Europe") +
xlab('Longitude') + ylab('Latitude'))

Dataset:

 Place  MetricA    MetricB   Lat   Lon
 A                   x       55    -40
 B      x                    60    -20
 C                   x       75     18
 D      x            x       68     3
 E                   x       35     18
 F                   x       74     42
 G      x                    62    -26
 H                   x       30    -30
 I      x            x       69     15

Answer 1

I believe I have an answer to half of your question. Essentially geom_sf behaves somewhat peculiarly relative to other ggplot2 geoms in that it enforces an aspect ratio for maps. I was somewhat frustrated by this at first, but it really makes sense when it comes to mapping. To change the margin size you just have to change the aspect ratio of the plot, either when you export it by adjusting the ratio of the height and width by entering different numbers or in an RStudio session by pulling around the plots pane.

As for your second question, I forget the specifics, but I know the answer is laid out clearly here https://www.r-spatial.org/r/2018/10/25/ggplot2-sf-2.html.

Have a good rest of your day, and good luck with your project.

Answer 2

You can overlay your map with points like so:

1. Convert your dataset to long format via e.g. tidyr::pivot_longer so that you metric variablea become categories of one variable
2. Add the points to your map via geom_point where you map the new metric variable on `shape´
3. Set the desired shapes via scale_shape_manual

Concerning your second question. I was not able to reproduce you error. Adjusting the plot margins worked fine. However, I'm not sure whether this will be sufficient to remove the margins when saving or plotting. As already pointed out by @OmriNewett, the issue is that maps are special in that the aspect ratio is fixed and to get rid of the margins you have to set the ratio of width and height so match the fixed aspect ratio.

library(tidyr)
library(dplyr)
library("ggplot2")
library("sf")
#> Linking to GEOS 3.8.0, GDAL 3.0.4, PROJ 6.3.1
library("rnaturalearth") 
library("rnaturalearthdata")

theme_set(theme_bw())

world <- ne_countries(scale = "medium", returnclass = "sf") 

d <- read.table(text = "Place  MetricA    MetricB   Lat   Lon
 A      NA             x       55    -40
 B      x           NA       60    -20
 C      NA           x       75     18
 D      x            x       68     3
 E      NA           x       35     18
 F      NA           x       74     42
 G      x           NA       62    -26
 H      NA           x       30    -30
 I      x            x       69     15", header = TRUE)

d <- d %>% 
  tidyr::pivot_longer(-c(Place, Lat, Lon), names_to = "Metric") %>% 
  filter(!is.na(value))

ggplot(data = world) +
  geom_sf(color = "black", fill = "grey") +
  geom_point(data = d, aes(x = Lon, y = Lat, shape = Metric)) +
  scale_shape_manual(values = c(MetricA = 4, MetricB = 1)) +
  theme(plot.margin=grid::unit(c(0,0,0,0), "mm")) +
  coord_sf(xlim = c(-40, 60), ylim = c(30, 90)) +
  ggtitle("Europe") +
  xlab('Longitude') + ylab('Latitude')