License: MIT
For more information please contact mark.ravina@austin.utexas.edu
To install from GitHub use the devtools package
devtools::install_github("histmr/sfhelper")The geolocate() function takes a data frame with a column of place names (toponyms) and a column of regions, specified by their two-digits ISO codes. The defualt colums names are “place” and “iso”. The function uses the API of the World Historical Gazeteer (https://whgazetteer.org/) to generate a four column data frame with the toponym, ISO code, longitude, and latitude. Many toponyms will retrun multple hits. You will probably want to import a data frame, but here’s an example of a data frame create within R
library(sfhelper)
df <- data.frame("place"=c("Tokyo","Edo","Prague","Prague"),"iso"=c("JP","JP","CZ",""))
geolocate(df,place,iso)
## toponym codes long lat
## 2 Tokyo JP 139.6917 35.6895
## 3 Tōkyō NULL 139.5000 35.7500
## 4 Tokyo NULL 139.7500 35.6667
## 5 Tokyo JP 139.6917 35.6895
## 6 Tōkyō NULL 139.5000 35.7500
## 7 Prague CZ NA NA
## 8 Prague US -92.2667 34.2833
## 9 Prague US -96.6833 35.4833
## 10 Prague US -96.8000 41.3000
## 11 Bohemia NULL -87.1500 30.4833
## 12 Bohemia NULL 14.4333 50.1000
## 13 Praha NULL -93.5667 44.5333Changing the meridian when re-projecting sf object often creates broken polygons and other unwanted artifacts. This function repairs most common errors. For example, using the map from rnaturalearth, the st_transform_repair() function corrects for the new meridian.
library(rnaturalearth)
library(tidyverse)
library(sf)
library(cowplot)
world.sf <- ne_countries(scale = "medium", returnclass = "sf")
new_crs <- "+proj=moll +lat_1=-10 +lon_0=-70"
new.sf <- st_transform(x = world.sf, crs = new_crs)
ggplot() + geom_sf(data=new.sf)
new.sf <- st_transform_repair(x = world.sf, crs= new_crs)
ggplot() + geom_sf(data=new.sf)
In an orthographic projection, the US disappears if we re-project to “+proj=ortho +lat_0=40 +lon_0=-12”.
new_crs <- "+proj=ortho +lat_0=40 +lon_0=-12"
ggplot() + geom_sf(data= world.sf %>% st_transform(new_crs))
The st_transform_repair() fixes the broken polygons
ggplot() + geom_sf(data=world.sf %>% st_transform_repair(new_crs))
As in the examples above, re-projected maps sometimes lack an “edge of the globe” graticule line. The st_transform_outline() function generates that line as an sf object. Using with fill in ggplot(), the new object can be used to represent oceans. At present it does NOT work for orthographic projections with changes in +lat_0=0
new_crs <- "+proj=ortho +lat_0=0 +lon_0=-12"
ggplot() +
geom_sf(data=st_transform_outline(new_crs), fill="lightblue") +
geom_sf(data=world.sf %>% st_transform_repair(new_crs))
new_crs <- "+proj=eck4 +lon_0=-70"
ggplot() +
geom_sf(data=st_transform_outline(new_crs), fill="lightblue") +
geom_sf(data=world.sf %>% st_transform_repair(new_crs))
These transforms make the graticule disappear, but we can restore it with st_graticule()
new_crs <- "+proj=ortho +lon_0=-20"
new.sf <- world.sf %>% st_transform_repair(new_crs)
new_graticule.sf <- st_graticule(new.sf) %>% st_transform(new_crs)
ggplot() + geom_sf(data=st_transform_outline(crs=new_crs), fill="lightblue") +
geom_sf(data=new.sf) +
geom_sf(data=new_graticule.sf, color="white", linewidth=0.3, alpha=0.7)
This function was inspired by Daniel Immerwahr’s How to Hide an Empire (2019), in which he combines,at the same scale, non-contiguous US states and territories. This is approach useful for appreciating the relative areas of different state, countries, etc. Following Immerwahr, let’s show the continental US, Alaska, Hawaii, and the Philippines at the same scale. The st_equal_grid() function takes a list of sf objects, all with projected coordinates in the same units. You can (likely will want to) use different projections for different places, but they must all be in the same units
library(spData)
## To access larger datasets in this package, install the spDataLarge
## package with: `install.packages('spDataLarge',
## repos='https://nowosad.github.io/drat/', type='source')`
logo.sf <- us_states %>%
st_transform("+proj=lcc +lon_0=-90 +lat_1=33 +lat_2=45") %>%
st_union()## continental US
alaska.sf <- alaska %>%
st_transform("+proj=aea +lat_0=50 +lon_0=-154 +lat_1=55 +lat_2=65 +x_0=0 +y_0=0 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs +type=crs") %>% st_union()
hawaii.sf <- hawaii %>% st_transform("+proj=utm +zone=32") %>% st_union()
world_ne.sf <- st_as_sf(ne_countries(scale="medium"))
philippines.sf <- world_ne.sf %>% filter(sov_a3=="PHL") %>% st_union() %>% st_transform(3121) # https://epsg.io/3121
places <- list(logo.sf,alaska.sf,hawaii.sf,philippines.sf)
plot_list <- st_equal_grid(places)To visualize the plots in a grid, use the cowplot package
cowplot::plot_grid(plotlist = plot_list, ncol = 4)