0

I want to calculate the distance from the centroid of every US state to the centroid of Kentucky. I achieve this using the following code:

library(rnaturalearth)
library(sf)
library(tidyverse)
library(nngeo)

us <- ne_states("United States of America", returnclass = "sf") |> 
  st_centroid()

kentucky <- us |> 
  filter(name == "Kentucky")

dist_kentucky <- st_connect(us, kentucky) |> 
  st_length() 

as.data.frame(dist_kentucky)

The problem now is that I'm unable to link the distance variable to the name of the state, although I assume that it's done row wise. How can I keep the names of the states when running st_length?

johnny
  • 423
  • 3
  • 10

2 Answers2

1

one approach:

library(rnaturalearth)
library(sf)
library(nngeo)

setNames(
  st_connect(us, kentucky) |> st_length(),
  us$name
)

or you could cbind the distances with state names, dplyr::mutate the us dataframe etc.

I_O
  • 4,983
  • 2
  • 2
  • 15
0

The use case - distances of centroids of multiple polygons from a common origin - is a frequent one. It does not even require sf::st_lenght(), mere sf::st_distance() is usually enough.

Consider this piece of code, built on top of the highly recommended {tigris} package (the gold standard for US admin area data, because Topologically Integrated Geographic Encoding and Referencing system and the US Census Bureau).

Note that:

  • the use of mutate creates a new variable within the existing usa dataframe (including all variables, such as the state abbreviation)
  • the use of dot in sf::st_centroid(.) that tells dplyr to calculate the centroid for current row (and use it as first input of distance measurement; the second will be the KY centroid)

You will get the somewhat expected warning that st_centroid assumes attributes are constant over geometries which should be OK

library(tigris)
library(dplyr)
library(sf)

usa <- states(resolution = "20m") %>% 
  select(STUSPS) # for sake of brevity...

kentucky_centroid <- usa %>% 
  filter(STUSPS == "KY") %>% 
  st_centroid()

usa %>% 
  mutate(distance_from_KY = st_distance(st_centroid(.), kentucky_centroid)) %>% 
  arrange(distance_from_KY)

# Simple feature collection with 56 features and 2 fields
# Geometry type: MULTIPOLYGON
# Dimension:     XY
# Bounding box:  xmin: -179.2311 ymin: -14.60181 xmax: 179.8597 ymax: 71.43979
# Geodetic CRS:  NAD83
# First 10 features:
#    STUSPS distance_from_KY                       geometry
# 1      KY          0.0 [m] MULTIPOLYGON (((-89.5712 36...
# 2      TN     208537.0 [m] MULTIPOLYGON (((-83.98762 3...
# 3      IN     277130.7 [m] MULTIPOLYGON (((-86.3296 38...
# 4      OH     389955.7 [m] MULTIPOLYGON (((-80.51915 4...
# 5      WV     427427.3 [m] MULTIPOLYGON (((-80.85847 3...
# 6      IL     438699.6 [m] MULTIPOLYGON (((-89.17208 3...
# 7      AL     550984.6 [m] MULTIPOLYGON (((-85.4883 30...
# 8      SC     571278.6 [m] MULTIPOLYGON (((-79.45002 3...
# 9      GA     571963.9 [m] MULTIPOLYGON (((-81.09538 3...
# 10     VA     584212.4 [m] MULTIPOLYGON (((-77.76715 3...
Jindra Lacko
  • 7,814
  • 3
  • 22
  • 44