Identify separate spatial lines on a map - r

Question

I have a map of multiple rivers, I want to add an ID column to the geometry dataframe adding a grouping variable defining each of the rivers. I only have the line geometry info, no other metadata, is this possible?
An example:
dat <- structure(list(geometry = structure(list(structure(c(169.023627307075,
169.02315299228, -45.3068517761089, -45.3081870363656), .Dim = c(2L,
2L), class = c("XY", "LINESTRING", "sfg")), structure(c(169.01854883529,
169.018930977689, 169.02315299228, -45.3083004691879, -45.3083134946477,
-45.3081870363656), .Dim = 3:2, class = c("XY", "LINESTRING",
"sfg")), structure(c(169.02315299228, 169.0330663306, -45.3081870363656,
-45.3144702778175), .Dim = c(2L, 2L), class = c("XY", "LINESTRING",
"sfg")), structure(c(169.015997396195, 169.022945130719, -45.3119974282578,
-45.3168289670259), .Dim = c(2L, 2L), class = c("XY", "LINESTRING",
"sfg")), structure(c(169.022945130719, 169.032555385154, -45.3168289670259,
-45.3163448854193), .Dim = c(2L, 2L), class = c("XY", "LINESTRING",
"sfg")), structure(c(169.01868555271, 169.022945130719, -45.3174947235968,
-45.3168289670259), .Dim = c(2L, 2L), class = c("XY", "LINESTRING",
"sfg"))), n_empty = 0L, crs = structure(list(input = "4326",
wkt = "GEOGCS[\"GCS_WGS_1984\",\n DATUM[\"WGS_1984\",\n SPHEROID[\"WGS_84\",6378137,298.257223563]],\n PRIMEM[\"Greenwich\",0],\n UNIT[\"Degree\",0.017453292519943295],\n AUTHORITY[\"EPSG\",\"4326\"]]"), class = "crs"), class = c("sfc_LINESTRING",
"sfc"), precision = 0, bbox = structure(c(xmin = 169.015997396195,
ymin = -45.3174947235968, xmax = 169.0330663306, ymax = -45.3068517761089
), class = "bbox"))), row.names = c(NA, 6L), class = c("sf",
"data.frame"), sf_column = "geometry", agr = structure(integer(0), .Label = c("constant",
"aggregate", "identity"), class = "factor"))
library(ggplot2)
library(sf)
ggplot() +
geom_sf(data = dat, size = 1) +
coord_sf()
In this trivial example there are two "rivers" but I don't have a way of identifying them in the dataframe. I want a column like:
dat$ID <- c("A", "A", "A", "B", "B", "B")
The only way I can think of is comparing coordinate values of line ends but this doesn't seem very efficient. Any help much appreciated.

Answer 1

Calculating cluster IDs, based on intersection, can be done with igraph as follows:
library(sf)
library(igraph)
dat <- structure(...)
# Detect clusters
m = st_intersects(dat, sparse = FALSE)
g = graph_from_adjacency_matrix(m)
dat$id = clusters(g)$membership
dat
## Simple feature collection with 6 features and 1 field
## geometry type: LINESTRING
## dimension: XY
## bbox: xmin: 169.016 ymin: -45.31749 xmax: 169.0331 ymax: -45.30685
## geographic CRS: WGS 84
## geometry id
## 1 LINESTRING (169.0236 -45.30... 1
## 2 LINESTRING (169.0185 -45.30... 1
## 3 LINESTRING (169.0232 -45.30... 1
## 4 LINESTRING (169.016 -45.312... 2
## 5 LINESTRING (169.0229 -45.31... 2
## 6 LINESTRING (169.0187 -45.31... 2
# Plot
plot(st_geometry(dat))
text(st_coordinates(st_centroid(dat)), as.character(dat$id))