从R+ sf中的GPS点创建线段

Question

我有时间戳的GPS动物运动坐标作为一个简单的特征集(8068个特征)，几何类型点。我需要将这些点转换为8067线段。使用grouping #321：https://github.com/r-spatial/sf/issues/321从点创建线的解决方案需要一个group-by字段来将点连接到多条线。但是，使用GPS数据时，点是连续的，并且没有分组字段。

ArcGIS有一个解决方案，XY TO Line：https://pro.arcgis.com/en/pro-app/tool-reference/data-management/xy-to-line.htm。有没有一个R等价于这个函数？

Answer 1

Plotting lines between two sf POINT features in r步骤的解释准确地描述了我的目标: sftraj :这个解决方案也非常有用:sftraj

library(readr)
library(sf)
library(tidyverse)

# Read-in GPS data. Data is in ascending order by date.
gps <- read_csv("X:/gps_xyt_data.csv")

# Create a unique ID column starting at 1 
gps <- tibble::rowid_to_column(gps, "KeyID")

# Create a second unique ID column starting at 0
gps$JoinID <- gps$KeyID - 1

# Select coordinates and KeyId 
start_xy <- gps %>% select(start_x = utm_e, start_y = utm_n, KeyID)

# Select coordinates and JoinID  
end_xy <- gps %>% select(end_x = utm_e, end_y = utm_n, JoinID) 

# Inner join to have start/end coordinate pairs for each record 
start_end_xy <- inner_join(start_xy, end_xy, by = c("KeyID" = "JoinID"))

# Select for start geometries and convert to sf object
pnts_start <- start_end_xy %>% st_as_sf( coords = c("start_x", "start_y"), crs = 26911)

# Select for end geometries and convert to sf object
pnts_end <- start_end_xy %>% st_as_sf(coords = c("end_x", "end_y"), crs = 26911)

# Combine start and end geometries
cbind(pnts_start,pnts_end) -> points_ready 

# Generate line segments via union of paired geometries
line_segments <- as.data.frame(st_sfc(mapply(function(a,b){st_cast(st_union(a,b),"LINESTRING")}, points_ready$geometry, points_ready$geometry.1, SIMPLIFY=FALSE)))

# Add a unique ID to line segments
line_segments <- tibble::rowid_to_column(trj_lines, "KeyID")

# Join the attribute data to line segments
line_segments <- inner_join(line_segments, gps, by = c("KeyID" = "KeyID"))

Answer 2

一个更简洁的整形解决方案：

source_file <- 'INSERT_FILENAME_HERE'

empty <- st_as_sfc("POINT(EMPTY)", crs = 4326)

sf::st_read(source_file) %>% 
  # great circle distances
  st_set_crs(4326) %>% 
  mutate(
    geometry_lagged = lag(geometry, default = empty)
  ) %>%
  # drop the NA row created by lagging
  slice(-1) %>% 
  mutate(
    line = st_sfc(purrr::map2(
      .x = geometry, 
      .y = geometry_lagged, 
      .f = ~{st_union(c(.x, .y)) %>% st_cast("LINESTRING")}
    ))) -> track_lines