sf分组点边界框生成、绘图问题及与ArcGIS Pro对比咨询

问题1：更简便的分组边界框生成方法 #

直接使用dplyr分组结合st_bbox和st_as_sfc即可，无需自定义函数，代码简洁易维护：

# 生成带分组ID的轴对齐边界框sf对象
bound_boxes <- Example %>%
  group_by(ID) %>%
  summarise(
    geometry = st_as_sfc(st_bbox(geometry)),  # 将分组点的边界框转为几何对象
    area = st_area(st_as_sfc(st_bbox(geometry))),  # 直接计算面积
    .groups = "drop"
  ) %>%
  st_set_crs(26917)  # 确保CRS与原数据一致

问题2：geom_sf报错原因分析 #

报错核心是额外的sfc列丢失了CRS信息：

• 自定义的st_bbox_by_feature函数中，st_as_sfc(st_bbox(y))生成的边界框sfc对象没有继承原数据的CRS；
• sf对象仅允许一个活跃几何列，将无CRS的sfc列BoundBox追加到原sf对象中，会导致ggplot处理数据时尝试转换所有sfc列，无CRS的列无法完成转换，从而抛出错误。

问题3：保留分组ID的边界框生成方法 #

在分组汇总时直接保留ID列，生成的边界框sf对象自带分组标识，无需额外处理：

# 生成带ID的边界框
bound_boxes <- Example %>%
  group_by(ID) %>%
  summarise(
    geometry = st_as_sfc(st_bbox(geometry)),
    area = st_area(geometry),
    .groups = "drop"
  ) %>%
  st_set_crs(26917)

后续绘图时直接调用该对象，即可按ID着色：

ggplot() +
  geom_sf(data = NC, fill = "lightgray", color = "white") +
  geom_sf(data = Example, aes(color = factor(ID))) +
  geom_sf(data = bound_boxes, aes(fill = factor(ID)), alpha = 0.3, color = NA)

问题4：st_bbox与ArcGIS Minimum Bounding Geometry的差异及对齐方法 #

• st_bbox默认生成轴对齐边界框（AABB）：矩形边与坐标轴平行，仅取x、y的最大最小值；
• ArcGIS默认生成最小旋转矩形（MRR）：能包围所有点的面积最小矩形，允许旋转，与坐标轴不一定平行。

要生成和ArcGIS一致的最小旋转矩形，需使用lwgeom包的st_minimum_bounding_rectangle函数：

# 安装并加载lwgeom
install.packages("lwgeom")
library(lwgeom)

# 生成最小旋转矩形（与ArcGIS默认结果一致）
rotated_boxes <- Example %>%
  group_by(ID) %>%
  summarise(
    geometry = st_minimum_bounding_rectangle(geometry),
    area = st_area(geometry),
    .groups = "drop"
  ) %>%
  st_set_crs(26917)

完整修正代码 #

library(tidyverse)
library(sf)
library(lwgeom)  # 用于生成最小旋转矩形

# 模拟数据
ID <- rep(seq(1:4), each = 4)
NC <- st_read(system.file("shape/nc.shp", package = "sf"), quiet = TRUE) %>% 
  filter(FIPS == c("37007", "37123", "37167", "37179")) %>% 
  st_transform(crs = 26917)
points <- st_sample(NC, size = c(4,4), type = "random") %>% st_transform(crs = 26917)
Example <- bind_rows(list(geometry = points)) %>% 
  mutate(ID = ID) %>% 
  st_as_sf(crs = 26917)

# 1. 生成轴对齐边界框（st_bbox）
axis_aligned_boxes <- Example %>%
  group_by(ID) %>%
  summarise(
    geometry = st_as_sfc(st_bbox(geometry)),
    area = st_area(geometry),
    .groups = "drop"
  ) %>%
  st_set_crs(26917)

# 2. 生成最小旋转矩形（与ArcGIS默认一致）
rotated_boxes <- Example %>%
  group_by(ID) %>%
  summarise(
    geometry = st_minimum_bounding_rectangle(geometry),
    area = st_area(geometry),
    .groups = "drop"
  ) %>%
  st_set_crs(26917)

# 轴对齐边界框绘图
ggplot() +
  geom_sf(data = NC, fill = "lightgray", color = "white") +
  geom_sf(data = Example, aes(color = factor(ID)), size = 2) +
  geom_sf(data = axis_aligned_boxes, aes(fill = factor(ID)), alpha = 0.3, color = "black") +
  labs(title = "轴对齐分组边界框", color = "分组ID", fill = "分组ID") +
  theme_minimal()

# 最小旋转矩形绘图
ggplot() +
  geom_sf(data = NC, fill = "lightgray", color = "white") +
  geom_sf(data = Example, aes(color = factor(ID)), size = 2) +
  geom_sf(data = rotated_boxes, aes(fill = factor(ID)), alpha = 0.3, color = "black") +
  labs(title = "最小旋转分组边界框（与ArcGIS一致）", color = "分组ID", fill = "分组ID") +
  theme_minimal()

内容的提问来源于stack exchange，提问作者Jeffrey Beauvais