无count列时在ggplot2中筛选Top10/Bottom10站点绘制分面子图

步骤1：统计站点总骑行量并筛选目标站点 #

先通过dplyr统计每个起始站点的总骑行次数，再提取计数最高和最低的10个站点名称：

library(dplyr)
library(ggplot2)

# 统计每个起始站点的总骑行次数
station_counts <- rides_stations %>%
  count(starting_station_name, name = "total_rides")

# 提取计数最高的10个站点（with_ties=FALSE确保仅取10个，如需保留并列可删除该参数）
top10_stations <- station_counts %>%
  slice_max(n = 10, order_by = total_rides, with_ties = FALSE) %>%
  pull(starting_station_name)

# 提取计数最低的10个站点
bottom10_stations <- station_counts %>%
  slice_min(n = 10, order_by = total_rides, with_ties = FALSE) %>%
  pull(starting_station_name)

步骤2：过滤原数据并添加分组标识 #

从原数据中筛选出目标站点，并标记其所属分组：

filtered_data <- rides_stations %>%
  filter(starting_station_name %in% c(top10_stations, bottom10_stations)) %>%
  mutate(group = case_when(
    starting_station_name %in% top10_stations ~ "Top 10 骑行量站点",
    starting_station_name %in% bottom10_stations ~ "Bottom 10 骑行量站点"
  ))

步骤3：绘制分面子图 #

通过facet_wrap生成两个子图，同时调整x轴标签避免拥挤：

ggplot(filtered_data, aes(x = starting_station_name, fill = member_status)) +
  geom_bar(position = "dodge") + # 可选：让会员/非会员柱子并排显示，默认是堆叠
  facet_wrap(~group, scales = "free_x") + # 两个子图使用独立x轴
  theme(axis.text.x = element_text(angle = 45, hjust = 1)) + # 旋转x轴标签
  labs(x = "起始站点", y = "骑行次数", fill = "会员状态")

可选：分开绘制两个独立图表 #

如果不需要分面，可单独生成两个图表后拼接：

# 绘制Top10站点图表
top_plot <- ggplot(filtered_data %>% filter(group == "Top 10 骑行量站点"), 
                   aes(x = starting_station_name, fill = member_status)) +
  geom_bar() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  labs(title = "Top 10 骑行量起始站点", x = "站点名称", y = "骑行次数")

# 绘制Bottom10站点图表
bottom_plot <- ggplot(filtered_data %>% filter(group == "Bottom 10 骑行量站点"), 
                      aes(x = starting_station_name, fill = member_status)) +
  geom_bar() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  labs(title = "Bottom 10 骑行量起始站点", x = "站点名称", y = "骑行次数")

# 用gridExtra拼接两个图表（需先安装：install.packages("gridExtra")）
library(gridExtra)
grid.arrange(top_plot, bottom_plot, ncol = 1)

内容的提问来源于stack exchange，提问作者Xray25