R语言中ntile分位数分组后出现多范围的问题排查

你遇到的问题核心在于分组区间的计算逻辑错误：

• ntile(height, 3)是基于整个数据集的身高值划分出3个等数量的组，但你后续在group_by(country, gender, height_group, age_group)后，计算的是每个细分小组内的身高极值——这相当于把同一个全局height_group拆成了更小的子集，每个子集的身高范围自然不同，因此会出现同一个height_group对应多个height_range的情况。

下面提供两种符合需求的修正代码，你可以根据偏好选择：

方案一：直接生成全局分位数区间（推荐） #

用cut()结合quantile()直接为每个观测分配全局统一的分位数区间，避免后续区间混乱：

library(dplyr)

set.seed(123)
n <- 100
country <- sample(c("USA", "Canada", "UK"), n, replace = TRUE)
gender <- sample(c("M", "F"), n, replace = TRUE)
age <- sample(18:100, n, replace = TRUE)
height <- runif(n, min = 150, max = 180)
owns_bicycle <- sample(c("Yes", "No"), n, replace = TRUE)

df <- data.frame(country, gender, age, height, owns_bicycle)

# 修正后的核心代码
df %>%
  # 生成全局统一的身高3分位数区间
  mutate(
    height_group = cut(height, 
                      breaks = quantile(height, probs = c(0, 1/3, 2/3, 1)),
                      include.lowest = TRUE, 
                      dig.lab = 3),
    # 生成全局统一的年龄5分位数区间
    age_group = cut(age, 
                   breaks = quantile(age, probs = seq(0, 1, 1/5)),
                   include.lowest = TRUE, 
                   dig.lab = 3)
  ) %>%
  group_by(country, gender, height_group, age_group) %>%
  summarise(
    count = n(),
    percent_own_bicycle = mean(owns_bicycle == "Yes") * 100
  ) %>%
  ungroup()

方案二：保留数字分组+关联全局区间 #

如果需要保留ntile()生成的数字分组编号，可以先提前计算全局分组对应的区间，再合并到结果中：

library(dplyr)

set.seed(123)
n <- 100
country <- sample(c("USA", "Canada", "UK"), n, replace = TRUE)
gender <- sample(c("M", "F"), n, replace = TRUE)
age <- sample(18:100, n, replace = TRUE)
height <- runif(n, min = 150, max = 180)
owns_bicycle <- sample(c("Yes", "No"), n, replace = TRUE)

df <- data.frame(country, gender, age, height, owns_bicycle)

# 提前计算全局身高分组对应的区间
height_intervals <- df %>%
  mutate(height_group = ntile(height, 3)) %>%
  group_by(height_group) %>%
  summarise(height_range = paste0(round(min(height), 1), "-", round(max(height), 1))) %>%
  ungroup()

# 提前计算全局年龄分组对应的区间
age_intervals <- df %>%
  mutate(age_group = ntile(age, 5)) %>%
  group_by(age_group) %>%
  summarise(age_range = paste0(min(age), "-", max(age))) %>%
  ungroup()

# 主计算流程
df %>%
  mutate(height_group = ntile(height, 3),
         age_group = ntile(age, 5)) %>%
  group_by(country, gender, height_group, age_group) %>%
  summarise(
    count = n(),
    percent_own_bicycle = mean(owns_bicycle == "Yes") * 100
  ) %>%
  ungroup() %>%
  # 合并全局区间
  left_join(height_intervals, by = "height_group") %>%
  left_join(age_intervals, by = "age_group") %>%
  # 调整列顺序
  select(country, gender, height_group, height_range, age_group, age_range, count, percent_own_bicycle)

内容的提问来源于stack exchange，提问作者stats_noob