如何在R中计算范围型身高与寿命列的相关性并可视化？

1. 数据预处理：将范围值转换为数值均值 #

你的Height..in.和Longevity..yrs.列是范围格式（比如"12-15"），无法直接用cor()计算。首先需要把这些范围转成可计算的数值，最常用的方式是取范围的平均值：

# 加载必要包
library(tidyverse)

# 假设你的数据集名为dog_breeds，先处理身高列
dog_breeds <- dog_breeds %>%
  # 拆分身高范围为上下限并转成数值
  separate(Height..in., into = c("height_low", "height_high"), sep = "-", convert = TRUE) %>%
  # 计算身高均值
  mutate(height_mean = (height_low + height_high)/2) %>%
  # 处理寿命列
  separate(Longevity..yrs., into = c("longevity_low", "longevity_high"), sep = "-", convert = TRUE) %>%
  mutate(longevity_mean = (longevity_low + longevity_high)/2) %>%
  # 移除临时辅助列（可选）
  select(-c(height_low, height_high, longevity_low, longevity_high))

如果数据里存在单值格式的异常值（比如"10"这种没有范围的），可以先做清洗：

# 把单值转换为自身范围，保证拆分逻辑统一
dog_breeds <- dog_breeds %>%
  mutate(
    Height..in. = ifelse(!str_detect(Height..in., "-"), paste0(Height..in., "-", Height..in.), Height..in.),
    Longevity..yrs. = ifelse(!str_detect(Longevity..yrs., "-"), paste0(Longevity..yrs., "-", Longevity..yrs.), Longevity..yrs.)
  )

2. 计算相关性 #

预处理完成后，就可以用cor()函数计算皮尔逊相关系数，记得处理缺失值：

# 计算身高均值与寿命均值的相关性，忽略缺失值
correlation <- cor(dog_breeds$height_mean, dog_breeds$longevity_mean, use = "complete.obs")
cat("身高与寿命的相关系数：", correlation, "\n")

如果小型犬寿命更长的结论成立，你会得到一个负相关系数（身高越低，寿命越长）。

3. 带犬种名称的可视化 #

用ggplot2绘制散点图，并借助ggrepel包避免犬种标签重叠：

# 若未安装ggrepel，先运行install.packages("ggrepel")
library(ggrepel)

# 绘制带标签的相关性散点图
ggplot(dog_breeds, aes(x = height_mean, y = longevity_mean)) +
  geom_point(color = "#2c3e50", size = 2) +
  # 自动调整标签位置避免重叠
  geom_text_repel(aes(label = Breed), size = 3, color = "#e74c3c") +
  # 添加线性趋势线
  geom_smooth(method = "lm", se = FALSE, color = "#3498db") +
  # 设置图表标题与坐标轴标签
  labs(
    x = "平均身高（英寸）",
    y = "平均寿命（年）",
    title = "犬种身高与寿命的相关性",
    subtitle = paste("相关系数：", round(correlation, 2))
  ) +
  theme_minimal()

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