如何在R语言中基于三个数值编码变量创建分组变量Group

你的代码存在几个关键问题：

• 多余的空group_by()调用，直接移除即可
• ifelse语法错误：它仅接受三个参数（判断条件、条件为真时的取值、条件为假时的取值），你传入多个参数导致逻辑完全混乱
• 逻辑运算符使用错误：未给每组条件添加括号明确优先级，且混用了向量级运算的|和标量级运算的||，此处应统一使用|
• 分组标签未加引号：Group0这类标识需要写成字符串格式"Group0"

根据你的分组规则（Group0/Group3各对应1种变量组合，Group1/Group2各对应3种变量组合），用dplyr的case_when实现会更清晰易读：

data <- 
  data %>%  
  mutate(Group = case_when(
    # Group0：race=1 且 gen=1 且 sexor=1
    race == 1 & gen == 1 & sexor == 1 ~ "Group0",
    # Group1：三种指定组合
    (race == 1 & gen == 2 & sexor == 1) | 
    (race == 2 & gen == 1 & sexor == 1) | 
    (race == 1 & gen == 1 & sexor == 2) ~ "Group1",
    # Group2：三种指定组合
    (race == 2 & gen == 2 & sexor == 1) | 
    (race == 2 & gen == 1 & sexor == 2) | 
    (race == 1 & gen == 2 & sexor == 2) ~ "Group2",
    # Group3：剩余的唯一组合
    race == 2 & gen == 2 & sexor == 2 ~ "Group3"
  ))

如果一定要用ifelse，需要嵌套使用，但可读性远不如case_when：

data <- 
  data %>%  
  mutate(Group = ifelse(race == 1 & gen == 1 & sexor == 1, "Group0",
                       ifelse((race == 1 & gen == 2 & sexor == 1) | 
                              (race == 2 & gen == 1 & sexor == 1) | 
                              (race == 1 & gen == 1 & sexor == 2), "Group1",
                              ifelse((race == 2 & gen == 2 & sexor == 1) | 
                                     (race == 2 & gen == 1 & sexor == 2) | 
                                     (race == 1 & gen == 2 & sexor == 2), "Group2", "Group3"))))

注意：确保你的data数据框中确实存在race、gen、sexor三个数值型变量，且取值符合你设定的1/2范围。

内容的提问来源于stack exchange，提问作者Bailey