使用R语言重编码分类字段后新列全为NA的问题求助

问题背景 #

我有个带education列的数据集，想把该列的分类名称转换成对应数值，但运行代码后，新增的education_numeric列全是NA。代码如下：

library(plyr)                 #Load plyr package 

edu.num <- revalue(x = bank_train$education,replace = 
                     c("illiterate" = 0,
                       "basic.4y" = 4,
                       "basic.6y" = 6,
                       "basic.9y" = 9,
                       "high.school" = 12,
                       "professional.course" = 12,
                       "university.degree" = 16,
                       "unknown" = NA))
bank_train$education_numeric <- as.numeric(levels(edu.num))[edu.num]

（运行后education_numeric列所有值均为NA）

错误原因 #

1. revalue的类型处理陷阱：plyr::revalue本质是修改因子的水平值。如果你的education列是字符型，它会自动转成因子，而且替换后的水平是字符形式的数值（比如"0"、"4"），不是真正的数值类型。后续用as.numeric(levels(edu.num))[edu.num]时，因子的底层整数编码和levels的索引不匹配，直接生成NA。
2. 未覆盖的类别或拼写问题：如果原数据里存在replace列表中没有的类别，或者类别名称拼写/大小写和列表不一致，revalue会把这些值转为NA，最终导致整列全NA。

修复方案 #

方案1：Base R直接映射（最简单） #

不用额外包，直接用命名向量做匹配：

# 定义分类到数值的映射
edu_map <- c(
  "illiterate" = 0,
  "basic.4y" = 4,
  "basic.6y" = 6,
  "basic.9y" = 9,
  "high.school" = 12,
  "professional.course" = 12,
  "university.degree" = 16,
  "unknown" = NA
)

# 生成数值列
bank_train$education_numeric <- edu_map[bank_train$education]

方案2：tidyverse风格（可读性强） #

用dplyr的case_when，逻辑更清晰，适合复杂映射：

library(dplyr)

bank_train <- bank_train %>%
  mutate(education_numeric = case_when(
    education == "illiterate" ~ 0,
    education == "basic.4y" ~ 4,
    education == "basic.6y" ~ 6,
    education == "basic.9y" ~ 9,
    education %in% c("high.school", "professional.course") ~ 12,
    education == "university.degree" ~ 16,
    education == "unknown" ~ NA_real_,
    # 处理未匹配的类别，可根据需求调整
    TRUE ~ NA_real_
  ))

方案3：修复原有plyr代码 #

如果一定要用plyr，调整类型转换逻辑：

library(plyr)

# 先把字符列转成因子（原列已是因子可跳过）
bank_train$education <- as.factor(bank_train$education)

# 替换因子水平为字符型数值，注意NA用NA_character_
edu.num <- revalue(bank_train$education, replace = list(
  "illiterate" = "0",
  "basic.4y" = "4",
  "basic.6y" = "6",
  "basic.9y" = "9",
  "high.school" = "12",
  "professional.course" = "12",
  "university.degree" = "16",
  "unknown" = NA_character_
))

# 先转字符再转数值，避免因子编码干扰
bank_train$education_numeric <- as.numeric(as.character(edu.num))

必做检查 #

• 用unique(bank_train$education)查看所有分类，确保没有遗漏或拼写错误（比如大小写、标点和映射列表不一致）。
• 用class(bank_train$education)确认列类型，字符和因子的处理逻辑有区别。

内容的提问来源于stack exchange，提问作者Riddle