如何根据season列因子匹配列名，将对应值替换为NA？

你可以通过字符串匹配结合批量列处理实现需求，下面是两种可行的方法：

方法一：使用dplyr的across+grepl #

这种方法和你设想的mutate()思路一致，利用across批量处理非season列，通过grepl精准匹配列名前缀：

library(dplyr)

# 加载示例数据
data <- data.frame(season = c("level1", "level1", "level2", "level2"),
                   level1_01 = c(12, 10, 0, 3),
                   level1_02 = c(12, 10, 0, 3),
                   level2_01 = c(12, 10, 0, 3),
                   level2_02 = c(12, 10, 0, 3),
                   level3_01 = c(12, 10, 0, 3),
                   level3_02 = c(12, 10, 0, 3))

# 处理数据
data_processed <- data %>%
  mutate(across(-season, ~ifelse(grepl(paste0("^", season), cur_column()), ., NA)))

# 查看结果
print(data_processed)

代码说明： #

• across(-season)：指定对除season外的所有列执行操作
• grepl(paste0("^", season), cur_column())：用^确保匹配列名的开头部分，避免类似level11和level1的误匹配
• ifelse(...)：匹配则保留原数值，不匹配则替换为NA

方法二：Base R循环实现 #

如果不想加载dplyr包，也可以用基础R的循环处理：

# 加载示例数据
data <- data.frame(season = c("level1", "level1", "level2", "level2"),
                   level1_01 = c(12, 10, 0, 3),
                   level1_02 = c(12, 10, 0, 3),
                   level2_01 = c(12, 10, 0, 3),
                   level2_02 = c(12, 10, 0, 3),
                   level3_01 = c(12, 10, 0, 3),
                   level3_02 = c(12, 10, 0, 3))

# 遍历非season列
for (col in colnames(data)[-1]) {
  # 提取列名的前缀（去除_及后续内容）
  col_prefix <- sub("_.*", "", col)
  # 把season不等于前缀的行设为NA
  data[data$season != col_prefix, col] <- NA
}

# 查看结果
print(data)

处理后结果： #

season level1_01 level1_02 level2_01 level2_02 level3_01 level3_02
1 level1        12        12        NA        NA        NA        NA
2 level1        10        10        NA        NA        NA        NA
3 level2        NA        NA         0         0        NA        NA
4 level2        NA        NA         3         3        NA        NA

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