R语言按分组删除首次出现yes前的no行的实现方法

核心逻辑是：按id分组后，计算每组内value == "yes"的累计和，累计和≥1的行即为首次出现yes及之后的所有行，直接过滤保留即可。

方法1：dplyr实现（适合常规数据量） #

先加载dplyr包，再分组过滤：

library(dplyr)

# 示例数据
DT <- data.frame(id = c(1,1,1,2,2,2,2,2,3,3,3,3,3,3),
                value = c("no","yes","yes","no","no","yes","yes","yes","no","yes","yes","yes","no","no"))

result <- DT %>%
  group_by(id) %>%
  # 累计和为0代表还未出现过yes，直接过滤
  filter(cumsum(value == "yes") >= 1) %>%
  ungroup()

运行后打印result即可得到你需要的输出。如果需要处理某组全为no的场景，可以调整过滤条件，比如要保留全no组就改成filter(cumsum(value == "yes") >= 1 | all(value == "no"))。

方法2：data.table实现（适合百万行以上的大数据量） #

用data.table语法运算效率更高：

library(data.table)

# 示例数据
DT <- data.frame(id = c(1,1,1,2,2,2,2,2,3,3,3,3,3,3),
                value = c("no","yes","yes","no","no","yes","yes","yes","no","yes","yes","yes","no","no"))
setDT(DT) # 转换为data.table格式

result <- DT[, .SD[cumsum(value == "yes") >= 1], by = id]

内容的提问来源于stack exchange，提问作者berehan