在R中为满足条件后的行添加标记的高效实现方法

你需要按id分组，为每个分组中首次出现value=99的行及其之后的所有行添加标记列new（符合条件为1，否则为0），以下是两种高效实现方式：

方法一：使用dplyr（tidyverse生态） #

适合熟悉tidyverse的用户，代码简洁易读：

library(dplyr)

# 构造示例数据
df <- data.frame(
  id = c(1,1,1,1,1,2,2,3,3,3),
  date = c("2026-01-24","2026-01-24","2026-01-24","2026-01-25","2026-01-25",
           "2026-01-24","2026-01-24","2026-01-24","2026-01-25","2026-01-25"),
  time = c("10:52:44","10:54:44","11:05:41","11:52:44","11:58:14",
           "10:02:44","10:22:44","10:52:12","10:52:44","11:12:23"),
  cost = c(234,236,353,5352,143,124,636,53,436,9473),
  value = c(34,68,99,78,99,99,99,75,99,13)
)

# 生成标记列
df_result <- df %>%
  group_by(id) %>%
  mutate(new = as.integer(cummax(value == 99))) %>%
  ungroup()

原理：

• value == 99生成逻辑向量，首次出现99的位置为TRUE，其余为FALSE
• cummax()对逻辑向量取累积最大值，一旦遇到TRUE，后续所有值都会保持为TRUE
• as.integer()将逻辑值转为1/0，得到最终的new列

方法二：使用data.table（大数据量首选） #

当处理百万级以上的大型数据集时，data.table的速度和内存效率优势明显：

library(data.table)

# 构造示例数据并转为data.table格式
df <- data.frame(
  id = c(1,1,1,1,1,2,2,3,3,3),
  date = c("2026-01-24","2026-01-24","2026-01-24","2026-01-25","2026-01-25",
           "2026-01-24","2026-01-24","2026-01-24","2026-01-25","2026-01-25"),
  time = c("10:52:44","10:54:44","11:05:41","11:52:44","11:58:14",
           "10:02:44","10:22:44","10:52:12","10:52:44","11:12:23"),
  cost = c(234,236,353,5352,143,124,636,53,436,9473),
  value = c(34,68,99,78,99,99,99,75,99,13)
)
setDT(df)

# 生成标记列（原地修改，节省内存）
df[, new := as.integer(cummax(value == 99)), by = id]

原理：

• 通过by = id按分组处理
• 逻辑与dplyr方法一致，cummax()实现累积标记，as.integer()转换为数值
• data.table采用原地修改的方式，避免额外内存开销，适合大数据场景

两种方法最终输出结果均与你期望的格式完全匹配。

内容的提问来源于stack exchange，提问作者AlexB