参考重复分组值修改data.frame中etc1列数值的方法咨询

核心思路 #

• 先对数据做分组：将name、原始etc1值、etc2三者完全相同的行划为同一组，对应你需要修改的同批次行
• 给每个组内的行按顺序生成行号
• 按需给每行的etc1赋值指定值即可

代码实现 #

1. 构造原始示例数据 #

library(dplyr)

df <- tibble(
  name = c("A","A","A","B","B","B","A","A","A"),
  value = c(9,10,11,2,40,50,2,40,50),
  etc1 = c(1,1,1,1,1,1,4,4,4),
  etc2 = c("X","X","X","Y","Y","Y","X","X","X")
)

2. 自定义指定值替换 #

如果每组的新值是提前定义好的，你可以用case_when直接匹配赋值：

# 先给每组内的行编序号
df <- df %>% 
  group_by(name, etc1, etc2) %>% 
  mutate(group_row = row_number()) %>% 
  ungroup()

# 按需求替换etc1的值
df <- df %>% 
  mutate(etc1 = case_when(
    name == "A" & etc2 == "X" & etc1 == 1 ~ group_row,
    name == "B" & etc2 == "Y" & etc1 == 1 ~ group_row,
    name == "A" & etc2 == "X" & etc1 == 4 ~ case_match(group_row, 1~4, 2~5, 3~7)
  )) %>% 
  select(-group_row) # 删除临时序号列

3. 固定规则自动生成新值 #

如果新值符合规则，比如每组以初始etc1值为起点依次递增1，不需要写死指定值，直接按组计算即可：

df %>% 
  group_by(name, etc1, etc2) %>% 
  mutate(etc1 = first(etc1) + row_number() - 1) %>% 
  ungroup()

4. base R 实现方案 #

不需要依赖第三方包的实现方法：

# 生成分组ID
df$group_id <- as.integer(interaction(df$name, df$etc1, df$etc2))
# 生成组内行号
df$row_in_group <- ave(df$group_id, df$group_id, FUN = seq_along)
# 赋值替换
df$etc1 <- ifelse(df$group_id == 3 & df$row_in_group == 3, 7, df$etc1 + df$row_in_group - 1)
# 删除临时列
df <- df[, !names(df) %in% c("group_id", "row_in_group")]

5. 批量替换方案 #

如果需要修改的组和值很多，可以先做映射表再左连接匹配，效率更高：

# 构造替换映射表
replace_map <- tibble(
  name = rep(c("A","B","A"), each=3),
  etc2 = rep(c("X","Y","X"), each=3),
  origin_etc1 = rep(c(1,1,4), each=3),
  row_in_group = rep(1:3, 3),
  new_etc1 = c(1,2,3,1,2,3,4,5,7)
)

# 左连接替换值
df %>% 
  group_by(name, etc1, etc2) %>% 
  mutate(row_in_group = row_number(), origin_etc1 = etc1) %>% 
  ungroup() %>% 
  left_join(replace_map, by = c("name", "etc2", "origin_etc1", "row_in_group")) %>% 
  mutate(etc1 = new_etc1) %>% 
  select(name, value, etc1, etc2)

内容的提问来源于stack exchange，提问作者Hadi Md Nor