R语言数据框按rank_code的S标记分组求和perc_fragments的实现方法

不需要做复杂的字符串正则匹配，按行顺序累计标记组边界即可：每当rank_code严格等于"S"时，为后续所有行开启新分组，直到遇到下一个"S"才切换分组。

• 关键技巧：用cumsum(rank_code == "S")直接生成每组唯一ID，无需循环、无需匹配物种名称，完美规避之前用grepl匹配不准的问题。
• 预处理注意：原始数据中perc_fragments为字符串类型，必须转为数值型才能完成求和计算。

Tidyverse 版本（简洁易维护） #

library(dplyr)

# 你的原始测试数据
df <- structure(list(ref = c("Sample1", "Sample1", "Sample1", "Sample1", 
"Sample1", "Sample1", "Sample1", "Sample1", "Sample1", "Sample1", 
"Sample1", "Sample1", "Sample1", "Sample1", "Sample1", "Sample1", 
"Sample1", "Sample1", "Sample1", "Sample1", "Sample1"), perc_fragments = c("73.43", 
"2.10", "1.40", "0.70", "1.40", "0.70", "0.70", "0.70", "0.70", 
"0.70", "0.70", "0.70", "0.70", "0.70", "0.70", "1.40", "0.70", 
"0.70", "0.70", "1.40", "1.40"), rank_code = c("S", "S1", "S1", 
"S2", "S1", "S1", "S1", "S1", "S1", "S2", "S1", "S1", "S1", "S1", 
"S1", "S", "S", "S", "S", "S", "S1"), scientific_name = c("Escherichia coli", 
"Escherichia coli P12b", "Escherichia coli O157:H7", "Escherichia coli O157:H7 str. SS17", 
"Escherichia coli CFT073", "Escherichia coli O6:H16", "Escherichia coli APEC IMT5155", 
"Escherichia coli UM146", "Escherichia coli C", "Escherichia coli ATCC 8739", 
"Escherichia coli FAP1", "Escherichia coli O27:H7", "Escherichia coli IAI1", 
"Escherichia coli O127:H6", "Escherichia coli PCN061", "Klebsiella pneumoniae", 
"Klebsiella variicola", "Citrobacter freundii", "Shigella dysenteriae", 
"Staphylococcus aureus", "Staphylococci haemolyticus"
)), row.names = c(NA, -21L), class = c("tbl_df", "tbl", "data.frame"
))

# 处理流程
result <- df |> 
  mutate(perc_fragments = as.numeric(perc_fragments)) |> 
  # 生成分组ID：遇到新的S即累计计数+1
  mutate(group_id = cumsum(rank_code == "S")) |> 
  group_by(group_id) |> 
  summarise(
    ref = first(ref),
    rank_code = "S",
    scientific_name = first(scientific_name),
    perc_fragments = sum(perc_fragments)
  ) |> 
  select(-group_id)

基础R版本（无需额外安装包） #

# 预处理
df$perc_fragments <- as.numeric(df$perc_fragments)
df$group_id <- cumsum(df$rank_code == "S")

# 聚合求和
perc_sum <- aggregate(perc_fragments ~ ref + group_id, df, sum)
# 提取每个分组对应的S行物种名
sci_name <- aggregate(scientific_name ~ group_id, df, \(x) x[1])
# 合并结果
result <- merge(perc_sum, sci_name, by = "group_id")
result$rank_code <- "S"
# 调整列顺序
result <- result[, c("ref", "perc_fragments", "rank_code", "scientific_name")]

如果数据包含多个不同的ref样本（如Sample1、Sample2等），需要在生成group_id时按ref先分组，避免跨样本串组：

# Tidyverse中调整后的分组ID生成逻辑
df |> 
  group_by(ref) |> 
  mutate(group_id = cumsum(rank_code == "S")) |> 
  ungroup()

以上代码运行后输出的结果完全匹配提出的分组规则：

• 第一组覆盖首个S行及其后续所有S1/S2行，求和后保留对应S行的物种名
• 独立的S行各自单独成组，保留自身对应的物种名
• S行之后紧跟的S1/S2行全部归到前序最近的S所在分组，最终保留S行的物种名作为组名称

内容的提问来源于stack exchange，提问作者Haakonkas