如何保留至少含3个字符串重复项的行？R语言数据处理求助

你的核心需求是筛选出同一行中存在某个非空/非NA字符串重复出现至少3次的行，而你之前使用的rowSums(!is.na(df)) >= 3只是统计每行非NA单元格的数量——既没有排除空字符串，也没有检查字符串的重复性，自然无法得到预期结果。

数据结构 #

df <- structure(list(First = c("TAGAA1", "TAGAA2", "GYI1", "GYI2", 
"BBB", "TTT", "GGG10", "NNN", "GGG4A", "GGG4B"), FD1 = c("TAGAA1", 
"TAGAA2", "GYI1", "GYI2", "", "", "GGG10", "", "GGG4A", "GGG4B"
), FD2 = c("TAGAA1", "TAGAA2", "GYI1", "GYI2", "", "", "GGG10", 
"", "GGG4A", "GGG4B"), FD3 = c("TAGAA1", "TAGAA2", "GYI1", "GYI2", 
"", "", "GGG10", "", "", ""), FD4 = c(NA, NA, "GYI1", "GYI2", 
"", "", "GGG10", "", "", ""), FD5 = c("TAGAA1", NA, "GYI1", "GYI2", 
"", "", "GGG10", "", "", ""), FD5.1 = c("TAGAA1", NA, "GYI1", 
"GYI2", "", "", "GGG10", "", "", "")), class = "data.frame", row.names = c(NA, 
-10L))

解决方案 #

1. 先将数据中的空字符串""转换为NA，统一处理无效值：

df[df == ""] <- NA

2. 定义逻辑判断，筛选符合条件的行：

# 对每行进行判断：是否存在某个非NA字符串出现≥3次
keep_rows <- apply(df, 1, function(row) {
  non_na_vals <- na.omit(row)
  if (length(non_na_vals) == 0) return(FALSE)
  # 统计每个字符串的出现次数
  val_counts <- table(non_na_vals)
  # 判断是否有字符串满足出现次数≥3
  any(val_counts >= 3)
})

# 提取符合条件的行
filtered_df <- df[keep_rows, ]

结果说明 #

运行上述代码后，会保留以下行：

• TAGAA1：对应字符串出现5次
• TAGAA2：对应字符串出现3次
• GYI1：对应字符串出现5次
• GYI2：对应字符串出现5次
• GGG10：对应字符串出现5次
• GGG4A：对应字符串出现3次

而BBB、TTT、NNN、GGG4B这些行，因为没有任何字符串重复≥3次，会被过滤掉。

内容的提问来源于stack exchange，提问作者nik