R语言：如何识别数据框同一行内的重复分组列数据

Got it, let's figure out how to spot duplicate (nameX, typeX, numX) groups within each row of your dataframe. First, I'll use a complete version of your sample data (since the original cuts off mid-entry) to demonstrate the solution clearly:

# Complete sample dataframe
df <- data.frame(
  key = c('1', '2', '3', '4', '5'),
  name1 = c('black','black','black','red','red'),
  type1 = c('chair','chair','sofa','sofa','plate'),
  num1 = c(4,5,12,4,3),
  name2 = c('black', 'red', 'black', 'green', 'blue'),
  type2 = c('chair','chair','sofa','bed','plate'),
  num2 = c(4,7,12,3,1),
  name3 = c('blue', 'green', 'black', 'blue', 'red'),
  type3 = c('chair','chair','sofa','bed','plate'),
  num3 = c(4,7,12,3,2),
  stringsAsFactors = FALSE
)

Basic Solution (Fixed Number of Groups) #

If you know exactly how many (name/type/num) groups you have, you can use a simple function to check for duplicates per row:

# Function to detect duplicate groups in a single row
find_duplicate_groups <- function(row) {
  # Combine each group's values into a single string for easy comparison
  groups <- list(
    group1 = paste(row["name1"], row["type1"], row["num1"], sep = "|"),
    group2 = paste(row["name2"], row["type2"], row["num2"], sep = "|"),
    group3 = paste(row["name3"], row["type3"], row["num3"], sep = "|")
  )
  # Find all groups that appear more than once
  duplicates <- groups[duplicated(groups) | duplicated(groups, fromLast = TRUE)]
  # Return group names if duplicates exist, else NA
  if (length(duplicates) > 0) {
    return(names(duplicates))
  } else {
    return(NA)
  }
}

# Apply the function to every row and add results as a new column
df$duplicate_groups <- apply(df, 1, find_duplicate_groups)

# View the output
print(df)

When you run this, you'll get a new column duplicate_groups that lists which groups are duplicates in each row (e.g., row 1 will show group1, group2 since those two groups are identical).

Flexible Solution (Any Number of Groups) #

If you might have more groups (like name4/type4/num4, etc.), use this dynamic version that automatically detects all existing groups:

# Generalized function for any number of (nameX, typeX, numX) groups
find_duplicate_groups_general <- function(row) {
  # Extract all group numbers from column names (e.g., 1,2,3 from name1, name2, name3)
  group_nums <- unique(sub("name(\\d+)", "\\1", grep("name\\d+", names(row), value = TRUE)))
  
  # Dynamically create groups for each number
  groups <- lapply(group_nums, function(x) {
    paste(row[paste0("name", x)], row[paste0("type", x)], row[paste0("num", x)], sep = "|")
  })
  names(groups) <- paste0("group", group_nums)
  
  # Find duplicate groups
  duplicates <- groups[duplicated(groups) | duplicated(groups, fromLast = TRUE)]
  return(if (length(duplicates) > 0) names(duplicates) else NA)
}

# Apply the generalized function
df$duplicate_groups <- apply(df, 1, find_duplicate_groups_general)

This version works no matter how many (name/type/num) groups you have in your dataframe—it automatically scans for all columns matching the nameX pattern and builds the corresponding groups.

内容的提问来源于stack exchange，提问作者Adam_S