使用R语言的for/if循环转换矩阵并提取列名求助

Got it, let's tackle this problem step by step. You don't need to mess with complicated for/if loops here—R has great vectorized tools to handle this cleanly, which is more efficient and readable than manual loops.

Step 1: Set up your sample data #

First, let's recreate your matrix as a data frame (note: R automatically renames duplicate column names with a suffix like .1/.2 to avoid conflicts, but we can rename them explicitly for clarity):

# Create your sample data
df <- data.frame(
  S = 1:5,
  K1 = c("A", "A", "A", "A", "A"),
  K1 = c("P", "A", "P", "P", "P"),
  K2 = c("A", "A", "A", "A", "A"),
  K2 = c("A", "A", "A", "A", "A"),
  K3 = c("A", "A", "A", "A", "A"),
  K3 = c("A", "A", "A", "A", "A"),
  K4 = c("P", "A", "P", "P", "A"),
  K4 = c("A", "P", "A", "A", "P"),
  K5 = c("A", "P", "A", "A", "A"),
  K5 = c("A", "P", "A", "A", "A")
)

# Rename columns to be explicit (optional but helpful)
colnames(df) <- c("S", "K1_1", "K1_2", "K2_1", "K2_2", "K3_1", "K3_2", "K4_1", "K4_2", "K5_1", "K5_2")

Step 2: Extract "P" column names for each sample #

We'll use apply() to process each row, find columns where the value is "P", and pull their names. Then we'll reshape the result into your desired S V1 V2 format:

Option 1: Keep full column names (with suffixes)

# Extract columns with "P" for each row (exclude the first "S" column)
p_columns <- apply(df[, -1], 1, function(row) {
  colnames(df[, -1])[row == "P"]
})

# Convert the list result to a data frame and combine with the "S" column
result <- cbind(df["S"], do.call(rbind, p_columns))

# Rename the output columns to match your desired format
colnames(result) <- c("S", "V1", "V2")

Option 2: Use original group names (e.g., "K1" instead of "K1_2")

If you want to strip the suffixes and keep just the original K1/K2 etc. labels, modify the function to extract the prefix:

p_columns <- apply(df[, -1], 1, function(row) {
  # Get column names where value is "P"
  cols_with_p <- colnames(df[, -1])[row == "P"]
  # Remove suffixes (e.g., "_1" or "_2") to get original group names
  sub("_.*", "", cols_with_p)
})

result <- cbind(df["S"], do.call(rbind, p_columns))
colnames(result) <- c("S", "V1", "V2")

Step 3: View the result #

Running Option 2 will give you this output, which matches your expected format:

S V1 V2
1 1 K1 K4
2 2 K4 K5
3 3 K1 K4
4 4 K1 K4
5 5 K1 K4

Why this works instead of loops #

• apply() handles the row-wise iteration for you, so you don't have to write a for loop manually.
• Vectorized operations (like row == "P") are faster in R than looping through each element with if statements, and they make the code shorter and easier to debug.

内容的提问来源于stack exchange，提问作者Mystra