R语言中如何将vector转换为matrix索引？求替代for循环的高效单行实现

Hey Jordan, awesome question—vectorized operations are where R really shines, so let’s ditch those slow for loops and solve both of your problems with clean, efficient code.

If you need to turn your vector into a set of (row, column) index pairs for a matrix, this is straightforward with seq_along() and cbind(). The row positions map directly to the order of elements in your vector, while the column positions are the values of the vector itself.

Example code:

# Sample input vector
vec <- c(3, 1, 2, 3, 2)

# Generate the index matrix (rows = element order, columns = vector values)
index_matrix <- cbind(row = seq_along(vec), col = vec)

# Check the output
index_matrix
#      row col
# [1,]   1   3
# [2,]   2   1
# [3,]   3   2
# [4,]   4   3
# [5,]   5   2

This gives you a 2-column matrix where each row represents a valid (row, column) index you can use to modify a target matrix later.

To create a matrix where each element’s row is its position in the vector, column is its value, and assign values without a for loop, we can use R’s native vectorized indexing. This method is drastically faster than looping, especially for large vectors.

Example 1: Assign a fixed value (e.g., 1) to target positions #

# Define your input and matrix dimensions
vec <- c(3, 1, 2, 3, 2)
num_rows <- length(vec)
num_cols <- max(vec)

# Create matrix and populate values in two clean lines
result_mat <- matrix(0, nrow = num_rows, ncol = num_cols)
result_mat[cbind(seq_along(vec), vec)] <- 1

# View the final matrix
result_mat
#      [,1] [,2] [,3]
# [1,]    0    0    1
# [2,]    1    0    0
# [3,]    0    1    0
# [4,]    0    0    1
# [5,]    0    1    0

If you want a true one-liner (though readability takes a small hit), you can wrap the assignment into the matrix creation:

result_mat <- `[<-`(matrix(0, nrow = length(vec), ncol = max(vec)), cbind(seq_along(vec), vec), value = 1)

Example 2: Assign the original vector values to target positions #

If you want to place the vector’s own values into their respective (row, column) positions instead of a fixed value, just adjust the right-hand side of the assignment:

result_mat <- matrix(0, nrow = length(vec), ncol = max(vec))
result_mat[cbind(seq_along(vec), vec)] <- vec

Why this works: R’s matrix indexing is vectorized by design—passing a 2-column index matrix to [ lets you modify multiple positions in a single operation, skipping the overhead of for loops entirely. This is especially impactful when working with large datasets.

内容的提问来源于stack exchange，提问作者Jordan