如何在R语言中对9×9矩阵内的分块执行求和操作？

Hey there! Let's walk through how to calculate the sum of sub-blocks in your 9×9 matrix. Since 9 is evenly divisible by 3, we'll focus on splitting the matrix into 3×3 sub-blocks (resulting in a 3×3 matrix of block sums) — this is the most common use case for a matrix like yours.

Method 1: Base R with Index Grouping #

This approach uses core R functions to manually define your blocks and calculate sums:

# First, recreate your matrix (for reproducible testing)
set.seed(123)
m <- matrix(runif(9*9), nrow = 9, ncol=9)

# Define the size of each sub-block (3x3 here)
block_size <- 3

# Split row and column indices into block groups
row_blocks <- split(1:nrow(m), ceiling(1:nrow(m)/block_size))
col_blocks <- split(1:ncol(m), ceiling(1:ncol(m)/block_size))

# Calculate the sum for each combination of row and column blocks
block_sum_results <- outer(row_blocks, col_blocks, function(r, c) sum(m[r, c]))

# Convert the result to a clean matrix
block_sum_matrix <- matrix(block_sum_results, nrow = length(row_blocks), ncol = length(col_blocks))

# View the final block sums
block_sum_matrix

What's happening here: #

• split() groups row/column numbers into chunks of size block_size (e.g., rows 1-3, 4-6, 7-9)
• outer() iterates over every pair of row and column blocks, using sum(m[r,c]) to calculate the total for each sub-matrix
• The final conversion to a matrix gives you a 3×3 output where each value is the sum of one 3×3 sub-block from your original matrix

Method 2: Simplified Base R with tapply() #

This is a more concise way using tapply() to group elements by their block membership:

set.seed(123)
m <- matrix(runif(81), nrow=9, ncol=9)
block_size <- 3

# Create grouping variables for each element's row and column block
row_group <- ceiling(row(m)/block_size)
col_group <- ceiling(col(m)/block_size)

# Calculate sums grouped by row and column block
block_sums <- tapply(m, list(row_group, col_group), sum)

block_sums

Why this works: #

• row(m) and col(m) return the row/column number for every element in the matrix
• ceiling(row(m)/block_size) assigns each element to a row block (1, 2, or 3 for your matrix)
• tapply() aggregates the matrix values by the combined row and column groups, automatically returning a matrix of block sums

Both methods will give you the same result — pick whichever feels more intuitive for you!

内容的提问来源于stack exchange，提问作者nymuffin