Partition (Block) Matrices

Big matrices are easier to think about once you draw a few lines through them and treat the rectangles inside as single objects. That’s a partitioned matrix — same numbers, just regrouped. The convenient surprise is that those blocks largely behave like scalars: you can add and multiply block-wise, as long as the sizes line up so the inner products make sense. It’s not just exam bookkeeping: the covariance matrix of two groups of features, a graph’s adjacency matrix, and the weight layout of a neural network all carry natural block structure that this same arithmetic exploits.

Why GATE cares: when the blocks land in a block-diagonal or block-triangular pattern, two normally-painful operations get easy. The determinant is just the product of the diagonal blocks’ determinants, and a block-diagonal matrix inverts block by block. That’s the entire shortcut.

Blocks behave like scalars

Multiplying two block matrices follows the ordinary row-times-column rule, but with blocks in place of numbers. For the 2-by-2 block layout above,

[ A  B ] [ E  F ]   [ AE + BG    AF + BH ]
[ C  D ] [ G  H ] = [ CE + DG    CF + DH ]

— identical in shape to scalar 2-by-2 multiplication, except each product like AE is a matrix product (so A’s column count must match E’s row count, and the order AE must be kept, never EA).

Block-triangular and block-diagonal determinants. If the lower-left block is zero (block upper-triangular) — or both off-diagonal blocks are zero (block-diagonal) — the determinant collapses to the product of the diagonal blocks:

det [ A  B ]  =  det(A) · det(D)            (block-triangular: lower-left = 0)
    [ 0  D ]

det diag(A, B, C, …)  =  det(A) · det(B) · det(C) · …   (block-diagonal)

And a block-diagonal matrix inverts block-by-block: diag(A, B)^(−1) = diag(A^(−1), B^(−1)), provided each diagonal block is itself invertible.

How GATE asks this

Usually an MCQ or MSQ: you are shown a matrix with an obvious block-triangular or block-diagonal structure (often a chunk of zeros in a corner) and asked for its determinant, or asked which block-matrix statements hold. The shortcut is to spot the structure and multiply the diagonal blocks’ determinants instead of expanding the whole matrix.

Worked example

Take the block-diagonal matrix M = diag(A, B) with

A = [ 2  0 ]   det(A) = 2·2 − 0·0 = 4
    [ 0  2 ]

B = [ 1  1 ]   det(B) = 1·3 − 1·0 = 3
    [ 0  3 ]

so, written out, M is the 4-by-4

M = [ 2  0 | 0  0 ]
    [ 0  2 | 0  0 ]
    [ ---------- ]
    [ 0  0 | 1  1 ]
    [ 0  0 | 0  3 ]

By the block-diagonal rule the determinant is the product of the diagonal blocks’ determinants:

det(M) = det(A) · det(B) = 4 · 3 = 12.

No 4-by-4 cofactor expansion needed — the structure does the work.

Quick check