Matrices for NDA — Operations, Types, Transpose, Inverse, and Adjoint

intermediate 18 min read

Concept

A matrix is a rectangular arrangement of numbers (or functions) organized into rows and columns, enclosed in brackets. That is the formal definition — but here is the intuition that actually matters for exam thinking.

Think of a matrix as a transformation machine. When you multiply a matrix by a column vector, you are transforming that vector — rotating it, scaling it, or shearing it in some geometric space. This is why matrices are central to linear algebra, and why every entrance exam at the 12th-pass level tests them heavily. The NDA paper is no exception.

A matrix with m rows and n columns is called an m × n matrix. Its size is called the order. The element sitting at the intersection of the ith row and jth column is denoted aᵢⱼ. So if someone says "the (2,3) element is 7", they mean row 2, column 3 contains 7.

Here is the analogy that works in the classroom: treat a matrix like a grid of spreadsheet cells. Addition combines two same-size spreadsheets cell by cell. Multiplication is more like a lookup-and-combine operation — far more powerful, and also far less obvious.

Two things that make matrices different from ordinary numbers:

  1. Multiplication is not commutative. AB ≠ BA in general. This surprises students who are used to 2 × 3 = 3 × 2.
  2. The determinant acts as a single-number summary of the whole matrix, and it governs whether the inverse exists at all.

For NDA specifically, matrices questions almost always reduce to one of three things: computing |AB| or |Aᵀ| using determinant product rules, performing a direct 2 × 2 or 3 × 3 multiplication, or identifying the type of matrix from its structure. The concept section below covers the types. The Deep Dive covers the operations and properties that translate directly into marks.


Deep Dive

Types of Matrices — Know Them Cold

| Name | Defining Property | |---|---| | Row matrix | Only 1 row (1 × n) | | Column matrix | Only 1 column (m × 1) | | Square matrix | m = n | | Zero (Null) matrix | All entries are 0 | | Identity matrix (I) | Square; aᵢᵢ = 1, all other aᵢⱼ = 0 | | Diagonal matrix | Square; aᵢⱼ = 0 whenever i ≠ j | | Scalar matrix | Diagonal with all diagonal entries equal | | Upper triangular | aᵢⱼ = 0 for all i > j | | Lower triangular | aᵢⱼ = 0 for all i < j | | Symmetric | A = Aᵀ, i.e., aᵢⱼ = aⱼᵢ | | Skew-symmetric | A = -Aᵀ, i.e., aᵢⱼ = -aⱼᵢ (diagonal entries must be 0) | | Orthogonal | AAᵀ = I, equivalently Aᵀ = A⁻¹ | | Idempotent | A² = A | | Involutory | A² = I | | Nilpotent | Aⁿ = O for some positive integer n |

NDA questions frequently drop one of these terms in an MCQ stem and ask for a conclusion. Recognise the definition on sight — no derivation needed.

Matrix Operations

Addition and Subtraction: Only defined when matrices have the same order. Add element by element. (A + B)ᵢⱼ = aᵢⱼ + bᵢⱼ. Commutative and associative.

Scalar Multiplication: (kA)ᵢⱼ = k · aᵢⱼ. Every element gets multiplied.

Matrix Multiplication: AB is defined only when the number of columns of A equals the number of rows of B. If A is m × p and B is p × n, then AB is m × n.

The formula: (AB)ᵢⱼ = Σₖ aᵢₖ · bₖⱼ

For 2 × 2 matrices:

(abcd)(efgh)=(ae+bgaf+bhce+dgcf+dh)\begin{pmatrix} a & b \\ c & d \end{pmatrix} \begin{pmatrix} e & f \\ g & h \end{pmatrix} = \begin{pmatrix} ae+bg & af+bh \\ ce+dg & cf+dh \end{pmatrix}

Repeat the row-into-column dot product mentally. With practice, a 2 × 2 product takes under 30 seconds.

Transpose

Aᵀ is obtained by swapping rows and columns: (Aᵀ)ᵢⱼ = aⱼᵢ. A 3 × 2 matrix becomes a 2 × 3 after transposing.

Key properties:

Determinant Properties Used in Matrix MCQs

These four properties are the backbone of every NDA "find the value of |...|" question:

  1. |AB| = |A| · |B|
  2. |Aᵀ| = |A|
  3. |kA| = kⁿ|A| for an n × n matrix (scalar comes out raised to the order, not just once)
  4. |A⁻¹| = 1/|A|

Look — these four lines answer the majority of NDA determinant-of-product questions in under 20 seconds each. Memorise them as a block.

Adjoint and Inverse

Cofactor Cᵢⱼ of element aᵢⱼ: delete row i and column j, compute the determinant of the remaining matrix (the minor Mᵢⱼ), then apply sign: Cᵢⱼ = (-1)^(i+j) · Mᵢⱼ.

Adjoint (adj A): Transpose of the cofactor matrix. That is, (adj A)ᵢⱼ = Cⱼᵢ.

Inverse: A⁻¹ = (adj A) / |A|, valid only when |A| ≠ 0 (i.e., A is non-singular).

Key results:

For a 2 × 2 matrix, the adjoint has a shortcut:

adj(abcd)=(dbca)\text{adj}\begin{pmatrix} a & b \\ c & d \end{pmatrix} = \begin{pmatrix} d & -b \\ -c & a \end{pmatrix}

Swap the diagonal, negate the off-diagonal. This is the fastest hand-calculation path to the inverse of a 2 × 2.

Elementary Operations and Rank

Elementary row operations (EROs): swap two rows, multiply a row by a non-zero scalar, add a scalar multiple of one row to another. These do not change the rank of a matrix. The rank is the number of non-zero rows in row-echelon form — but for NDA purposes, rank questions are less common than the product and inverse questions above.


Memory Tricks & Shortcuts

patternReverse-Order Rule for Products

Whenever you see a formula involving a product of matrices — transpose, adjoint, or inverse — the order reverses: (AB)ᵀ = BᵀAᵀ, (AB)⁻¹ = B⁻¹A⁻¹, adj(AB) = (adj B)(adj A).

One mental hook: think of putting on shoes and socks — you take them off in reverse order. Socks go on first (A first), shoes go on second (B second). To undo: remove shoes (B) first, then socks (A). Standard method of deriving this each time: 60s. Pattern recall: 3s.

patternScalar Out of Determinant — Count the n

Students routinely write |kA| = k|A|. Wrong. For an n × n matrix, each row gets multiplied by k, so |kA| = kⁿ|A|.

Micro-example: A is 3 × 3, k = 2. |2A| = 2³|A| = 8|A|. If |A| = 5, then |2A| = 40, not 10.

Standard error rate on this in mock tests is high. The fix: always write n as a superscript before substituting. Step count: 1 step vs. the common 1-step-but-wrong approach. Gets the answer right 100% vs. roughly 40% without the rule.

pattern2×2 Inverse in Four Moves

For A = [[a,b],[c,d]]:

  1. Compute |A| = ad - bc.
  2. Write adj A: swap a and d, negate b and c.
  3. Divide every element of adj A by |A|.
  4. Done.

No cofactor expansion, no 3×3 determinant. Standard cofactor method for 2×2: 6-8 steps. This method: 4 steps. Time: standard 50s vs. shortcut 20s. Works for any invertible 2×2.

patternDeterminant of Transpose = Determinant of Original

|Aᵀ| = |A| always. So |AAᵀ| = |A| · |Aᵀ| = |A| · |A| = |A|².

When the question gives you |A| and asks for |AAᵀ|, square the given value immediately. If |A| = -2, then |AAᵀ| = (-2)² = 4. This is a one-second read once you see the pattern. Standard expansion approach: 30-40s. Pattern recall: 5s.

eliminationIdentify Matrix Type from Diagonal Pattern

Check these in order and stop as soon as one fits:

  1. All entries zero → Zero matrix.
  2. Square + non-zero only on main diagonal → Diagonal matrix.
  3. Square + all diagonal entries equal + off-diagonal zero → Scalar matrix.
  4. Square + aᵢⱼ = aⱼᵢ → Symmetric.
  5. Square + aᵢⱼ = -aⱼᵢ (so diagonal = 0) → Skew-symmetric.

This elimination chain resolves 90% of "what type is this matrix" MCQs in under 10 seconds without computing anything. Standard approach of checking all definitions: 30-40s.


Fast-Solving Framework

When you see a matrix question in the exam hall, run this decision tree:

Step 1 — What is the question asking?

Step 2 — Is there a given condition you haven't used? If the question states A is symmetric, A is orthogonal, or A² = I, that condition almost always unlocks a one-line solution. Don't ignore it.

Step 3 — Does the answer need to be a number or a matrix? If it is a number (like |AB|), verify sign. Determinants of real matrices can be negative; |A|² is always non-negative.

Step 4 — Check order consistency. Before multiplying, write (m × p)(p × n) = m × n. If sizes don't match, the product is undefined — and that itself might be the answer.


Solved PYQs

Why this question: This is a high-frequency NDA pattern — |AAᵀ| combines the transpose determinant property with the product rule. If you know the two-line rule, this is a guaranteed mark.

Previous Year Questionपिछले वर्ष का प्रश्न2026
If A is a square matrix such that |A| = −2, then |AAᵀ|, where Aᵀ is the transpose of A, is equal to
  1. −4
  2. 1
  3. 2
  4. 4
Solutionसमाधान
|AAᵀ| = |A|·|Aᵀ| = |A|·|A| = |A|² = (−2)² = 4.

Solving path: |Aᵀ| = |A| = -2. Apply |AAᵀ| = |A| · |Aᵀ| = (-2)(-2) = 4. Notice the answer is positive — squaring removes the negative. Do not write -4 just because |A| is negative.


Why this question: The product rule |AB| = |A||B| is tested directly. Students who haven't internalized this rule waste time computing the full product matrix and then its determinant — an approach that is both slower and error-prone for larger matrices.

Previous Year Questionपिछले वर्ष का प्रश्न
If |A| = 3 and |B| = 2, then |AB| = ?
यदि |A| = 3 और |B| = 2 है, तो |AB| = ?
  1. 6
  2. 5
  3. 1
  4. 8
  1. 6
  2. 5
  3. 1
  4. 8
Solutionसमाधान
For square matrices, the determinant of a product equals the product of determinants: |AB| = |A| × |B| = 3 × 2 = 6.
वर्ग मैट्रिक्स के लिए, गुणनफल का सारणिक सारणिकों के गुणनफल के बराबर होता है: |AB| = |A| × |B| = 3 × 2 = 6।

Solving path: |AB| = |A| × |B| = 3 × 2 = 6. No matrix computation needed. If this took you more than 5 seconds, drill the four determinant product rules until they are automatic.


Why this question: Direct 2 × 2 matrix multiplication. Tests whether you can read the product rule without errors under time pressure. The diagonal structure of A makes this slightly faster than a general 2 × 2 case.

Previous Year Questionपिछले वर्ष का प्रश्न
If A = [2 0; 0 3] and B = [1 1; 1 1], then AB = ?
यदि A = [2 0; 0 3] और B = [1 1; 1 1] हैं, तो AB = ?
  1. [2 2; 3 3]
  2. [2 3; 2 3]
  3. [3 3; 2 2]
  4. [2 1; 0 3]
  1. [2 2; 3 3]
  2. [2 3; 2 3]
  3. [3 3; 2 2]
  4. [2 1; 0 3]
Solutionसमाधान
Matrix multiplication: AB = [2×1+0×1 2×1+0×1; 0×1+3×1 0×1+3×1] = [2 2; 3 3].
मैट्रिक्स गुणन: AB = [2×1+0×1 2×1+0×1; 0×1+3×1 0×1+3×1] = [2 2; 3 3]।

Solving path: A = [[2,0],[0,3]], B = [[1,1],[1,1]].

Row 1 of AB: (2×1 + 0×1, 2×1 + 0×1) = (2, 2). Row 2 of AB: (0×1 + 3×1, 0×1 + 3×1) = (3, 3). Result: [[2,2],[3,3]].

Shortcut observation: since A is diagonal, multiplying A by B simply scales row 1 of B by 2 and row 2 of B by 3. A diagonal matrix diag(d₁, d₂) scales rows — this halves your mental calculation time whenever you spot a diagonal matrix.


Common Mistakes


Related Topics


Practice on SarkariRise

Sign up + get 3 free mocks →