Chapter 3 · Matrices and Determinants

3.1 Matrices and Matrix Algebra

A matrix is a rectangular array of numbers that can be added, scaled and multiplied — row by column. This section builds the algebra of matrices, the types of matrices JEE Main asks about, the transpose with its symmetric and skew-symmetric parts, and the traces and powers of square matrices.

A bakery has two outlets. On Saturday outlet P sold $40$ loaves, $25$ cakes and $60$ buns; outlet Q sold $35$ loaves, $30$ cakes and $50$ buns. A loaf costs $\text{₹}\,30$, a cake $\text{₹}\,120$ and a bun $\text{₹}\,10$. To find each outlet's takings you multiply every count by its price and add: P takes $40\cdot30+25\cdot120+60\cdot10=\text{₹}\,4800$ and Q takes $35\cdot30+30\cdot120+50\cdot10=\text{₹}\,5150$. Write the counts as a rectangular table with one row per outlet and the prices as a column, and this "multiply along a row, down a column, and add" becomes a single operation — the product of two matrices. The same operation rotates and stretches pictures on a screen, moves money between accounts in an economic model and tracks a population from one year to the next.

Matrices are the arithmetic of tables. In this section we define them and name the types that appear in JEE Main questions, learn to add, scale and multiply them — and see which rules of ordinary arithmetic survive and which do not ($AB$ need not equal $BA$, and $AB=O$ does not force $A=O$ or $B=O$). Then we meet the transpose, symmetric and skew-symmetric matrices, the trace, and the special matrices whose powers follow a pattern. JEE Main questions on this part of the unit test exactly these skills: a product or a power computed quickly, a property of the transpose used correctly, a count of symmetric or skew-symmetric matrices.

Matrices, Order and Types

📘 Definition 3.1 — Matrix, order and equality

A matrix is a rectangular array of numbers (or expressions) arranged in rows and columns. A matrix with $m$ rows and $n$ columns has order $m\times n$ (read "$m$ by $n$") and $mn$ entries (elements). The entry in row $i$ and column $j$ is written $a_{ij}$, and the whole matrix $A=[a_{ij}]_{m\times n}$:

$$A=\begin{bmatrix}a_{11}&a_{12}&\cdots&a_{1n}\\a_{21}&a_{22}&\cdots&a_{2n}\\\vdots&\vdots& &\vdots\\a_{m1}&a_{m2}&\cdots&a_{mn}\end{bmatrix}.$$

Two matrices are equal, $A=B$, when they have the same order and every entry of $A$ equals the corresponding entry of $B$: $a_{ij}=b_{ij}$ for all $i$, $j$.

So a matrix equation between two $2\times2$ matrices is really four ordinary equations, one per position. The first index is always the row and the second the column: in $A=\begin{bmatrix}5&-1&0\\2&7&3\end{bmatrix}$, of order $2\times3$, we have $a_{12}=-1$ and $a_{21}=2$. A matrix is not a number — it has no single "value" — but a square matrix will have one special number attached to it, its determinant, in Section 3.2.

📘 Definition 3.2 — Types of matrices
  • A row matrix has one row ($1\times n$); a column matrix has one column ($m\times1$).
  • The zero matrix $O$ has every entry $0$ (of any order).
  • A square matrix has $m=n$; its entries $a_{11},a_{22},\dots,a_{nn}$ form the principal diagonal.
  • A square matrix is diagonal if $a_{ij}=0$ whenever $i\ne j$; scalar if it is diagonal with all diagonal entries equal ($a_{ii}=k$); the identity matrix $I_n$ (or $I$) is the scalar matrix with $k=1$.
  • A square matrix is upper triangular if $a_{ij}=0$ whenever $i\gt j$ (everything below the diagonal is $0$) and lower triangular if $a_{ij}=0$ whenever $i\lt j$.

The types are nested: identity $\Rightarrow$ scalar $\Rightarrow$ diagonal $\Rightarrow$ both upper and lower triangular. Conversely, a matrix that is both upper and lower triangular is diagonal. Notice that "diagonal", "scalar", "identity" and "triangular" are only defined for square matrices, while a zero matrix can have any order.

$$\underbrace{\begin{bmatrix}3&0&0\\0&-1&0\\0&0&5\end{bmatrix}}_{\text{diagonal}}\qquad\underbrace{\begin{bmatrix}2&0\\0&2\end{bmatrix}}_{\text{scalar }=2I_2}\qquad\underbrace{\begin{bmatrix}1&4&-2\\0&3&7\\0&0&6\end{bmatrix}}_{\text{upper triangular}}$$
✏️ Example 1 — Building, comparing and counting matrices
(a) Construct the $2\times3$ matrix $A=[a_{ij}]$ with $a_{ij}=|2i-3j|$. (b) Find $x$, $y$, $z$, $w$ if $\begin{bmatrix}x+y&2\\x-y&z+w\end{bmatrix}=\begin{bmatrix}7&z-w\\1&8\end{bmatrix}$. (c) How many different orders can a matrix with $18$ entries have? (d) How many $2\times2$ matrices with every entry from $\{0,1,2\}$ are there? How many of them are diagonal, and how many are upper triangular?
(a) Row $1$: $|2-3|=1$, $|2-6|=4$, $|2-9|=7$. Row $2$: $|4-3|=1$, $|4-6|=2$, $|4-9|=5$. So $A=\begin{bmatrix}1&4&7\\1&2&5\end{bmatrix}$.
(b) Equal matrices have equal entries: $x+y=7$, $x-y=1$, $z-w=2$, $z+w=8$. Adding and subtracting: $x=4$, $y=3$, $z=5$, $w=3$.
(c) An order $m\times n$ needs $mn=18$, and $m$ can be any divisor of $18$: $1,2,3,6,9,18$. So there are $6$ orders: $1\times18$, $2\times9$, $3\times6$, $6\times3$, $9\times2$, $18\times1$.
(d) All: each of the $4$ entries has $3$ choices, $3^4=81$. Diagonal: the two off-diagonal entries must be $0$, the two diagonal entries are free: $3^2=9$. Upper triangular: only $a_{21}=0$ is forced, so $3^3=27$.
Check In (b), substitute back: $\begin{bmatrix}7&2\\1&8\end{bmatrix}$ on both sides ✓. In (c), $2\times9$ and $9\times2$ are different orders — a matrix with $2$ rows is not a matrix with $9$ rows.

Which one of these statements is true?

Addition, Scalar Multiples and the Product

📘 Definition 3.3 — Sum, difference and scalar multiple

If $A=[a_{ij}]$ and $B=[b_{ij}]$ have the same order, their sum is $A+B=[a_{ij}+b_{ij}]$ — add entry by entry. For a number (scalar) $k$, the scalar multiple is $kA=[k\,a_{ij}]$ — multiply every entry by $k$. Then $-A=(-1)A$ and $A-B=A+(-B)$. Matrices of different orders cannot be added.

📐 Theorem 3.1 — Rules for addition and scalar multiples

For matrices $A$, $B$, $C$ of the same order and scalars $k$, $l$:

Commutative, associative$A+B=B+A$
$(A+B)+C=A+(B+C)$
Zero and negatives$A+O=A,\quad A+(-A)=O$
Distributive$k(A+B)=kA+kB$
$(k+l)A=kA+lA$
Scalars$k(lA)=(kl)A,\quad 1A=A,\quad 0A=O$

Proof. Each rule is the corresponding rule for real numbers applied in every position: for example the $(i,j)$ entry of $k(A+B)$ is $k(a_{ij}+b_{ij})=ka_{ij}+kb_{ij}$, which is the $(i,j)$ entry of $kA+kB$. $\blacksquare$ So matrix equations of the form $2X+A=B$ are solved exactly like number equations: $X=\tfrac12(B-A)$.

Multiplication is different, and it is the operation of the bakery: each outlet's takings come from a row of counts times the column of prices. We use the summation sign: $\sum_{k=1}^{n}a_kb_k$ means $a_1b_1+a_2b_2+\dots+a_nb_n$.

📘 Definition 3.4 — Product of two matrices

Let $A=[a_{ik}]$ be $m\times n$ and $B=[b_{kj}]$ be $n\times p$ — the number of columns of $A$ equals the number of rows of $B$. Then $AB$ is the $m\times p$ matrix whose $(i,j)$ entry is row $i$ of $A$ times column $j$ of $B$:

$$(AB)_{ij}=a_{i1}b_{1j}+a_{i2}b_{2j}+\dots+a_{in}b_{nj}=\sum_{k=1}^{n}a_{ik}b_{kj}.$$

If the inner numbers differ, $AB$ is not defined. Memory aid: $(m\times\underline{n})(\underline{n}\times p)=m\times p$ — the inner numbers must match and the outer numbers give the order.

For the bakery, with $S$ the $2\times3$ sales matrix and $P$ the $3\times1$ price column,

$$SP=\begin{bmatrix}40&25&60\\35&30&50\end{bmatrix}\begin{bmatrix}30\\120\\10\end{bmatrix}=\begin{bmatrix}4800\\5150\end{bmatrix},$$

a $(2\times3)(3\times1)=2\times1$ column of takings. The product $PS$ is not defined: $P$ has $1$ column and $S$ has $2$ rows.

📐 Theorem 3.2 — Rules for multiplication (and the rules that fail)

Whenever the products and sums below are defined, and $k$ is a scalar:

Associative$(AB)C=A(BC)$
Distributive$A(B+C)=AB+AC$
$(A+B)C=AC+BC$
Scalars and identity$k(AB)=(kA)B=A(kB)$
$I_mA=A=AI_n$ for $A$ of order $m\times n$
Zero$OA=O,\quad AO=O$

But, in general: (i) $AB$ may be defined while $BA$ is not; (ii) even for square matrices $AB\ne BA$; (iii) $AB=O$ is possible with $A\ne O$ and $B\ne O$; (iv) $AB=AC$ with $A\ne O$ does not imply $B=C$.

Proof of associativity. Let $A$, $B$, $C$ be $m\times n$, $n\times p$, $p\times q$. The $(i,j)$ entry of $(AB)C$ is $\sum_{l}\big(\sum_{k}a_{ik}b_{kl}\big)c_{lj}=\sum_k\sum_l a_{ik}b_{kl}c_{lj}$, and the $(i,j)$ entry of $A(BC)$ is $\sum_k a_{ik}\big(\sum_l b_{kl}c_{lj}\big)$ — the same double sum of all products $a_{ik}b_{kl}c_{lj}$, added in a different order. The distributive laws follow in the same way from $a(b+c)=ab+ac$. $\blacksquare$ Because of associativity we may write $ABC$ without brackets — but we may not change the order of the factors. Two square matrices with $AB=BA$ are said to commute; for them, and only for them, $(A+B)^2=A^2+2AB+B^2$, since in general

$$(A+B)^2=A^2+AB+BA+B^2,\qquad (A+B)(A-B)=A^2-AB+BA-B^2.$$
✏️ Example 2 — Multiplying, and what can go wrong
(a) For $A=\begin{bmatrix}2&-1\\1&3\end{bmatrix}$ and $B=\begin{bmatrix}1&4\\0&-2\end{bmatrix}$ find $AB$ and $BA$. (b) For $P=\begin{bmatrix}1&1\\2&2\end{bmatrix}$ and $Q=\begin{bmatrix}1&-1\\-1&1\end{bmatrix}$ find $PQ$ and $QP$. (c) Find $CD$ for $C=\begin{bmatrix}1&0&2\\-1&3&1\end{bmatrix}$ and $D=\begin{bmatrix}3\\-1\\2\end{bmatrix}$. Is $DC$ defined?
(a) Row times column: $(AB)_{11}=2\cdot1+(-1)\cdot0=2$, $(AB)_{12}=2\cdot4+(-1)(-2)=10$, $(AB)_{21}=1\cdot1+3\cdot0=1$, $(AB)_{22}=1\cdot4+3(-2)=-2$:
$$AB=\begin{bmatrix}2&10\\1&-2\end{bmatrix},\qquad BA=\begin{bmatrix}1\cdot2+4\cdot1&1(-1)+4\cdot3\\0\cdot2+(-2)1&0(-1)+(-2)3\end{bmatrix}=\begin{bmatrix}6&11\\-2&-6\end{bmatrix}.$$
Not commutative $AB\ne BA$: not a single entry agrees.
(b) $PQ=\begin{bmatrix}1-1&-1+1\\2-2&-2+2\end{bmatrix}=O$, although neither $P$ nor $Q$ is $O$. And $QP=\begin{bmatrix}1-2&1-2\\-1+2&-1+2\end{bmatrix}=\begin{bmatrix}-1&-1\\1&1\end{bmatrix}\ne O$. So also $PQ=PO$ without $Q=O$: no cancelling.
(c) $(2\times3)(3\times1)=2\times1$: $CD=\begin{bmatrix}3+0+4\\-3-3+2\end{bmatrix}=\begin{bmatrix}7\\-4\end{bmatrix}$. $DC$ would be $(3\times1)(2\times3)$: the inner numbers $1$ and $2$ differ, so $DC$ is not defined.
Check Sizes first, always: in (a) both products are $2\times2$, in (c) only one exists.

$A$ is $2\times3$ and $B$ is $3\times2$. Computing $AB$ entry by entry takes a certain number of multiplications of numbers; so does computing $BA$. How many in each case?

Choose (2×3)(3×2) and let Play run: $AB$ is $2\times2$, each of its $4$ entries is a row of length $3$ times a column of length $3$ — $3$ multiplications — and the readout ends at "$12$ of $12$". Press swap order: now the lab computes $BA$, which is $3\times3$; each of its $9$ entries needs $2$ multiplications, and the readout ends at "$18$ of $18$". In general an $(m\times n)(n\times p)$ product needs $mnp$ multiplications. $AB$ and $BA$ are not even the same size, let alone equal.
Interactive 1 — Row-by-Column Lab

Choose the orders of the two factors. Press Play and the lab builds the product one multiplication at a time: the row of the first factor slides along the column of the second, the products are added, and the result lands in the cell where that row and column cross. swap order multiplies the other way round — sometimes it gives a different matrix, sometimes a different size, sometimes nothing at all. 🎲 gives new entries.

order of product—
current entry—
multiplications—
the other order—

$A$ is a $3\times2$ matrix and $B$ is a $2\times4$ matrix. Which statement is correct?

Transpose, Symmetric and Skew-symmetric Matrices

📘 Definition 3.5 — Transpose

The transpose of an $m\times n$ matrix $A=[a_{ij}]$ is the $n\times m$ matrix $A^{T}$ (also written $A'$) obtained by turning rows into columns: $(A^{T})_{ij}=a_{ji}$. For example $\begin{bmatrix}1&2&3\\4&5&6\end{bmatrix}^{T}=\begin{bmatrix}1&4\\2&5\\3&6\end{bmatrix}$.

📐 Theorem 3.3 — Properties of the transpose

Whenever the operations are defined:

$$(A^{T})^{T}=A,\qquad (A\pm B)^{T}=A^{T}\pm B^{T},\qquad (kA)^{T}=kA^{T},$$ $$(AB)^{T}=B^{T}A^{T}\quad\text{(the order reverses)},\qquad (ABC)^{T}=C^{T}B^{T}A^{T}.$$

Proof of the reversal law. Let $A$ be $m\times n$ and $B$ be $n\times p$. First the orders: $(AB)^{T}$ is $p\times m$, and $B^{T}A^{T}$ is $(p\times n)(n\times m)=p\times m$ — while $A^{T}B^{T}$ need not even be defined. Now the entries:

$$\big((AB)^{T}\big)_{ij}=(AB)_{ji}=\sum_{k}a_{jk}b_{ki}=\sum_k (B^{T})_{ik}(A^{T})_{kj}=(B^{T}A^{T})_{ij}.$$

For three factors, $(ABC)^{T}=\big((AB)C\big)^{T}=C^{T}(AB)^{T}=C^{T}B^{T}A^{T}$. $\blacksquare$ Think of putting on socks and then shoes: to undo it you take off the shoes first. The same reversal will appear for inverses in Section 3.3.

📘 Definition 3.6 — Symmetric and skew-symmetric matrices

A square matrix $A$ is symmetric if $A^{T}=A$, i.e. $a_{ij}=a_{ji}$ for all $i$, $j$ (it is a mirror image of itself in the principal diagonal), and skew-symmetric if $A^{T}=-A$, i.e. $a_{ij}=-a_{ji}$ for all $i$, $j$.

📐 Theorem 3.4 — Facts about symmetric and skew-symmetric matrices
  1. Every diagonal entry of a skew-symmetric matrix is $0$.
  2. Every square matrix $A$ can be written in exactly one way as a symmetric matrix plus a skew-symmetric matrix: $$A=\underbrace{\tfrac12\big(A+A^{T}\big)}_{\text{symmetric}}+\underbrace{\tfrac12\big(A-A^{T}\big)}_{\text{skew-symmetric}}.$$
  3. For every matrix $A$ (any order), $AA^{T}$ and $A^{T}A$ are symmetric.
  4. If $A$ and $B$ are symmetric of the same order, then $AB+BA$ is symmetric, $AB-BA$ is skew-symmetric, and $AB$ is symmetric if and only if $AB=BA$.
  5. If $A$ is symmetric (skew-symmetric), so is $P^{T}AP$ for every square $P$ of the same order.

Proofs. (1) $a_{ii}=-a_{ii}$ gives $a_{ii}=0$. (2) $\big(A+A^{T}\big)^{T}=A^{T}+A$ and $\big(A-A^{T}\big)^{T}=A^{T}-A=-\big(A-A^{T}\big)$; the two halves add up to $A$. For uniqueness, suppose $A=S+K$ with $S$ symmetric and $K$ skew. Transposing, $A^{T}=S-K$; adding and subtracting the two equations gives $S=\tfrac12(A+A^{T})$ and $K=\tfrac12(A-A^{T})$ — there was no other choice. (3) $(AA^{T})^{T}=(A^{T})^{T}A^{T}=AA^{T}$ by the reversal law. (4) $(AB)^{T}=B^{T}A^{T}=BA$, so $(AB+BA)^{T}=BA+AB$ and $(AB-BA)^{T}=BA-AB$; and $AB$ is symmetric exactly when $AB=(AB)^{T}=BA$. (5) $(P^{T}AP)^{T}=P^{T}A^{T}P$, which is $\pm P^{T}AP$ when $A^{T}=\pm A$. $\blacksquare$

✏️ Example 3 — Splitting a matrix, and counting symmetric matrices
(a) Write $A=\begin{bmatrix}2&5&-1\\1&0&4\\3&-2&6\end{bmatrix}$ as the sum of a symmetric and a skew-symmetric matrix. (b) Write the $3\times3$ skew-symmetric matrix with $a_{12}=4$, $a_{13}=-1$, $a_{23}=2$. (c) How many $3\times3$ symmetric matrices have every entry from $\{0,1,2\}$? How many $3\times3$ skew-symmetric matrices have every entry from $\{-1,0,1\}$?
Find the number of symmetric matrices in (c):
(a) $A^{T}=\begin{bmatrix}2&1&3\\5&0&-2\\-1&4&6\end{bmatrix}$. Average and half-difference:
$$S=\tfrac12(A+A^{T})=\begin{bmatrix}2&3&1\\3&0&1\\1&1&6\end{bmatrix},$$ $$K=\tfrac12(A-A^{T})=\begin{bmatrix}0&2&-2\\-2&0&3\\2&-3&0\end{bmatrix}.$$
Check (a) $S$ is a mirror image in its diagonal, $K$ has zero diagonal and $k_{ij}=-k_{ji}$, and $S+K$: $3+2=5$, $1-2=-1$, $3-2=1$, $1+3=4$, $1+2=3$, $1-3=-2$ — exactly $A$ ✓.
(b) The diagonal is $0$ and $a_{ji}=-a_{ij}$: $\begin{bmatrix}0&4&-1\\-4&0&2\\1&-2&0\end{bmatrix}$.
(c) A symmetric $3\times3$ matrix is fixed by its $3$ diagonal entries and the $3$ entries above the diagonal: $3^6=729$. A skew-symmetric one has zero diagonal and is fixed by the $3$ entries above it; each may be $-1$, $0$ or $1$ (and the entry below is then forced, and allowed, since $-1,0,1$ is closed under change of sign): $3^3=27$.
⏱ Shortcut An $n\times n$ symmetric matrix has $\tfrac{n(n+1)}{2}$ free entries, a skew-symmetric one $\tfrac{n(n-1)}{2}$. Same numbers as in the count of symmetric relations in Section 1.1 — a relation on an $n$-set is an $n\times n$ matrix of $0$s and $1$s.
✏️ Example 4 — The transpose of a product
(a) For $A=\begin{bmatrix}1&2\\0&-1\end{bmatrix}$ and $B=\begin{bmatrix}3&1\\-2&4\end{bmatrix}$, verify that $(AB)^{T}=B^{T}A^{T}$ and show that $(AB)^{T}\ne A^{T}B^{T}$. (b) Prove that if $A$ is skew-symmetric then $A^2$ is symmetric. (c) Prove that if $A$ is symmetric and $B$ is skew-symmetric (same order), then $AB-BA$ is symmetric.
(a) $AB=\begin{bmatrix}3-4&1+8\\0+2&0-4\end{bmatrix}=\begin{bmatrix}-1&9\\2&-4\end{bmatrix}$, so $(AB)^{T}=\begin{bmatrix}-1&2\\9&-4\end{bmatrix}$.
Reversed order $B^{T}A^{T}=\begin{bmatrix}3&-2\\1&4\end{bmatrix}\begin{bmatrix}1&0\\2&-1\end{bmatrix}=\begin{bmatrix}3-4&0+2\\1+8&0-4\end{bmatrix}=\begin{bmatrix}-1&2\\9&-4\end{bmatrix}$ ✓.
Same order fails $A^{T}B^{T}=\begin{bmatrix}1&0\\2&-1\end{bmatrix}\begin{bmatrix}3&-2\\1&4\end{bmatrix}=\begin{bmatrix}3&-2\\5&-8\end{bmatrix}\ne(AB)^{T}$.
(b) $(A^2)^{T}=(AA)^{T}=A^{T}A^{T}=(-A)(-A)=A^2$. $\blacksquare$ (Similarly $(A^3)^{T}=(-A)^3=-A^3$: odd powers of a skew-symmetric matrix are skew-symmetric.)
(c) $(AB-BA)^{T}=B^{T}A^{T}-A^{T}B^{T}=(-B)A-A(-B)=-BA+AB=AB-BA$. $\blacksquare$

$A$ and $B$ are symmetric matrices of the same order. Then $AB-BA$ is

Trace, Special Square Matrices and Powers

📘 Definition 3.7 — Trace, powers and special square matrices

For a square matrix $A$ of order $n$, the trace is the sum of its diagonal entries, $\operatorname{tr}A=a_{11}+a_{22}+\dots+a_{nn}$. Powers are repeated products: $A^0=I$, $A^1=A$, $A^{k+1}=A^kA$. A square matrix $A$ is

  • idempotent if $A^2=A$;
  • nilpotent if $A^k=O$ for some positive integer $k$ (the least such $k$ is its index);
  • involutory if $A^2=I$;
  • orthogonal if $AA^{T}=A^{T}A=I$.
📐 Theorem 3.5 — Properties of the trace

For square matrices $A$, $B$ of order $n$ and a scalar $k$:

$$\operatorname{tr}(A+B)=\operatorname{tr}A+\operatorname{tr}B,\quad \operatorname{tr}(kA)=k\operatorname{tr}A,\quad \operatorname{tr}A^{T}=\operatorname{tr}A,$$ $$\operatorname{tr}(AB)=\operatorname{tr}(BA),\qquad \operatorname{tr}\big(AA^{T}\big)=\sum_{i,j}a_{ij}^2\ \text{(the sum of the squares of all entries)}.$$

The last two also hold for rectangular $A$ ($m\times n$) and $B$ ($n\times m$).

Proof. The first three are immediate. For the product, $\operatorname{tr}(AB)=\sum_i(AB)_{ii}=\sum_i\sum_k a_{ik}b_{ki}$, and $\operatorname{tr}(BA)=\sum_k\sum_i b_{ki}a_{ik}$ — the same numbers added in a different order. Finally $(AA^{T})_{ii}=\sum_j a_{ij}a_{ij}=\sum_j a_{ij}^2$ (row $i$ times itself), and adding over $i$ gives every entry squared once. $\blacksquare$ A neat consequence: there are no square matrices with $AB-BA=I$, because the left side has trace $0$ and the right side has trace $n$.

📐 Theorem 3.6 — Powers, and powers of commuting matrices

For a square matrix $A$ and non-negative integers $p$, $q$: $A^pA^q=A^{p+q}$ and $(A^p)^q=A^{pq}$. If $AB=BA$, then $(AB)^n=A^nB^n$ and the binomial expansion holds:

$$(A+B)^n=A^n+\tbinom n1A^{n-1}B+\tbinom n2A^{n-2}B^2+\dots+B^n,\qquad \tbinom n2=\tfrac{n(n-1)}{2}.$$

In particular, if $N^2=O$ then $(I+N)^n=I+nN$; if $N^3=O$ then $(I+N)^n=I+nN+\tfrac{n(n-1)}{2}N^2$.

Why. Expanding $(A+B)^n$ gives a sum of products of $n$ factors, each $A$ or $B$. When $AB=BA$ the factors in each product can be shuffled into the form $A^{n-r}B^r$, and the number of products that shuffle to $A^{n-r}B^r$ is the number $\tbinom nr$ of ways of choosing which $r$ of the $n$ brackets supply a $B$ — exactly as for numbers (Chapter 5 studies these coefficients). $I$ commutes with every matrix, so the theorem always applies to $I+N$; and when $N^2=O$ every term from $N^2$ onwards vanishes. Without $AB=BA$ the expansion is false: already $(A+B)^2=A^2+AB+BA+B^2$. $\blacksquare$

✏️ Example 5 — Powers by spotting the pattern
(a) For $A=\begin{bmatrix}1&3\\0&1\end{bmatrix}$ find $A^n$, and find $n$ if $A^n=\begin{bmatrix}1&60\\0&1\end{bmatrix}$. (b) For $B=\begin{bmatrix}2&1\\0&2\end{bmatrix}$ find $B^n$ and hence $B^5$. (c) For $J=\begin{bmatrix}1&1\\1&1\end{bmatrix}$ show that $J^n=2^{n-1}J$ and find the sum of the entries of $J^{10}$.
Find the $(1,2)$ entry of $B^5$ in (b):
(a) $A=I+N$ with $N=\begin{bmatrix}0&3\\0&0\end{bmatrix}$ and $N^2=O$. By Theorem 3.6, $A^n=I+nN=\begin{bmatrix}1&3n\\0&1\end{bmatrix}$. Then $3n=60$ gives $n=20$.
(b) $B=2I+N$ with $N=\begin{bmatrix}0&1\\0&0\end{bmatrix}$, $N^2=O$. Only two terms survive: $B^n=(2I)^n+n(2I)^{n-1}N=2^nI+n2^{n-1}N=\begin{bmatrix}2^n&n2^{n-1}\\0&2^n\end{bmatrix}$. For $n=5$: $B^5=\begin{bmatrix}32&80\\0&32\end{bmatrix}$.
(c) $J^2=\begin{bmatrix}2&2\\2&2\end{bmatrix}=2J$. If $J^n=2^{n-1}J$ then $J^{n+1}=2^{n-1}J^2=2^{n-1}\cdot2J=2^nJ$, so the pattern holds for every $n\ge1$ (induction). $J^{10}=2^9J=512J$, whose four entries add up to $4\times512=2048$.
Check (b) with $n=2$: $B^2=\begin{bmatrix}4&4\\0&4\end{bmatrix}$ directly, and the formula gives $\begin{bmatrix}2^2&2\cdot2\\0&2^2\end{bmatrix}$ ✓.

$N=\begin{bmatrix}2&-4\\1&-2\end{bmatrix}$ has no zero entry. What is $N^2$?

Choose the preset nilpotent. At $n=1$ the unit square is flattened onto a segment of the line $y=\tfrac x2$; at $n=2$ the readouts show $A^2=\begin{bmatrix}0&0\\0&0\end{bmatrix}$, status "$A^n=O$", and the picture is a single point at the origin. By hand: $N^2=\begin{bmatrix}4-4&-8+8\\2-2&-4+4\end{bmatrix}=O$. $N$ sends every vector onto the line $y=\tfrac x2$, and sends every vector on that line to $0$. The first option squares each entry — that is not how matrices multiply.
Interactive 2 — Powers Lab: What does $A^n$ do?

A $2\times2$ matrix moves every point $(x,y)$ of the plane to $A\begin{bmatrix}x\\y\end{bmatrix}$. The dashed unit square with its letter F is moved to the shaded shape by $A^n$. Choose a matrix and the power $n$, or press Play to step through $n=0,1,2,\dots$ and watch the pattern: shears add up, rotations go round, idempotent and nilpotent matrices flatten, involutory ones undo themselves. The grid rescales when the picture grows.

1
$A^n$—
$\det$ (area factor)—
status—
✏️ Example 6 — Hard example: idempotents, a trace count and a period
(a) Let $A$ be idempotent ($A^2=A$). Prove that $(I+A)^n=I+(2^n-1)A$ for every positive integer $n$, and find $n$ if $(I+A)^n=I+127A$ (with $A\ne O$). (b) How many $2\times3$ matrices $A$ with every entry from $\{0,1,2\}$ satisfy $\operatorname{tr}\big(AA^{T}\big)=6$? (c) For $C=\begin{bmatrix}0&1\\-1&0\end{bmatrix}$ find $C^{2026}$.
Find the count in (b):
(a) Induction $n=1$: $I+A=I+(2-1)A$ ✓. If $(I+A)^n=I+(2^n-1)A$, then
$$(I+A)^{n+1}=\big(I+(2^n-1)A\big)(I+A)=I+A+(2^n-1)A+(2^n-1)A^2.$$
Use $A^2=A$ The right side is $I+\big(1+2(2^n-1)\big)A=I+(2^{n+1}-1)A$ ✓. So $(I+A)^n=I+(2^n-1)A$ for all $n$. Then $2^n-1=127$ gives $n=7$.
(b) By Theorem 3.5, $\operatorname{tr}(AA^{T})$ is the sum of the squares of the $6$ entries, each square being $0$, $1$ or $4$. A total of $6$ arises as $4+1+1$ (one entry $2$, two entries $1$, the rest $0$): $6\times\tbinom52=6\times10=60$ ways (place the $2$, then choose two of the other five places for the $1$s); or as $1+1+1+1+1+1$ (all six entries $1$): $1$ way. Total $61$.
(c) $C^2=\begin{bmatrix}-1&0\\0&-1\end{bmatrix}=-I$, so $C^4=I$ and the powers repeat with period $4$. Since $2026=4\cdot506+2$, $C^{2026}=C^2=-I$.
⏱ Shortcut For any idempotent $A$, every power $A^r$ ($r\ge1$) is $A$, so $(I+A)^n=I+\big[\tbinom n1+\tbinom n2+\dots+\tbinom nn\big]A=I+(2^n-1)A$ in one line. For periodic powers, reduce the exponent modulo the period.
⚠️ JEE Main traps

1. Order matters. $AB\ne BA$ in general, so $(A+B)^2\ne A^2+2AB+B^2$ and $(A+B)(A-B)\ne A^2-B^2$ unless $A$ and $B$ commute. Use the binomial theorem only for commuting matrices — $I$ and $kI$ commute with everything, which is why $(I+N)^n$ and $(2I+N)^n$ are safe.

2. No cancelling, no zero-product rule. $AB=O$ does not give $A=O$ or $B=O$, and $AB=AC$ does not give $B=C$ (Example 2). Cancelling is allowed only when $A$ has an inverse (Section 3.3).

3. The transpose reverses products. $(AB)^{T}=B^{T}A^{T}$, never $A^{T}B^{T}$.

4. $A^2$ is not "square every entry". $\begin{bmatrix}2&-4\\1&-2\end{bmatrix}^2=O$. Only for diagonal matrices do powers act entry by entry.

5. Counting symmetric and skew-symmetric matrices. Free entries: $\tfrac{n(n+1)}2$ for symmetric, $\tfrac{n(n-1)}2$ for skew-symmetric (diagonal forced to $0$). For skew-symmetric matrices the allowed set must contain $0$ and the negatives of its members: with entries from $\{0,1,2\}$, the only skew-symmetric matrix is $O$.

6. $kA$ scales every entry. So $\operatorname{tr}(kA)=k\operatorname{tr}A$ — but beware Section 3.2, where the determinant of $kA$ is $k^n|A|$.

$A$ is a square matrix with $A^2=A$. Then $(I-A)^2$ equals

Exercises

All matrices have real entries. $I$ is the identity and $O$ the zero matrix of the appropriate order.

I
Matrices and their types.
  • 1 Construct the $3\times2$ matrix $A=[a_{ij}]$ with $a_{ij}=i+2j$.
  • 2 How many different orders can a matrix with $24$ entries have?
  • 3 Find $a$, $b$, $c$, $d$ if $\begin{bmatrix}a-b&2a+c\\2a-b&3c+d\end{bmatrix}=\begin{bmatrix}-1&5\\0&13\end{bmatrix}$.
  • 4 Name every type from Definition 3.2 that fits: (a) $\begin{bmatrix}3&0&0\\0&3&0\\0&0&3\end{bmatrix}$ (b) $\begin{bmatrix}0&0\\0&0\end{bmatrix}$ (c) $\begin{bmatrix}1&2&3\end{bmatrix}$ (d) $\begin{bmatrix}1&0\\4&5\end{bmatrix}$.
  • 5 How many $3\times3$ diagonal matrices have their diagonal entries from $\{1,2,3,4,5\}$?
  • 6 How many $3\times3$ upper triangular matrices have every entry from $\{0,1\}$?
  • 7 True or false: (a) every identity matrix is a scalar matrix; (b) every scalar matrix is diagonal; (c) a matrix that is both upper and lower triangular is diagonal; (d) a square matrix whose diagonal entries are all $0$ is the zero matrix.
  • 8 Find $x$ and $y$ if $\begin{bmatrix}x^2&y\\3&x+y\end{bmatrix}=\begin{bmatrix}4&x\\3&-4\end{bmatrix}$.
1 $a_{11}=3$, $a_{12}=5$, $a_{21}=4$, $a_{22}=6$, $a_{31}=5$, $a_{32}=7$: $A=\begin{bmatrix}3&5\\4&6\\5&7\end{bmatrix}$.
2 One order for each divisor $m$ of $24$: $1,2,3,4,6,8,12,24$ — $8$ orders.
3 $a-b=-1$ and $2a-b=0$ give $a=1$, $b=2$; then $2+c=5$ gives $c=3$, and $9+d=13$ gives $d=4$.
4 (a) Scalar ($3I_3$), hence diagonal, upper and lower triangular (and symmetric). (b) Zero matrix; also square, diagonal, scalar ($k=0$) and triangular. (c) Row matrix. (d) Square and lower triangular.
5 Off-diagonal entries are $0$; each of the $3$ diagonal entries has $5$ choices: $5^3=125$.
6 The $3$ entries below the diagonal are $0$; the other $6$ are free: $2^6=64$.
7 (a) True. (b) True. (c) True. (d) False — $\begin{bmatrix}0&1\\0&0\end{bmatrix}$.
8 $x^2=4$, $y=x$ and $x+y=-4$. The last two give $2x=-4$, so $x=-2$, $y=-2$, and $x^2=4$ ✓. (Using only $x^2=4$ would wrongly allow $x=2$: every one of the four equations must hold.)
II
The algebra of matrices. In parts 1–5, $A=\begin{bmatrix}1&-2\\3&0\end{bmatrix}$, $B=\begin{bmatrix}2&1\\-1&4\end{bmatrix}$, $C=\begin{bmatrix}0&3\\1&-1\end{bmatrix}$.
  • 1 Find $2A-3B+C$.
  • 2 Find the matrix $X$ with $2X+A=B$.
  • 3 Find $AB$ and $BA$. Do $A$ and $B$ commute?
  • 4 Verify that $A(B+C)=AB+AC$.
  • 5 Show that $A^2-A+6I=O$.
  • 6 Find $\begin{bmatrix}1&2&3\end{bmatrix}\begin{bmatrix}4\\5\\6\end{bmatrix}$ and $\begin{bmatrix}4\\5\\6\end{bmatrix}\begin{bmatrix}1&2&3\end{bmatrix}$.
  • 7 Find all $x$ with $\begin{bmatrix}x&1\end{bmatrix}\begin{bmatrix}1&2\\3&0\end{bmatrix}\begin{bmatrix}x\\-2\end{bmatrix}=\begin{bmatrix}0\end{bmatrix}$.
  • 8 For $M=\begin{bmatrix}1&-1\\-1&1\end{bmatrix}$, show that $M^2=2M$ and find $k$ with $M^5=kM$.
  • 9 School P buys $120$ notebooks, $200$ pens and $30$ bags; school Q buys $80$ notebooks, $150$ pens and $45$ bags. A notebook costs $\text{₹}\,40$, a pen $\text{₹}\,10$ and a bag $\text{₹}\,350$. Write the bills as a matrix product and find each school's bill.
1 $2A-3B=\begin{bmatrix}2-6&-4-3\\6+3&0-12\end{bmatrix}=\begin{bmatrix}-4&-7\\9&-12\end{bmatrix}$; adding $C$: $\begin{bmatrix}-4&-4\\10&-13\end{bmatrix}$.
2 $X=\tfrac12(B-A)=\tfrac12\begin{bmatrix}1&3\\-4&4\end{bmatrix}=\begin{bmatrix}\tfrac12&\tfrac32\\-2&2\end{bmatrix}$.
3 $AB=\begin{bmatrix}2+2&1-8\\6&3\end{bmatrix}=\begin{bmatrix}4&-7\\6&3\end{bmatrix}$, $BA=\begin{bmatrix}2+3&-4\\-1+12&2\end{bmatrix}=\begin{bmatrix}5&-4\\11&2\end{bmatrix}$. They do not commute.
4 $B+C=\begin{bmatrix}2&4\\0&3\end{bmatrix}$, $A(B+C)=\begin{bmatrix}2&-2\\6&12\end{bmatrix}$. $AC=\begin{bmatrix}-2&5\\0&9\end{bmatrix}$ and $AB+AC=\begin{bmatrix}4-2&-7+5\\6+0&3+9\end{bmatrix}=\begin{bmatrix}2&-2\\6&12\end{bmatrix}$ ✓.
5 $A^2=\begin{bmatrix}1-6&-2\\3&-6\end{bmatrix}=\begin{bmatrix}-5&-2\\3&-6\end{bmatrix}$; $A^2-A=\begin{bmatrix}-6&0\\0&-6\end{bmatrix}=-6I$, so $A^2-A+6I=O$.
6 $(1\times3)(3\times1)$: $[4+10+18]=[32]$. $(3\times1)(1\times3)$: the $3\times3$ matrix $\begin{bmatrix}4&8&12\\5&10&15\\6&12&18\end{bmatrix}$.
7 $\begin{bmatrix}x&1\end{bmatrix}\begin{bmatrix}1&2\\3&0\end{bmatrix}=\begin{bmatrix}x+3&2x\end{bmatrix}$; times $\begin{bmatrix}x\\-2\end{bmatrix}$: $x^2+3x-4x=x^2-x$. So $x(x-1)=0$: $x=0$ or $x=1$.
8 $M^2=\begin{bmatrix}2&-2\\-2&2\end{bmatrix}=2M$. Then $M^n=2^{n-1}M$ (as for $J$ in Example 5), so $M^5=16M$: $k=16$.
9 $\begin{bmatrix}120&200&30\\80&150&45\end{bmatrix}\begin{bmatrix}40\\10\\350\end{bmatrix}=\begin{bmatrix}4800+2000+10500\\3200+1500+15750\end{bmatrix}=\begin{bmatrix}17300\\20450\end{bmatrix}$: P pays $\text{₹}\,17\,300$ and Q pays $\text{₹}\,20\,450$.
III
Transpose, symmetric and skew-symmetric matrices.
  • 1 For $A=\begin{bmatrix}1&2&-3\\0&4&5\end{bmatrix}$ find $AA^{T}$ and $A^{T}A$, check that both are symmetric, and find $\operatorname{tr}(AA^{T})$.
  • 2 Write $\begin{bmatrix}3&-1\\5&2\end{bmatrix}$ as the sum of a symmetric and a skew-symmetric matrix.
  • 3 $\begin{bmatrix}0&x&3\\-2&0&y\\z&5&0\end{bmatrix}$ is skew-symmetric. Find $x+y+z$.
  • 4 $\begin{bmatrix}2&a&3\\4&5&b\\c&-1&7\end{bmatrix}$ is symmetric. Find $a$, $b$, $c$.
  • 5 Prove that for every square matrix $A$, $A+A^{T}$ is symmetric and $A-A^{T}$ is skew-symmetric.
  • 6 Prove that if $A$ is skew-symmetric, then $A^3$ is skew-symmetric and $A^4$ is symmetric.
  • 7 Prove that if $A$ is symmetric, then $B^{T}AB$ is symmetric for every square $B$ of the same order.
  • 8 How many $3\times3$ matrices with every entry from $\{-1,0,1\}$ are (a) symmetric, (b) skew-symmetric?
  • 9 $A$ is the $3\times3$ skew-symmetric matrix with $a_{12}=2$, $a_{13}=-1$, $a_{23}=4$. Find $\operatorname{tr}(AA^{T})$.
1 Rows times rows: $AA^{T}=\begin{bmatrix}14&-7\\-7&41\end{bmatrix}$. Columns times columns: $A^{T}A=\begin{bmatrix}1&2&-3\\2&20&14\\-3&14&34\end{bmatrix}$. Both are mirror images in the diagonal, and both traces are $55$ — the sum of the squares of all six entries, $1+4+9+0+16+25$ (Theorem 3.5).
2 $S=\tfrac12(A+A^{T})=\begin{bmatrix}3&2\\2&2\end{bmatrix}$, $K=\tfrac12(A-A^{T})=\begin{bmatrix}0&-3\\3&0\end{bmatrix}$; $S+K=A$ ✓.
3 $a_{ij}=-a_{ji}$: $x=-(-2)=2$, $y=-5$, $z=-3$. So $x+y+z=-6$.
4 $a_{ij}=a_{ji}$: $a=4$, $b=-1$, $c=3$.
5 $(A+A^{T})^{T}=A^{T}+A$ and $(A-A^{T})^{T}=A^{T}-A=-(A-A^{T})$, by Theorem 3.3. $\blacksquare$
6 $(A^3)^{T}=(A^{T})^3=(-A)^3=-A^3$ and $(A^4)^{T}=(-A)^4=A^4$. $\blacksquare$
7 $(B^{T}AB)^{T}=B^{T}A^{T}(B^{T})^{T}=B^{T}AB$. $\blacksquare$
8 (a) $6$ free entries: $3^6=729$. (b) $3$ free entries above the zero diagonal: $3^3=27$.
9 $\operatorname{tr}(AA^{T})$ is the sum of the squares of all entries; the entries are $0$ on the diagonal and $\pm2$, $\pm1$, $\pm4$ off it, each value twice: $2(4+1+16)=42$.
IV
Traces, powers and special matrices (harder).
  • 1 For $A=\begin{bmatrix}1&0\\2&1\end{bmatrix}$ find $A^n$, and the $n$ for which $A^n=\begin{bmatrix}1&0\\50&1\end{bmatrix}$.
  • 2 Show that $A=\begin{bmatrix}3&-6\\1&-2\end{bmatrix}$ is idempotent and find $(I+A)^5$. Give its $(1,1)$ entry.
  • 3 Let $N=\begin{bmatrix}0&2&1\\0&0&3\\0&0&0\end{bmatrix}$. Show that $N$ is nilpotent of index $3$ and find the $(1,3)$ entry of $(I+N)^4$.
  • 4 Show that $A=\begin{bmatrix}4&3\\-5&-4\end{bmatrix}$ is involutory and find $A^{2027}$.
  • 5 For $R=\begin{bmatrix}\cos\theta&-\sin\theta\\\sin\theta&\cos\theta\end{bmatrix}$ show that $R^2$ is the same matrix with $2\theta$ in place of $\theta$. Assuming $R^n$ has $n\theta$ in place of $\theta$, find the least positive $n$ with $R^n=I$ when $\theta=40^\circ$.
  • 6 Prove that there are no $2\times2$ matrices $A$, $B$ with $AB-BA=I$.
  • 7 For $J=\begin{bmatrix}1&1\\1&1\end{bmatrix}$, find the sum of all the entries of $J+J^2+J^3+\dots+J^6$.
  • 8 How many $2\times2$ matrices with every entry from $\{0,1\}$ satisfy $A^2=O$?
1 $A=I+N$, $N=\begin{bmatrix}0&0\\2&0\end{bmatrix}$, $N^2=O$: $A^n=I+nN=\begin{bmatrix}1&0\\2n&1\end{bmatrix}$. $2n=50$ gives $n=25$.
2 $A^2=\begin{bmatrix}9-6&-18+12\\3-2&-6+4\end{bmatrix}=A$. By Example 6(a), $(I+A)^5=I+31A=\begin{bmatrix}94&-186\\31&-61\end{bmatrix}$; the $(1,1)$ entry is $94$.
3 $N^2=\begin{bmatrix}0&0&6\\0&0&0\\0&0&0\end{bmatrix}\ne O$ and $N^3=O$: index $3$. By Theorem 3.6, $(I+N)^4=I+4N+6N^2$, whose $(1,3)$ entry is $0+4\cdot1+6\cdot6=40$.
4 $A^2=\begin{bmatrix}16-15&12-12\\-20+20&-15+16\end{bmatrix}=I$. Odd powers equal $A$: $A^{2027}=(A^2)^{1013}A=A$.
5 $R^2=\begin{bmatrix}\cos^2\theta-\sin^2\theta&-2\sin\theta\cos\theta\\2\sin\theta\cos\theta&\cos^2\theta-\sin^2\theta\end{bmatrix}=\begin{bmatrix}\cos2\theta&-\sin2\theta\\\sin2\theta&\cos2\theta\end{bmatrix}$. $R^n=I$ needs $40n$ to be a multiple of $360$: least $n=9$.
6 $\operatorname{tr}(AB-BA)=\operatorname{tr}(AB)-\operatorname{tr}(BA)=0$ (Theorem 3.5), while $\operatorname{tr}I=2$. $\blacksquare$
7 $J^n=2^{n-1}J$ has entry sum $4\cdot2^{n-1}=2^{n+1}$. Total: $4+8+16+32+64+128=252$.
8 With $A=\begin{bmatrix}a&b\\c&d\end{bmatrix}$, $A^2=\begin{bmatrix}a^2+bc&b(a+d)\\c(a+d)&d^2+bc\end{bmatrix}$. All terms are $\ge0$, so $a^2+bc=0$ forces $a=0$ and $bc=0$, and likewise $d=0$. Then $(b,c)$ is $(0,0)$, $(1,0)$ or $(0,1)$: $3$ matrices.
🎯 Exam corner JEE Main · MCQJEE Main · Numerical

E1. JEE Main · MCQ If $A=\begin{bmatrix}2&-1\\3&-2\end{bmatrix}$, then $A^{51}$ equals (A) $I$ (B) $-A$ (C) $A$ (D) $2^{25}A$

Square it $A^2=\begin{bmatrix}4-3&-2+2\\6-6&-3+4\end{bmatrix}=I$: $A$ is involutory.
Reduce $A^{51}=(A^2)^{25}A=I^{25}A=A$.
Answer (C).
⏱ Shortcut For a $2\times2$ matrix with trace $0$ and determinant $-1$ (Section 3.2), $A^2=I$ always. Then odd powers are $A$ and even powers are $I$.

E2. JEE Main · Numerical Let $A=\begin{bmatrix}1&1&1\\0&1&1\\0&0&1\end{bmatrix}$. For a positive integer $n$, the $(1,3)$ entry of $A^n$ is $55$. Find $n$. Give the answer to the nearest integer.

Split $A=I+N$ with $N=\begin{bmatrix}0&1&1\\0&0&1\\0&0&0\end{bmatrix}$, $N^2=\begin{bmatrix}0&0&1\\0&0&0\\0&0&0\end{bmatrix}$, $N^3=O$.
Expand Theorem 3.6: $A^n=I+nN+\tfrac{n(n-1)}2N^2$. Its $(1,3)$ entry is $n\cdot1+\tfrac{n(n-1)}2\cdot1=\tfrac{n(n+1)}2$.
Answer $\tfrac{n(n+1)}2=55$ gives $n^2+n-110=0$, $(n-10)(n+11)=0$: $n=10$.
Check $n=2$: $A^2=\begin{bmatrix}1&2&3\\0&1&2\\0&0&1\end{bmatrix}$ and $\tfrac{2\cdot3}{2}=3$ ✓.

E3. JEE Main · MCQ $A$ is a symmetric and $B$ a skew-symmetric matrix, both of order $3$. Which of the following is skew-symmetric? (A) $AB+BA$ (B) $AB-BA$ (C) $A^2+B^2$ (D) $BAB$

Transpose each $(AB+BA)^{T}=B^{T}A^{T}+A^{T}B^{T}=-BA-AB=-(AB+BA)$: skew-symmetric.
The others $(AB-BA)^{T}=-BA+AB$ (symmetric); $(A^2+B^2)^{T}=A^2+(-B)^2$ (symmetric); $(BAB)^{T}=B^{T}A^{T}B^{T}=(-B)A(-B)=BAB$ (symmetric).
Answer (A).
⏱ Shortcut Count the skew factors: each $B$ contributes a factor $-1$ to the transpose. One $B$ per term ($AB$, $BA$) and a $+$ sign between terms that swap into each other gives "skew"; two $B$s give "symmetric".

E4. JEE Main · Numerical Find the number of $3\times3$ matrices $A$, with every entry from $\{-1,0,1\}$, for which $\operatorname{tr}\big(AA^{T}\big)=2$. Give the answer to the nearest integer.

Meaning $\operatorname{tr}(AA^{T})=\sum a_{ij}^2$ (Theorem 3.5). Each square is $0$ or $1$, so the condition says exactly two entries are non-zero.
Count Choose the two positions: $\tbinom92=36$; each non-zero entry is $1$ or $-1$: $2^2=4$.
Answer $36\times4=144$.
⏱ Shortcut Read $\operatorname{tr}(AA^{T})$ as "sum of squares of all entries" at once; never multiply the matrices out.

E5. JEE Main · MCQ $A$ and $B$ are square matrices of the same order with $AB=A$ and $BA=B$. Then $A^2+B^2$ equals (A) $O$ (B) $I$ (C) $AB$ (D) $A+B$

$A^2$ $A^2=A\cdot A=(AB)A=A(BA)=AB=A$, using associativity and the two given equations.
$B^2$ $B^2=(BA)B=B(AB)=BA=B$.
Answer $A^2+B^2=A+B$: (D).
⏱ Shortcut Test with $A=B=I$: $AB=A$ ✓, $BA=B$ ✓, and $A^2+B^2=2I$ — only (D) gives $2I$.

E6. JEE Main · Numerical Let $A=\begin{bmatrix}3&-1\\2&1\end{bmatrix}$ and suppose $A^2=\alpha A+\beta I$ for real numbers $\alpha$, $\beta$. Find $\alpha^2+\beta^2$. Give the answer to the nearest integer.

Square $A^2=\begin{bmatrix}9-2&-3-1\\6+2&-2+1\end{bmatrix}=\begin{bmatrix}7&-4\\8&-1\end{bmatrix}$.
Compare entries $\alpha A+\beta I=\begin{bmatrix}3\alpha+\beta&-\alpha\\2\alpha&\alpha+\beta\end{bmatrix}$. The $(1,2)$ entries give $\alpha=4$; then $12+\beta=7$ gives $\beta=-5$, and the other entries agree ($8$ and $-1$ ✓).
Answer $16+25=41$.
⏱ Shortcut Every $2\times2$ matrix satisfies $A^2=(\operatorname{tr}A)\,A-|A|\,I$ (expand $A^2$ to check; $|A|=ad-bc$ is its determinant, Section 3.2). Here $\operatorname{tr}A=4$ and $|A|=3+2=5$: $\alpha=4$, $\beta=-5$ at sight.
🔬 Research–Teaching NexusResearch-tutored · optional, about 45 minutes

Research question. Wildlife managers protect endangered species with limited money. For an animal whose life has three stages — young, juvenile, adult — is it better to protect the young or the adults? A matrix answers this.

Why it matters. Population ecologists describe a stage-structured population by a column $X_t$ (the numbers in each stage in year $t$) and a Leslie matrix $L$ of birth and survival rates, so that $X_{t+1}=LX_t$ and $X_t=L^tX_0$. Studies of sea turtles, whales and forest trees used exactly this model to show that protecting the stage with the highest "leverage" matters far more than protecting the most visible one.

What to do.

  1. Take $L=\begin{bmatrix}0&4&3\\0.5&0&0\\0&0.25&0\end{bmatrix}$: each juvenile produces $4$ young a year and each adult $3$; half the young survive to become juveniles; a quarter of the juveniles survive to become adults. Start from $X_0=\begin{bmatrix}100\\40\\20\end{bmatrix}$.
  2. In a spreadsheet compute $X_1=LX_0$, $X_2=LX_1$, …, $X_{12}$ (one matrix product per year). Record the total population and the ratio total$(t+1)$/total$(t)$.
  3. The ratio settles down. Estimate its limit $\lambda$ and the long-run proportions of young, juveniles and adults.
  4. Now spend a "conservation budget": raise one survival rate by $0.1$ (either $0.5\to0.6$ or $0.25\to0.35$) and repeat. Which change raises the long-run growth rate more?

Tools: A spreadsheet (MMULT does a matrix product) or a few lines of Python.

Hand in: Your table of $X_t$, the limiting ratio, the stage proportions, and a short recommendation to a wildlife agency based on your two experiments.

Going further. If you replace $L$ by $kL$, how do the populations change? Show that the long-run ratio $\lambda$ satisfies $\lambda^3=0.5\cdot4\,\lambda+0.5\cdot0.25\cdot3$, and check your estimate against it.

Instructor note: for the matrix given, $\lambda^3-2\lambda-0.375=0$ has the root $\lambda=1.5$ exactly (since $3.375-3-0.375=0$), so the population eventually grows by $50\%$ a year whatever the starting column — a first glimpse of eigenvalues, which lie beyond JEE Main but explain why the ratio settles. The survival experiment shows that a rate's leverage depends on the whole matrix, not on its size alone.

🧪 Self-check quiz

0/5 correct

Q1. $A$ is a $2\times3$ matrix and $B$ is a $3\times4$ matrix. The order of $AB$ is

Q2. For square matrices $A$, $B$ of the same order, $(A+B)^2=A^2+2AB+B^2$ holds exactly when

Q3. The number of $2\times2$ skew-symmetric matrices with every entry from $\{-2,-1,0,1,2\}$ is

Q4. For matrices $A$, $B$, $C$ for which $ABC$ is defined, $(ABC)^{T}$ equals

Q5. If $A=\begin{bmatrix}2&0&0\\0&-1&0\\0&0&3\end{bmatrix}$, then $A^3$ is