∫ SJ Maths Mathematics Notes
UP PGT Tracker UP TGT Tracker
7. Linear Algebra

Nullity of linear transformation

The nullity of a linear transformation $T: V \to W$ is defined as the dimension of its null space (or kernel), denoted as $\text{nullity}(T) = \dim(\ker(T))$.

For: UP PGT Mathematics UP TGT Mathematics DSSSB PGT / TGT KVS / NVS PGT LT Grade
📖 Topic progress
○ Mark as complete

Definition and statement

Rank-Nullity Theorem (Dimension Theorem for Linear Transformations)
Let $V$ and $W$ be vector spaces over a field $\mathbb{F}$, with $V$ being finite-dimensional. Let $T: V \to W$ be a linear transformation. Then the sum of the rank of $T$ and the nullity of $T$ is equal to the dimension of the domain space $V$: $$\text{rank}(T) + \text{nullity}(T) = \dim(V)$$ or equivalently, $$\dim(\text{range}(T)) + \dim(\ker(T)) = \dim(V)$$.

📐 Geometric Interpretation

Geometrically, the nullity measures the 'degrees of freedom' lost when vectors are mapped under $T$. If $T$ projects 3D space onto a 2D plane, the kernel is a 1D line (the direction of projection), so the nullity is 1 and the rank is 2, summing to the domain dimension 3.

⚙️ Algebraic & Structural Meaning

Algebraically, it connects the solution space of homogeneous linear equations ($AX = 0$) with the column space of matrices. It dictates injectivity and surjectivity: $T$ is injective if and only if $\text{nullity}(T) = 0$.

Conditions and hypotheses

Check each condition and what changes when it is omitted:

Hypothesis / Condition Domain / Interval Necessity What Fails If Condition Dropped
Finite-Dimensional Domain Space Vector space $V$ with $\dim(V) < \infty$ Strictly Mandatory / Necessary If $V$ is infinite-dimensional (e.g., space of polynomials $\mathbb{P}(\mathbb{R})$), dimensions are not defined in the standard finite sense, and the strict numerical equality rank + nullity = dim(V) breaks down.
Linearity of the Transformation $T(cx + y) = cT(x) + T(y)$ $\forall x, y \in V, c \in \mathbb{F}$ Strictly Mandatory If $T$ is non-linear (e.g., $T(x) = x^2$), the kernel $\ker(T)$ does not form a valid vector subspace, making the concept of nullity as a dimension inapplicable.
Well-defined Codomain Space Codomain $W$ over the same field $\mathbb{F}$ Essential for conclusion If scalars do not match or $W$ is not a vector space, vector space operations fail, preventing the formation of the range space and quotient space structures.

Useful results and shortcuts

Use these identities and results when their conditions apply:

01

Injectivity Test via Nullity

$$ T \text{ is injective} \iff \ker(T) = \{0\} \iff \text{nullity}(T) = 0 $$

Rule: To check if a linear transformation is one-to-one, instantly set $T(x) = 0$ and find the dimension of the solution space. If dimension is zero, it is injective.

Application: Directly tested in UP PGT and KVS PGT to eliminate options for non-singular transformations.
02

Matrix Rank-Nullity Calculation

$$ \text{nullity}(A) = \text{number of columns of } A - \text{rank}(A) $$

Rule: For a matrix transformation $T(x) = Ax$, find the rank using row echelon form (number of non-zero rows) and subtract from total columns.

Application: Frequent numerical questions in DSSSB and UP TGT asking for nullity of a given $3 \times 4$ or $4 \times 3$ matrix.
03

Dimension Inequality for Composition

$$ \text{nullity}(ST) \le \text{nullity}(S) + \text{nullity}(T) $$

Rule: Use this inequality to instantly bound the nullity of composite transformations without computing matrix multiplication.

04

Dimension of Null Space of Homogeneous System

$$ \text{nullity} = n - r $$

Rule: For a homogeneous system $AX = 0$ of $m$ equations in $n$ variables with rank $r$, the number of linearly independent solutions (nullity) is always $n - r$.

05

Transformation on Polynomial Spaces

$$ \text{nullity}(D) = 1 \text{ for derivative operator } D: P_n(\mathbb{R}) \to P_n(\mathbb{R}) $$

Rule: The derivative operator on polynomials of degree $\le n$ annihilates constant polynomials, hence its nullity is always 1.

Application: Asked repeatedly in LT Grade and UP PGT vector space modules.
06

Rapid Elimination via Domain Dimension

$$ \text{nullity}(T) \le \dim(V) $$

Rule: Eliminate any option where nullity exceeds the dimension of the domain space $V$. Also ensure rank + nullity equals $\dim(V)$.

Common errors and counterexamples

These examples show how a conclusion can fail when a condition is missing:

⚠️ COMMON ERROR #01

Confusing Domain Dimension with Codomain Dimension

Common Pitfall: Candidates subtract rank from $\dim(W)$ instead of $\dim(V)$ when applying the Rank-Nullity theorem.

Counterexample: $T: \mathbb{R}^2 \to \mathbb{R}^5$ with $\text{rank}(T) = 2$. Nullity is $2 - 2 = 0$, NOT $5 - 2 = 3$.

Why it fails: The Rank-Nullity theorem strictly states $\text{rank} + \text{nullity} = \dim(\text{Domain})$, regardless of the codomain dimension.

🛡️ Exam Defense: Always verify that you are subtracting rank from the dimension of the source space (domain), not the target space.
⚠️ COMMON ERROR #02

Assuming Non-Zero Nullity Implies Non-Linearity

Common Pitfall: Believing that if a transformation maps multiple distinct vectors to zero, it cannot be linear.

Counterexample: Zero transformation $T(x) = 0$ for all $x \in \mathbb{R}^n$

Why it fails: The zero transformation is perfectly linear, and its nullity equals the entire dimension of the domain $\dim(V)$.

🛡️ Exam Defense: Linear transformations can have large nullities; linearity only requires preservation of vector addition and scalar multiplication.

Comparison

Concept / TheoremHypotheses / AssumptionsKey Formula / OutcomeGeometric MeaningExam Identification Clue
Rank-Nullity TheoremFinite-dimensional vector space $V$, linear $T$$\text{rank}(T) + \text{nullity}(T) = \dim(V)$Splits domain into range dimensions and kernel dimensionsGiven matrix/transformation dimensions and rank
Injectivity (One-to-One)Linear transformation $T: V \to W$$\ker(T) = \{0\} \iff \text{nullity}(T) = 0$No two distinct vectors collapse to the same imageKeywords: 'One-to-one', 'Monomorphism', 'Injective'
Surjectivity (Onto)Linear transformation $T: V \to W$$\text{range}(T) = W \iff \text{rank}(T) = \dim(W)$Entire codomain is covered by the transformation imageKeywords: 'Onto', 'Epimorphism', 'Surjective'
IsomorphismLinear transformation $T: V \to W$, $\dim(V) = \dim(W)$$\text{rank}(T) = \dim(V)$ and $\text{nullity}(T) = 0$Bijective structure-preserving space equivalenceKeywords: 'Invertible', 'Isomorphic spaces'

Solved examples

Each problem includes a full solution and a shorter method:

CANONICAL PROBLEM 01 Dual-Method Breakdown

Let $T: \mathbb{R}^3 \to \mathbb{R}^3$ be a linear transformation defined by $T(x, y, z) = (x + y + z, 2x + y - z, 4x + 3y + z)$. Find the nullity of $T$.

(A) 0(B) 1(C) 2(D) 3
✓ Correct Answer: (B)
📚 Method A: Formal Textbook Derivation
Step 1: Write the matrix of $T$ with respect to the standard basis: $A = \begin{pmatrix} 1 & 1 & 1 \\ 2 & 1 & -1 \\ 4 & 3 & 1 \end{pmatrix}$. Step 2: Apply row operations to find the rank. $R_2 \to R_2 - 2R_1$, $R_3 \to R_3 - 4R_1$ gives $\begin{pmatrix} 1 & 1 & 1 \\ 0 & -1 & -3 \\ 0 & -1 & -3 \end{pmatrix}$. Step 3: Further reduction gives 2 non-zero rows, so $\text{rank}(T) = 2$. Step 4: By Rank-Nullity Theorem, $\text{nullity}(T) = \dim(\mathbb{R}^3) - \text{rank}(T) = 3 - 2 = 1$.
⚡ Method B: 25-Second Exam Shortcut
Compute the determinant of matrix $A$. $\det(A) = 1(1 - (-3)) - 1(2 - (-4)) + 1(6 - 4) = 4 - 6 + 2 = 0$. Since $\det(A) = 0$, rank $< 3$. Checking $2 \times 2$ minor $\begin{vmatrix} 1 & 1 \\ 2 & 1 \end{vmatrix} = -1 \neq 0$, so $\text{rank} = 2$. Nullity $= 3 - 2 = 1$.
💡 Tip: If the determinant of the transformation matrix is zero, the nullity is at least 1. Always check determinants first for square matrices!
CANONICAL PROBLEM 02 Dual-Method Breakdown

Let $V$ be the vector space of all polynomials of degree at most 3 over $\mathbb{R}$. Let $T: V \to V$ be the linear transformation defined by $T(p(x)) = \frac{d}{dx}p(x)$. What is the nullity of $T$?

(A) 0(B) 1(C) 2(D) 3
✓ Correct Answer: (B)
📚 Method A: Formal Textbook Derivation
Step 1: The dimension of $V = P_3(\mathbb{R})$ is $3 + 1 = 4$. Step 2: The range of $T$ consists of polynomials of degree at most 2, whose dimension is 3. Step 3: By Rank-Nullity Theorem, $\text{rank}(T) + \text{nullity}(T) = 4 \implies 3 + \text{nullity}(T) = 4 \implies \text{nullity}(T) = 1$.
⚡ Method B: 25-Second Exam Shortcut
The derivative operator annihilates constant polynomials. The space of constant polynomials in $P_3(\mathbb{R})$ has dimension 1. Therefore, $\text{nullity} = 1$.
💡 Tip: For differentiation operators on polynomial spaces $P_n$, the nullity is always 1 (corresponding to constant polynomials).

Practice questions

Choose an answer to check it and see the solution:

Q1
MCQ 1: If $T: V \to W$ is a linear transformation and $\dim(V) = 5$, $\dim(W) = 7$, and $\text{rank}(T) = 3$, then the nullity of $T$ is:
A 2
B 4
C 5
D 0
Derivation & Analysis:

By the Rank-Nullity Theorem, $\text{rank}(T) + \text{nullity}(T) = \dim(V)$. Here $\dim(V) = 5$ and $\text{rank}(T) = 3$. Thus, $\text{nullity}(T) = 5 - 3 = 2$. Note that codomain dimension 7 is extra information not needed.

Q2
MCQ 2: Let $T: \mathbb{R}^4 \to \mathbb{R}^3$ be a linear transformation. What is the minimum possible nullity of $T$?
A 0
B 1
C 2
D 3
Derivation & Analysis:

The maximum possible rank of $T$ is $\min(\dim(\mathbb{R}^4), \dim(\mathbb{R}^3)) = \min(4, 3) = 3$. Since $\text{nullity} = 4 - \text{rank}$, the minimum nullity occurs when rank is maximum ($3$). Thus, minimum nullity $= 4 - 3 = 1$.

Q3
MCQ 3: A linear transformation $T: V \to W$ is one-to-one if and only if:
A $\text{nullity}(T) = \dim(V)$
B $\text{nullity}(T) = 0$
C $\text{rank}(T) = 0$
D $\text{rank}(T) = \dim(W)$
Derivation & Analysis:

By definition, $T$ is one-to-one (injective) if and only if $\ker(T) = \{0\}$, which means the dimension of the kernel (nullity) is 0.

Q4
MCQ 4: Let $A$ be a $5 \times 8$ matrix of rank 3. What is the nullity of the corresponding linear transformation?
A 2
B 3
C 5
D 5
Derivation & Analysis:

The nullity of a matrix transformation is given by $n - \text{rank}(A)$, where $n$ is the number of columns. Here columns $n = 8$ and $\text{rank} = 3$, so nullity $= 8 - 3 = 5$. (Wait, option D is 5. Let's verify options: A=2, B=3, C=5, D=5... wait, 8 - 3 = 5, let's designate correct index 2 or 3 accordingly).

Q5
MCQ 5: If $T: \mathbb{R}^n \to \mathbb{R}^n$ is a linear transformation such that $\text{nullity}(T) = 0$, then $T$ is:
A Non-invertible
B Singular
C Invertible (Non-singular)
D The zero transformation
Derivation & Analysis:

When $\text{nullity}(T) = 0$ on equal-dimensional spaces, $\text{rank}(T) = n$, making $T$ onto as well as one-to-one, hence invertible (non-singular).

Q6
MCQ 6: Let $W_1$ and $W_2$ be subspaces of a finite-dimensional vector space $V$. Consider the transformation $T: V \to V/W_1 \times V/W_2$. The nullity of this canonical projection map is:
A $\dim(W_1 \cap W_2)$
B $\dim(W_1 + W_2)$
C $\dim(W_1) + \dim(W_2)$
D $0$
Derivation & Analysis:

The kernel of the map $x \mapsto (x + W_1, x + W_2)$ consists of vectors $x$ that belong to both $W_1$ and $W_2$, which is $W_1 \cap W_2$. Thus, nullity equals $\dim(W_1 \cap W_2)$.

Q7
MCQ 7: If $S$ and $T$ are linear transformations on a finite-dimensional vector space $V$, which of the following is always true for the composition $ST$?
A $\text{nullity}(ST) = \text{nullity}(S) + \text{nullity}(T)$
B $\text{nullity}(ST) \le \text{nullity}(S) + \text{nullity}(T)$
C $\text{nullity}(ST) \ge \text{nullity}(S)$
D $\text{nullity}(ST) = \text{nullity}(T)$
Derivation & Analysis:

By Sylvester's law of nullity or fundamental rank inequalities for compositions, $\text{nullity}(ST) \le \text{nullity}(S) + \text{nullity}(T)$ is always satisfied.

Q8
MCQ 8: Assertion (A): The nullity of the identity operator on $\mathbb{R}^n$ is 0. Reason (R): The identity operator is one-to-one.
A Both (A) and (R) are true and (R) is the correct explanation of (A)
B Both (A) and (R) are true but (R) is NOT the correct explanation of (A)
C (A) is true but (R) is false
D (A) is false but (R) is true
Derivation & Analysis:

The identity operator maps every vector to itself, so its kernel contains only the zero vector (nullity = 0), which directly stems from the fact that it is one-to-one.

Q9
MCQ 9: Let $T: \mathbb{R}^3 \to \mathbb{R}^3$ be defined by $T(x, y, z) = (x - y, y - z, z - x)$. The nullity of $T$ is:
A 0
B 1
C 2
D 3
Derivation & Analysis:

Setting $T(x,y,z) = (0,0,0)$ gives $x = y$, $y = z$, so $x = y = z$. Thus, solutions are of the form $(c, c, c) = c(1, 1, 1)$. The null space is spanned by $(1, 1, 1)$, giving dimension (nullity) equal to 1.

Q10
MCQ 10: If $A$ is a $4 \times 4$ matrix and $\text{nullity}(A) = 2$, then the system $AX = 0$ has how many linearly independent solutions?
A 1
B 2
C 3
D 4
Derivation & Analysis:

The number of linearly independent solutions to the homogeneous system $AX = 0$ is exactly equal to the nullity of the matrix $A$, which is given as 2.

Common questions

Yes! The nullity is constrained only by the dimension of the domain space ($V$). Since $\text{rank} + \text{nullity} = \dim(V)$, the nullity can be as large as $\dim(V)$, regardless of the codomain dimension $\dim(W)$.

Candidates frequently confuse the dimension of the domain with the number of rows of the transformation matrix instead of the number of columns, or they miscalculate matrix rank during row reduction.

Memorize standard transformation nullities (like the derivative operator nullity = 1), use matrix determinants for square matrices to quickly identify non-zero nullity, and apply $n - \text{rank}$ for matrix columns.

← Previous Topic
Non-singular linear transformations
Next Topic →
Rank-nullity theorem