Matrix Trace Calculator
Enter the entries of a 2 × 2 or 3 × 3 square matrix and instantly get its trace — the sum of the main diagonal elements a₁₁ + a₂₂ (+ a₃₃).
Find the Trace of a Square Matrix
Matrix Trace Formula
Where:
- \(a_{ii}\) — the entry of matrix \(A\) in row \(i\), column \(i\) (an element of the main diagonal)
- \(n\) — the size of the square matrix (number of rows = number of columns)
The trace of a square matrix is the sum of the entries on its main diagonal — the elements whose row index equals their column index. Only square matrices have a trace; a rectangular matrix has no well-defined main diagonal of equal length and width.
Matrix Trace – Calculation Example
Example: find the trace of the 3 × 3 matrix whose entries run from 1 to 9, row by row (these are the calculator's default values).
\[
A = \begin{pmatrix} 1 & 2 & 3 \\ 4 & 5 & 6 \\ 7 & 8 & 9 \end{pmatrix}
\]
\[
\operatorname{tr}(A) = a_{11} + a_{22} + a_{33} = 1 + 5 + 9 = 15
\]
Result: \(\operatorname{tr}(A) = 15\).
How to use the calculator
- Open the Matrix size dropdown and choose 2 × 2 or 3 × 3. The input grid adjusts automatically: for a 2 × 2 matrix the third row and third column are hidden.
- Fill in the fields of Matrix A (a₁₁, a₁₂, …) with the entries of your matrix, going row by row. Each label tells you the position: a₂₃ is row 2, column 3.
- Read the result: Trace (tr A) is the sum of the main diagonal entries, and Diagonal entries summed shows exactly which numbers were added together.
- Change any value and the result updates immediately — no button to press.
What is the trace of a matrix?
The trace is one of the simplest yet most important numbers attached to a square matrix. Despite looking like a minor summary, it captures deep structural information: the trace equals the sum of the matrix's eigenvalues, counting multiplicities. For a 2 × 2 matrix \(A\) with eigenvalues \(\lambda_1\) and \(\lambda_2\), you always have \(\operatorname{tr}(A) = \lambda_1 + \lambda_2\), while the determinant gives \(\det(A) = \lambda_1 \lambda_2\). This is why the trace appears constantly in the characteristic polynomial, stability analysis of dynamical systems, and quantum mechanics, where the trace of an operator gives the expected value of an observable.
The trace also has convenient algebraic properties that make it a workhorse in matrix proofs and computations:
- Linearity: \(\operatorname{tr}(A + B) = \operatorname{tr}(A) + \operatorname{tr}(B)\) and \(\operatorname{tr}(cA) = c\,\operatorname{tr}(A)\).
- Cyclic property: \(\operatorname{tr}(ABC) = \operatorname{tr}(BCA) = \operatorname{tr}(CAB)\), which implies \(\operatorname{tr}(AB) = \operatorname{tr}(BA)\) even when \(AB \neq BA\).
- Invariance under similarity: if \(B = P^{-1}AP\), then \(\operatorname{tr}(B) = \operatorname{tr}(A)\). The trace is the same in every basis.
Geometrically, if the matrix represents a linear transformation, the trace measures how the transformation scales volume locally along the diagonal direction: positive traces stretch, negative traces compress, and a trace of zero means the diagonal contributions cancel out. In graph theory, the trace of \(A^k\) counts closed walks of length \(k\) in the graph described by \(A\).
Assumptions and practical tips
This calculator works with 2 × 2 and 3 × 3 matrices and accepts any real numbers as entries, including negatives and decimals. The off-diagonal entries you type do not affect the result — the trace only depends on the main diagonal — but you still need to fill them in so the matrix is fully specified.
Keep in mind the following when interpreting your result:
- Square matrices only. If your data forms a rectangular matrix (say 2 × 3), the trace is not defined. Reconsider what your data represents or trim it to a square form.
- The trace is a single number, not a matrix. A trace of 0 does not mean the matrix is zero or trivial — many nonzero matrices (e.g. rotation matrices) have zero trace.
- Trace and determinant together fully determine the eigenvalues of any 2 × 2 matrix through the quadratic \(\lambda^2 - \operatorname{tr}(A)\,\lambda + \det(A) = 0\). Complex eigenvalues always come in conjugate pairs — see the Complex Conjugate calculator — so for a real 2 × 2 matrix with \(\operatorname{tr}(A)^2 < 4\det(A)\) the eigenvalues are complex.
Working with number grids in other contexts? The diagonals of a square also matter in puzzles — try the Magic Square generator, where every row, column, and diagonal must share the same sum. If your matrix work leads to solving systems of linear equations, the Elimination Method calculator can handle the next step.
[/wp_code]
Frequently Asked Questions
How do I find the trace of a matrix by hand?
Circle the entries that sit on the main diagonal — the ones running from the top-left corner to the bottom-right corner — and add them up. For a 3 × 3 matrix that means \(a_{11} + a_{22} + a_{33}\). No multiplication or sign rules are involved; it is a plain sum. Double-check that you took the top-left to bottom-right diagonal, not the anti-diagonal, which is a very common slip.
Can I use this calculator for a 4 × 4 or larger matrix?
Not directly — the tool supports 2 × 2 and 3 × 3 matrices. For a larger square matrix, the rule is exactly the same: add the entries \(a_{11} + a_{22} + a_{33} + a_{44} + \dots\). Since the trace only involves the diagonal, you can compute an \(n \times n\) trace by hand in seconds, or split the matrix into blocks whose traces add up, because the trace is linear.
Does the trace change if I transpose the matrix?
No. The transpose \(A^T\) swaps rows and columns, but every diagonal entry \(a_{ii}\) stays in place, so \(\operatorname{tr}(A^T) = \operatorname{tr}(A)\). The same holds for other similarity-based rearrangements: conjugating a matrix by an invertible \(P\), i.e. \(P^{-1}AP\), leaves the trace unchanged. This invariance is why the trace is considered an intrinsic property of a linear transformation, not of a particular matrix representation.
What is the trace of an identity matrix?
The trace of the \(n \times n\) identity matrix equals \(n\), because all diagonal entries are 1. For example, \(\operatorname{tr}(I_3) = 1 + 1 + 1 = 3\). More generally, \(\operatorname{tr}(cI) = cn\) for a scalar multiple of the identity. A useful related fact: \(\operatorname{tr}(A)\) equals the sum of eigenvalues, and the identity matrix has all eigenvalues equal to 1, which is consistent.
Is the trace the same as the sum of all matrix entries?
Usually not — this is the most common mistake. The trace only adds the diagonal entries; the off-diagonal entries are ignored. For the 3 × 3 matrix filled with 1 through 9, the sum of all nine entries is 45, but the trace is just \(1 + 5 + 9 = 15\). Only for a diagonal matrix (where all off-diagonal entries are zero) do the two coincide.
Can the trace be negative, zero, or a decimal?
Yes to all three. The trace is just a sum of real numbers, so any of your diagonal entries being negative or decimal flows straight into the result. A trace of zero is perfectly normal and does not mean the matrix itself is zero — for example, every 2 × 2 rotation matrix \(\begin{pmatrix} \cos\theta & -\sin\theta \\ \sin\theta & \cos\theta \end{pmatrix}\) has trace \(2\cos\theta\), which is zero for a 90° rotation.
How are the trace and determinant used together?
For any 2 × 2 matrix, the trace \(t\) and determinant \(d\) give you both eigenvalues at once, since they satisfy \(\lambda^2 - t\lambda + d = 0\). Solving: \(\lambda = \frac{t \pm \sqrt{t^2 - 4d}}{2}\). If \(t^2 < 4d\), the square root is negative and the eigenvalues form a complex conjugate pair. This pairing of trace and determinant is the backbone of 2 × 2 stability analysis in differential equations.