Visual Tools
Calculators
Tables
Mathematical Keyboard
Converters
Other Tools

How to Check If a Set Is Orthogonal or Orthonormal

A set is orthogonal when every pair of different vectors has dot product zero. It is orthonormal when, in addition, every vector has length 11.

Recognize the task

A list of vectors or the columns of a matrix, and the question: orthogonal? orthonormal? an orthogonal basis? an orthogonal matrix? Forms: {(1,1,0),(1,−1,1),(−1,1,2)}\{(1, 1, 0), (1, -1, 1), (-1, 1, 2)\}, {(3,4),(−4,3)}\{(3, 4), (-4, 3)\}, Q=13(12221−22−21)Q = \frac{1}{3}\begin{pmatrix} 1 & 2 & 2 \\ 2 & 1 & -2 \\ 2 & -2 & 1 \end{pmatrix}. Test: kk vectors need k(k−1)2\frac{k(k-1)}{2} pair checks.
Orthogonal and orthonormal sets are defined in the orthogonal sets lesson; the pair test is the orthogonality condition.

Two ways to do it

Pairwise dot productsa few vectors
vi⋅vj=0for all i≠j\mathbf{v}_i \cdot \mathbf{v}_j = 0 \quad \text{for all } i \neq j
Check each pair once; one nonzero product ends the test.
Gram matrixcolumns of a matrix
QTQ=I  ⟺  the columns of Q are orthonormalQ^TQ = I \iff \text{the columns of } Q \text{ are orthonormal}
Put the vectors in the columns of QQ and compute QTQQ^TQ: zeros off the diagonal mean orthogonal, ones on it mean orthonormal (orthogonal matrix).

Steps

StepOn (1,1,0)(1, 1, 0), (1,−1,1)(1, -1, 1), (−1,1,2)(-1, 1, 2)
1List the pairs. Three vectors give three pairs; four vectors give six.(v1,v2)(\mathbf{v}_1, \mathbf{v}_2), (v1,v3)(\mathbf{v}_1, \mathbf{v}_3), (v2,v3)(\mathbf{v}_2, \mathbf{v}_3)
2Compute every dot product. Use the dot product. Stop at the first nonzero result: the set is not orthogonal.1−1+0=01 - 1 + 0 = 0, −1+1+0=0-1 + 1 + 0 = 0, −1−1+2=0-1 - 1 + 2 = 0
3Check the lengths. All pair products zero means orthogonal. For orthonormal, every vi⋅vi\mathbf{v}_i \cdot \mathbf{v}_i must equal 11 as well.∥v1∥2=2\|\mathbf{v}_1\|^2 = 2, ∥v2∥2=3\|\mathbf{v}_2\|^2 = 3, ∥v3∥2=6\|\mathbf{v}_3\|^2 = 6: orthogonal, not orthonormal
4Normalize if an orthonormal set is needed. Divide each vector by its length (normalization). Scaling does not change orthogonality.12(1,1,0)\frac{1}{\sqrt{2}}(1, 1, 0), 13(1,−1,1)\frac{1}{\sqrt{3}}(1, -1, 1), 16(−1,1,2)\frac{1}{\sqrt{6}}(-1, 1, 2)

Worked examples

Example 1 · two vectors in the plane

normalize
(3,4)⋅(−4,3)=−12+12=0(3, 4) \cdot (-4, 3) = -12 + 12 = 0
orthogonal
∥(3,4)∥=∥(−4,3)∥=5\|(3, 4)\| = \|(-4, 3)\| = 5
not unit length
{(35,45),(−45,35)}\{(\frac{3}{5}, \frac{4}{5}), (-\frac{4}{5}, \frac{3}{5})\} is orthonormal
divide both by 55

Example 2 · one pair fails

not orthogonal
(1,2,1)⋅(2,−1,0)=2−2+0=0(1, 2, 1) \cdot (2, -1, 0) = 2 - 2 + 0 = 0
first pair passes
(1,2,1)⋅(1,1,−1)=1+2−1=2(1, 2, 1) \cdot (1, 1, -1) = 1 + 2 - 1 = 2
second pair fails
not an orthogonal set
one nonzero product is enough

Example 3 · an orthogonal matrix

Gram matrix
Q=13(12221−22−21)Q = \frac{1}{3}\begin{pmatrix} 1 & 2 & 2 \\ 2 & 1 & -2 \\ 2 & -2 & 1 \end{pmatrix}
the columns are the vectors
column products: 2+2−4=02 + 2 - 4 = 0, 2−4+2=02 - 4 + 2 = 0, 4−2−2=04 - 2 - 2 = 0
off-diagonal entries of QTQQ^TQ
each column has length 33=1\frac{3}{3} = 1
diagonal entries of QTQQ^TQ
QTQ=IQ^TQ = I, so Q−1=QTQ^{-1} = Q^T
orthonormal columns
With an orthogonal basis, coordinates come free: ci=w⋅vivi⋅vic_i = \frac{\mathbf{w} \cdot \mathbf{v}_i}{\mathbf{v}_i \cdot \mathbf{v}_i} (coordinates via dot products). For w=(2,0,4)\mathbf{w} = (2, 0, 4) and the set above, c=(1,2,1)c = (1, 2, 1).

See it happen

Orthogonal vectors point along perpendicular directions, like the edges of a box meeting at a corner. The dot product measures how much one vector leans along another; zero lean in every pair means every pair meets at a right angle. Test a matrix in the orthogonal matrices visualizer.

Common mistakes

Checking only neighboring pairs

✗ checking v1⋅v2\mathbf{v}_1 \cdot \mathbf{v}_2 and v2⋅v3\mathbf{v}_2 \cdot \mathbf{v}_3 only
✓ check all three pairs, including v1⋅v3\mathbf{v}_1 \cdot \mathbf{v}_3

Calling an orthogonal set orthonormal

✗ {(3,4),(−4,3)}\{(3, 4), (-4, 3)\} is orthonormal
✓ it is orthogonal; divide by 55 to make it orthonormal

Orthogonal columns are not enough for an orthogonal matrix

✗ (3−443)\begin{pmatrix} 3 & -4 \\ 4 & 3 \end{pmatrix} is an orthogonal matrix
✓ QTQ=25I≠IQ^TQ = 25I \neq I; an orthogonal matrix needs orthonormal columns

Keeping the zero vector in a basis

✗ {(1,0),(0,0)}\{(1, 0), (0, 0)\} is an orthogonal basis of R2\mathbb{R}^2
✓ it is an orthogonal set, but a basis cannot contain the zero vector
Nonzero orthogonal vectors are always independent (orthogonal sets are independent).