Orthogonality
Orthogonality is the relationship between vectors, directions, or states whose inner product is zero, corresponding to perpendicularity in Euclidean space.
Data Without Labels
Orthogonality is the relationship between vectors or directions whose inner product is zero, corresponding to a right angle in Euclidean space.
- How it is used
- Supports principal component analysis by separating successive components while retaining important dataset variation.
- How deep it goes
- Defines orthogonality geometrically and applies it to PCA selection, but does not derive inner-product or optimization machinery.
- Examples
- Uses house-price variables, vector-space projections, candidate lines, and successive principal components.
- What you can do after
- Evaluate/understand why PCA chooses orthogonal components to maximize combined captured variance.
Geometry for Programmers
Orthogonal describes vectors or directions that are perpendicular to each other.
- How it is used
- Uses perpendicularity to constrain smooth curve joins, organize coordinate bases, and construct plane normals.
- How deep it goes
- Explains right-angle intuition, the zero dot-product condition, basis implications, and a solved arc-line construction.
- Examples
- Demonstrates an arc-line join, coordinate-based dot-product equations, orthonormal coordinates, and a nonorthonormal ray-casting basis.
- What you can do after
- Build smooth arc-line joins and choose or recognize orthogonal bases and normal vectors in geometric algorithms.
Quantum Programming in Depth
An orthogonal state is a quantum state whose inner product with another state is zero, allowing the two states to be perfectly distinguished by a suitable measurement.
- How it is used
- Supports quantum state discrimination by identifying when measurement can distinguish promised input states perfectly.
- How deep it goes
- Explains perfect distinguishability, overlapping superpositions, and basis-changing measurements, but does not develop inner-product calculations in detail.
- Examples
- Uses |+⟩ and |−⟩, four three-qubit superpositions, and the four Bell states as concrete orthogonal-state examples.
- What you can do after
- Apply it via a suitable basis transformation to distinguish orthogonal quantum states with perfect accuracy.
Read how Quantum Programming in Depth explains orthogonality →
Math and Architectures of Deep Learning
Perpendicular in the geometric sense; here used for vectors or matrices whose directions are mutually independent or right-angled.
- How it is used
- Uses perpendicular directions to interpret vectors, rotations, eigenvectors, PCA, and dimensionality reduction.
- How deep it goes
- Develops the dot-product definition, proves the symmetric-matrix result, implements rotations, and applies orthogonality to diagonalization and PCA.
- Examples
- Uses document feature vectors, gravity and push directions, PyTorch rotations, a hyperellipse, PCA axes, and SVD.
- What you can do after
- Evaluate orthogonality with dot products and PyTorch, interpret rotations, and understand orthogonal directions in PCA and matrix decompositions.
Read how Math and Architectures of Deep Learning explains orthogonality →
What to know first
- transpositionTransposition is an operation that exchanges a matrix’s rows and columns or, more generally, permutes the axes of a multidimensional array.
Related concepts
- inner productmeasured by · An inner product is a scalar-valued operation on two vectors that generalizes the dot product, which for real coordinate vectors is the sum of products of corresponding components.
- measurement basispart of · A measurement basis is a pair of orthogonal quantum states that defines the possible outcomes of a quantum measurement.
- nonorthogonal statecontrasts with · A nonorthogonal state is a quantum state having a nonzero inner product with another state, so the two states cannot be perfectly distinguished by measurement.
- Bell stateincludes · A Bell state is one of four maximally entangled two-qubit quantum states whose measurements exhibit correlated outcomes.
- eigenvectorcontrasts with · An eigenvector is a nonzero vector that a linear transformation maps to a scalar multiple of itself.
- rotation matrixequivalent to · A rotation matrix is a square matrix representing a linear transformation that rotates vectors while preserving their lengths and angles.
Questions about orthogonality
What is orthogonality?
Orthogonality is the relationship between vectors, directions, or states whose inner product is zero, corresponding to perpendicularity in Euclidean space.
Which Manning books explain orthogonality?
Data Without Labels, Geometry for Programmers, Quantum Programming in Depth and Math and Architectures of Deep Learning.
What should I know before learning orthogonality?
Manning books that teach orthogonality usually introduce transposition first.
What is the difference between orthogonality and nonorthogonal state?
Orthogonality is the relationship between vectors, directions, or states whose inner product is zero, corresponding to perpendicularity in Euclidean space. A nonorthogonal state is a quantum state having a nonzero inner product with another state, so the two states cannot be perfectly distinguished by measurement.
What kinds of orthogonality are there?
Kinds of orthogonality include Bell state.
What is orthogonality part of?
Orthogonality is part of measurement basis. A measurement basis is a pair of orthogonal quantum states that defines the possible outcomes of a quantum measurement.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.



