A Hilbert space is a real or complex inner product space that is complete in the norm . Completeness means that every Cauchy sequence has a limit in the space: if its terms become arbitrarily close in this norm, they converge to an allowed vector. The inner product supplies lengths and orthogonality; completeness ensures that limiting approximation procedures do not leave the space.
Finite-dimensional Euclidean spaces are examples, but the same geometry also works for sequences and functions. In the l2 sequence space,In the L2 space on with Lebesgue measure,These formulas use the convention linear in the first argument. Elements of the L2 space are equivalence classes of square-integrable functions agreeing almost everywhere, so zero distance really identifies the same element. A vector can therefore be a whole function, and its squared length can measure an integrated squared error.
For example, approximating by is a nearest-point problem in this Hilbert space. The allowed approximations form the linear span of , and the orthogonal projection is the constant . Its residual is orthogonal to both allowed directions because its integral and its inner product with vanish. The geometry therefore identifies the best approximation without searching through every pair of coefficients.
This is useful whenever a complicated vector must be represented by a tractable collection of directions. Fourier series approximate functions with orthogonal oscillations; the Parseval identity relates the sum of squared coefficients to the squared norm. Least-squares solutions approximate attainable data using a linear operator. The Hilbert projection theorem explains why a nearest element exists and is unique for a nonempty closed convex set. Distinct application settings share the same geometry of useful components and orthogonal residuals. Casey Rodriguez's MIT lectures develop these connections between Hilbert spaces, Fourier series, and nearest-point problems.[ref]
https://ocw.mit.edu/courses/18-102-introduction-to-functional-analysis-spring-2021/pages/lecture-notes-and-readings/. Rodriguez, MIT 18.102, lectures 14–17, 2021
A historical mathematical setting is David Hilbert's Grundzüge einer allgemeinen Theorie der linearen Integralgleichungen (1912). Its preface states that it reprints six communications published during 1904–1910 on linear integral equations. This provides a concrete source for the integral-equation setting associated with the name, while the abstract definition above incorporates later developments.[ref]
https://archive.org/details/grundzugeallg00hilbrich. Hilbert, Grundzüge einer allgemeinen Theorie der linearen Integralgleichungen, 1912, preface
For a planar domain , the Bergman space consists of holomorphic functions with finite area L2 norm, using the inherited inner product. It is a closed subspace of . Indeed the disk mean-value estimate bounds values locally by the L2 norm. An -Cauchy sequence of holomorphic functions consequently converges uniformly on compact subsets to a holomorphic function, agreeing almost everywhere with its limit.
On the unit disk, the normalized monomials , , form an orthonormal basis. Angular integration makes them orthogonal and gives . Applying the same integration to a Taylor expansion gives , proving completeness of this basis.
In the area Hilbert space of the unit disk, conjugates of Bergman space functions vanishing at zero are orthogonal to all holomorphic functions, since . This follows from the angular mean-value property and the Cauchy-Schwarz inequality, which makes the product integrable. But the radial function is also orthogonal to every holomorphic function: its radial integral against the constant angular mean is zero. It is not conjugate holomorphic, since it is real-valued and nonconstant. Thus the conjugate-holomorphic family is a proper subset of the orthogonal complement.
A closed subspace of a Hilbert space reduces a bounded operator when both the operator and its adjoint leave the subspace invariant. Equivalently, the orthogonal projection onto it commutes with the operator: the adjoint invariance makes its orthogonal complement invariant too. For a star-closed algebra of operators, the closed span of the orbit of any vector is reducing for the whole algebra. This enables an orthogonal direct-sum decomposition into cyclic actions when constructing Borel functional calculus for a normal operator.
The Hilbert tensor product is the completion of the algebraic tensor product of Hilbert spaces with , extended sesquilinearly. Orthonormal bases identify this inner product with the Euclidean inner product on coefficient arrays, proving positivity. In particular, and . These identities turn scalar power series into vector embeddings, as in the proof of the Grothendieck inequality.
The Hardy space of the circle is the closed subspace of L2 space whose negative Fourier coefficients vanish. The orthogonal projection onto it is denoted . A continuous symbol defines a Toeplitz operator . For continuous , is compact: it is finite rank for Laurent polynomials, and uniform approximation extends the result.
The Bargmann-Fock space is a Hilbert space of entire holomorphic functions with Gaussian-weighted area inner product. If , angular integration and the Gaussian integral give . Conversely any such square-summable coefficient sequence defines an entire function, since . This proves completeness and shows that is an orthonormal basis. The same bound shows boundedness of evaluation and gives reproducing kernel .
The closed unit ball of a Hilbert space is compact in norm exactly when the space is finite-dimensional. Finite-dimensional compactness follows from the Heine-Borel theorem. In infinite dimension, an orthonormal sequence has pairwise distance and no convergent subsequence. Restricting a compact operator that equals a nonzero multiple of the identity on a subspace therefore forces that subspace to be finite-dimensional.
A vector subspace closed in the norm topology of a Hilbert space is complete in the inherited norm and inner product. Its orthogonal complement gives a decomposition of the ambient Hilbert space into an orthogonal direct sum. This closedness permits orthogonal projection and ensures limits used in range arguments remain in the subspace. Every finite-dimensional vector subspace is closed.
A linear map is an isometry when it preserves inner products, equivalently . It need not be surjective. It maps an orthonormal basis to an orthonormal family, and preserves the nonzero eigenvalues and the Von Neumann entropy of a density operator.
A linear isometry from a subspace of a finite-dimensional Hilbert space into the same ambient space extends to a unitary operator. Complete an orthonormal basis of the input subspace and its isometric image to full orthonormal bases, then map the first to the second. The equal complement dimensions make this possible. Orthogonality of the specified image columns, not normalization alone, is essential. In an infinite-dimensional space arbitrary isometries need not be surjective; the finite-dimensional same-space hypothesis cannot be dropped.
An N-term approximation in an orthonormal basis uses at most basis functions. For a chosen index set , its optimal coefficients are the inner products with those functions, and its squared error is the sum of the omitted squared coefficients by the Parseval identity.
The best N-term approximation in an orthonormal basis retains coefficients of greatest absolute value. It minimizes the squared Hilbert space error over all index sets of size at most . The selection makes the approximation nonlinear in the data, even though coefficient extraction is linear.
For a bounded function on an interval with finitely many jumps and piecewise bounds, localized orthonormal wavelets with sufficient vanishing moments give squared best N-term approximation error . There are smooth-region coefficients of size , and only jump-crossing coefficients of size per level. Retain all coefficients through level and the jump-crossing coefficients through when . This costs and leaves squared tails . Choosing comparable to proves the rate by the Parseval identity.
A linear N-term approximation fixes its index set independently of the approximated function. Wavelet approximation normally orders by increasing resolution, whereas Fourier series approximation normally keeps the lowest frequencies. An arbitrary reordering can change or destroy a claimed convergence rate.
The Hilbert space completion of an inner product space is a Hilbert space containing an isometric dense copy of . It can be constructed from Cauchy sequences in , identifying two sequences when the norm of their difference tends to zero.
The space of finitely supported sequences with the l2 sequence space norm illustrates the need for completion. Let . For ,This is a Cauchy sequence, but its coordinatewise limit is not finitely supported. It belongs to the l2 sequence space, which is the completion. The same construction permits infinitely many small components while controlling their total squared size. An infinite orthonormal basis of a Hilbert space consequently permits norm-convergent series; a Hamel basis represents each vector by a finite sum, so these two uses of "basis" must be distinguished.
Start with finite real linear combinations of , and with arbitrary positive real frequencies, and use the inner product . Product-to-sum identities give existence of these averages and mutual orthogonality of distinct frequencies. The squared norm is . Its Hilbert space completion contains an uncountable orthonormal set, so is not a separable Hilbert space. This construction does not equip all locally square-integrable functions with an inner product: compactly supported nonzero functions have zero mean square, and other functions have divergent or nonexistent averages.
A Hilbert space is separable when it has a countable dense subset, equivalently a countable Hilbertian basis.
For every nonempty closed convex set in a Hilbert space and every in that space, there is a unique satisfyingClosedness and completeness let a limiting approximation attain the infimum; convexity prevents two different minimizers. The proof idea is to choose a sequence approaching the infimum. The parallelogram identity and the admissible midpoint make it a Cauchy sequence; applying the same identity to two putative minimizers proves uniqueness.
For a closed subspace of a Hilbert space, this nearest-point map is an orthogonal projection and is linear. For a general closed convex set it need not be linear: projecting the real line onto clips values below zero to zero and values above one to one. The theorem is therefore useful for both linear approximation and constrained nearest-point problems.[ref]
https://ocw.mit.edu/courses/18-102-introduction-to-functional-analysis-spring-2021/resources/18102-sp21-lecture-17/. Rodriguez, MIT 18.102, lecture 17: minimizers and orthogonal complements
Closedness is essential even in familiar examples. For the open interval and , the distance infimum is zero but no nearest point exists. The same obstruction appears when approximating by finitely supported sequences: the infimum is zero, but the limit lies outside that subspace until we take its Hilbert space completion.
For a closed subspace of a Hilbert space,The closest point gives the first component, and differentiating at zero shows that the remainder is orthogonal to every .
The orthogonal projection onto a closed subspace of a Hilbert space sends to its component in the orthogonal decompositionThe residual is orthogonal to every permitted direction. For any , the Pythagorean theorem givesThus uniquely minimizes the distance to . It is a linear map, is idempotent, and is nonexpansive: ; taking gives .
In , project onto the line spanned by . The inner product givesThe check confirms orthogonality, and . Projecting again changes nothing, since the result already lies on the line.
For a finite orthonormal basis of , the same operation is , using the convention linear in the first argument. Replacing coordinate vectors by functions gives exactly the same approximation principle. For in the real L2 space on and , the directions are orthogonal, with squared norms and . Their inner products with are and zero, soDirect integration gives the squared error . For any competing affine polynomial, the orthogonal residual yieldsEvery deviation from , adds a nonnegative error. The best function need not pass through any selected data point: it minimizes an integrated error across the whole interval.
Projecting x squared onto affine functions gives the constant one third and minimum integrated squared error 8/45
. Details. If idempotent matrices satisfy whenever , then . Thus the sum is idempotent. For a real orthonormal set , the matrices meet this condition. Their sum is the orthogonal projector onto their span; the symmetric matrix property follows by transposing the sum. Arbitrary sums of idempotent matrices need not be idempotent.
The foot from to the line , with , is its orthogonal projection onto that line. Solving gives the displayed formula and proves uniqueness. It belongs to the line segment with endpoints and exactly when .
For a nonzero closed subspace of a Hilbert space and a closed subspace of a Hilbert space , this angle measures the worst loss under projection from to . Its positive cosine is a uniform lower bound for . The order of the subspaces matters in general. This is distinct from the smallest angle between two subspaces, whose cosine is a supremum. For equal finite dimensions the cosine is the smallest singular value of the cross Gram matrix of orthonormal bases. When is zero, the infimum is over an empty unit sphere and does not define an angle in ; use a lower-bound formulation instead.
For closed subspaces of a Hilbert space, set . The two positive cosines bound and its adjoint operator below. The first bound gives a closed range and injectivity; the second makes the range dense by image-kernel orthogonality for an adjoint. Thus is a bounded bijection, and yields the unique direct sum decomposition . In equal finite dimensions, a lower bound on alone suffices by the rank-nullity theorem.
For two nonzero finite-dimensional subspaces with projections , define , with vectors drawn from the respective subspaces. The angle is zero when they intersect nontrivially. To measure separation after removing a common intersection, restrict both subspaces to its orthogonal complement. The resulting angle controls sums of positive operators in the Kitaev geometrical lemma. For a zero subspace, the vector supremum is empty; the projection-norm convention is still usable when explicitly adopted.
Let be positive operators on a finite-dimensional Hilbert space, with every nonzero eigenvalue at least . Let be the orthogonal projections onto and define . ThenWhen the kernels have no common nonzero vector, this gives a positive lower bound. The projection formula also covers a zero kernel without taking a supremum over an empty unit sphere.
Indeed, and . For , , . The map consequently has squared operator norm at most , so . Combining these inequalities proves the result. With common kernel , restrict to and use the kernel projections there to bound the positive spectral gap.
In infinite dimension, replace the eigenvalue condition by the actual operator inequalities and . An assertion only about eigenvalues can be vacuous in the presence of continuous spectrum, and does not ensure this uniform gap.
For orthogonal projections with , the identity and imply for . When the projections have a common fixed subspace, first remove its orthogonal projection. This estimate is stronger than generic submultiplicativity and is useful for alternating local projection layers.
For a real or complex Hilbert space with inner product linear in its first argument, every bounded linear functional has a unique representing vector such thatThe Cauchy-Schwarz inequality gives , and evaluation at gives equality when . Uniqueness follows because for every implies . Existence is proved by the Kernel-orthogonal proof of the Riesz representation theorem.[ref]
18.102 Introduction to Functional Analysis, Spring 2021 (lecture notes) by Andrew Lin. lecture18, Theorem185, bounded functional representation
For example, on , is represented by , with . By contrast, point evaluation is not a bounded functional on this L2 space: elements are equivalence classes modulo equality almost everywhere, so point values are not even well-defined. On continuous functions, triangular peaks can also have value one at zero and arbitrarily small L2 norm.
On a complex Hilbert space, is conjugate-linear. Thus the representation is an isometric conjugate-linear identification with the continuous dual space, under the declared convention; it should not silently be called a complex-linear map.
Let be a nonzero bounded linear functional on a real or complex Hilbert space whose inner product is linear in its first argument. Its kernel is a proper closed subspace of a Hilbert space. The orthogonal decomposition supplies a nonzero , and . For every ,Taking its inner product with gives . Therefore the correct representing vector isIn the real case the conjugation disappears. The Cauchy-Schwarz inequality and evaluation along prove equality of functional and vector norms. For , take . This proof makes both closedness of the kernel and the complex conjugate in the representing vector explicit.
For a bounded linear operator between real or complex Hilbert spaces, its adjoint is the unique bounded linear operator satisfyingThe inner product here is linear in its first argument. The adjoint operator transfers an output measurement back to the input: measuring against the vector gives exactly the same scalar as measuring against . It depends on both inner products.
In this real example,Both inner products equal . The dashed lines collect all vectors giving that same measurement: the output line becomes the input line . The adjoint operator therefore moves the direction that describes the measurement. It generally differs from the inverse matrix: here .
For an explicitly complex example, use the standard inner product on . In the standard orthonormal basis, the adjoint operator is represented by the conjugate transpose:For and , direct matrix multiplication givesand henceThe complex conjugate is essential to this agreement.
Existence follows from the Riesz representation theorem: for fixed , the bounded linear functional has a unique representing vector . The resulting adjoint operator has the same operator norm as , andMIT's lecture 18 develops the representation argument and the conjugate transpose example.[ref]
https://ocw.mit.edu/courses/18-102-introduction-to-functional-analysis-spring-2021/7f23bfc43ee2e1efd238a5384e5b9d51_MIT18_102s21_lec18.pdf. Rodriguez, notes by Andrew Lin, MIT 18.102, lecture 18, Theorem 185 and Example 186, 2021
This transfer explains the normal equation for a linear inverse problem: a least-squares solution has residual orthogonal to every attainable direction , so for every , equivalently . An output residual becomes an input condition through the same adjoint operator. All identities in this section concern bounded operators defined on the whole source Hilbert space; for a densely defined operator that is not bounded, use the adjoint of a densely defined operator and check its operator domain.
On the Hilbert space , the unilateral shift is the bounded linear operatorIt preserves the norm, so . Its adjoint operator shifts the other way: . Consequently , where is the orthogonal projection onto the first coordinate.
The range consists of sequences whose first coordinate is zero. Thus is an isometry but is not onto and is not unitary. Its adjoint has a one-dimensional kernel even though has zero kernel. These formulas give a concrete infinite-dimensional distinction: in finite dimensions an isometry from a space to itself is automatically onto, whereas the unilateral shift loses one output coordinate.
Give the real vector space of degree-at-most-one polynomials the inner product . In coordinates , , this is . Since , the adjoint identity requires for every . Comparing the two coefficients yields the displayed formula. The adjoint operator depends on both the inner product and the domain; a formal integration-by-parts sign without the boundary term would give the wrong answer.
On a complex inner product space , with the inner product conjugate-linear in its first argument, an everywhere-defined linear map admits an adjoint on if there is a linear map satisfyingThe adjoint, when it exists, is unique: the difference of two candidate values at is orthogonal to every vector and hence to itself. In a finite-dimensional vector space, it always exists, and its matrix in an orthonormal basis is the conjugate transpose. In a space that is not complete, existence must instead be checked.
The identity for every supplies such an adjoint directly, namely . This algebraic statement should be distinguished from self-adjointness of a linear operator on a completed Hilbert space, which additionally compares the domains of an operator and its Hilbert-space adjoint. The Gaussian polynomial operator is a useful example where the identity holds on the entire polynomial inner product space and can be used on every finite-dimensional invariant polynomial subspace.
Suppose a linear map between finite-dimensional Hilbert spaces has matrix , and their inner products in the chosen bases are and . The Gram matrices are positive-definite matrices. Writing for the ordinary conjugate transpose, the coordinate matrix of the adjoint operator isIndeed, the defining inner product identity says for all . Thus simply transposing and conjugating is valid in orthonormal bases, where both Gram matrices are the identity matrix.
For the weighted inner product and the coordinate swap , the adjoint operator has matrix . With and , both inner products equal . The coordinate swap is its own ordinary transpose, but is not self-adjoint in this weighted inner product.
For a densely defined operator on a Hilbert space, belongs to exactly when the functional is bounded in the ambient Hilbert norm on . Density makes its representing vector unique. Differential expressions alone do not determine this domain.
// Target: analysis.bigb
For bounded operators on a Hilbert space, moving products across the inner product givesThe order reversal and minus sign matter. If are self-adjoint, their commutator is skew-adjoint, and is self-adjoint.
For unbounded densely defined operators, the same manipulation gives a pairing identity only on vectors for which all the displayed compositions are defined. It does not by itself prove equality of the full adjoint operator domains. For example, a formally self-adjoint Dolbeault Laplacian and mutually formal-adjoint Lefschetz operators give the formal identity on suitable smooth test functions; operator realizations additionally require compatible domains or boundary conditions.
A densely defined operator on a Hilbert space is symmetric whenEquivalently, agrees with its adjoint operator on , while may be larger. A self-adjoint operator additionally requires equality of these operator domains.
Its adjoint is on all of , with no endpoint conditions. Every complex is an adjoint eigenvalue, with eigenfunction . Thus every complex obstructs surjectivity of . The original operator has no eigenvalues because its zero initial value kills any solution of .
// Target: analysis.bigb
For a linear map between finite-dimensional inner product spaces,The first equality follows directly from the defining identity for the adjoint operator; the second follows by taking orthogonal complements. In an infinite-dimensional Hilbert space, the second equality generally requires closure of the image.
A bounded operator on a Hilbert space is invertible exactly when both and are bounded below. A lower bound on makes it injective with closed range, while a lower bound on gives and hence dense range throughClosed dense range is the whole space, and the lower bound controls the inverse.
Hermitian conjugation is the operation of taking the adjoint operator, written . For a matrix in orthonormal bases, it takes the complex conjugate of each entry and exchanges rows with columns:For a real matrix it is the ordinary transpose. For the complex matrix in the adjoint operator example, its second row becomes the second column of . This is the same conjugate transpose used in MIT's treatment of complex inner products and Hermitian matrices.[ref]
https://ocw.mit.edu/courses/18-06sc-linear-algebra-fall-2011/53f17d47bf99d26fdc472b3177c256a8_MIT18_06SCF11_Ses3.2sum.pdf. Strang, MIT 18.06SC, lecture 26 summary, Complex matrices, pp. 1–2, 2011
For bounded operators with matching source and target Hilbert spaces, Hermitian conjugation satisfiesThe reversal of function composition is forced by moving both linear operators across an inner product:The calculation identifies the adjoint operator by the uniqueness theorem in Melrose's MIT notes.[ref]
https://ocw.mit.edu/courses/18-102-introduction-to-functional-analysis-spring-2021/3d4cc88026d44a01f936cd6a0aa995cb_MIT18_102s20_lec_FA.pdf. Melrose, MIT 18.102, Functional Analysis, Chapter 3, §11, 2020
For a computed order check, let and reuse . ThenKeeping the original order gives a different matrix. Likewise : taking the adjoint operator is conjugate-linear as an operation on linear operators, although each individual is a linear map on vectors.
The formal adjoint of a differential operator is characterized by moving between factors under an integral by integration by parts, while discarding boundary terms for compactly supported functions.
The formal adjoint of an infinite matrix does not specify the domain of the true adjoint. For the weighted difference generator , its formal positive-eigenvalue solution is . This is square summable but does not belong to the true adjoint domain of the maximal operator: testing against contradicts the adjoint identity. The nonzero limiting boundary term is the obstruction.
// Target: analysis.bigb
For smooth positive , let . Its unweighted bilinear transpose is the displayed operator. If satisfies , its transpose boundary condition is . At parameters for which the corresponding boundary-value inverses exist, their Green kernels satisfy . Equivalently , exhibiting weighted acoustic Green-function reciprocity.
For an order- operator with smooth coefficients on a bounded piecewise smooth domain, integration by parts has no boundary remainder when at each boundary point either the entire boundary jet of through order vanishes or the entire such jet of vanishes. Every boundary summand pairs derivatives of the two functions of order at most . It is insufficient to require only that, for each same multi-index , one of and vanish: for , and on a rectangle, all those same-index products vanish on the boundary but .
A bounded operator on a real or complex Hilbert space is Hermitian, or self-adjoint, when . Equivalently,Its directional values are real, since this identity makes equal its complex conjugate. Reality allows either sign; a positive operator additionally requires nonnegative values in every direction.
Self-adjoint directional values may change sign, touch zero, or stay strictly positive
. Details. The diagram compares Hermitian quadratic forms along paths of unit vectors. The Hermitian matrix takes the values and on the standard basis vectors; takes the values and ; andsatisfies and . These eigenvectors form an orthonormal basis, and every has . Thus is a positive operator, with eigenvalues and operator norm .
Three conditions describe different geometry. A Hermitian operator satisfies ; a normal operator satisfies ; a unitary operator satisfies , where is the identity operator. Every Hermitian operator and every unitary operator is a normal operator. Concrete matrices separate the conditions:
- The Hermitian matrix above is a normal operator and a positive operator, but not a unitary matrix, because .
- is a unitary matrix and normal matrix, but not a Hermitian matrix, since .
- is a normal matrix, but neither a Hermitian matrix nor a unitary matrix: its first standard basis vector doubles in norm.
- is not a normal matrix, since and .
The spectrum of a bounded Hermitian operator is real. In finite dimension, the spectral theorem supplies an orthonormal basis of eigenvectors, turning its action into independent real scalings. This makes the geometry easy to interpret: each eigenvalue records a signed response along an orthogonal direction. An arbitrary bounded Hermitian operator in infinite dimension need not have any eigenvectors; the general spectral theorem uses a projection-valued measure. MIT's lecture 21 develops the real spectrum and the orthogonality of eigenspaces.[ref]
https://ocw.mit.edu/courses/18-102-introduction-to-functional-analysis-spring-2021/345c4a5141559a4069aeb3105acd8ef3_MIT18_102s21_lec21.pdf. Rodriguez, notes by Andrew Lin, MIT 18.102, lecture 21, Theorems 223 and 227, 2021
A historical connection is the study of forms with conjugate variables. In his 1901 account of Charles Hermite's work, Camille Jordan describes these forms and emphasizes their role in reducing forms with real or complex coefficients. In today's notation, is the diagonal value of a Hermitian form; the conjugate variables make a real-valued expression possible even with complex coefficients.[ref]
https://www.numdam.org/item/JMPA_1901_5_7__91_0.pdf. Jordan, Notice sur M. Ch. Hermite, Journal de mathématiques pures et appliquées 7, pp. 92 and 95, 1901
For a densely defined operator outside this bounded setting, the displayed pairing identity alone defines a symmetric operator. A self-adjoint operator must also have the same operator domain as its adjoint operator.
A bounded Hermitian operator determines a Hermitian form and its real-valued diagonal expression . On a complex Hilbert space,Thus it is a quadratic form on the underlying real vector space; it is generally not a quadratic form over the complex field, whose scaling rule would involve . This distinction explains why replacing by loses the geometry of a complex inner product.
For a finite-dimensional Hermitian operator with an orthonormal basis of eigenvectors and corresponding eigenvalues , the spectral theorem givesFor a unit vector, this is an average of the eigenvalues with nonnegative weights summing to one. It is also the Rayleigh quotient, so its minimum and maximum are the smallest and largest eigenvalues.
In the Hermitian operator diagram, take for , and for . The computed curves areThese paths illustrate the extremal values; the spectral theorem supplies the bounds for every complex unit vector, including those outside the plotted paths.
The Hermitian part of a square matrix is half its sum with its conjugate transpose. It is a Hermitian matrix and satisfies . Its largest eigenvalue is the numerical abscissa, controlling instantaneous growth of the Euclidean norm under .
For a complex matrix , the real linear map on the -dimensional real vector space of Hermitian matrices has determinant . A real basis of Hermitian matrices is also a complex basis of all matrices, so the matrix of this restriction equals the matrix of the complex-linear congruence map. Left and right multiplication separately have determinants and .
A bounded Hermitian operator on a Hilbert space is positive, also called a positive semidefinite operator, whenThe value must be real as well as nonnegative. This permits zero values on nonzero vectors: the zero linear operator is positive. For a bounded Hermitian operator, positivity is equivalent to its spectrum being contained in ; in finite dimension, this is the condition that all eigenvalues are nonnegative.[ref]
https://ocw.mit.edu/courses/18-102-introduction-to-functional-analysis-spring-2021/345c4a5141559a4069aeb3105acd8ef3_MIT18_102s21_lec21.pdf. Rodriguez, notes by Andrew Lin, MIT 18.102, lecture 21, Corollary 225, 2021
The Hermitian operator diagram separates three cases. is self-adjoint but not positive, because . The orthogonal projection is positive semidefinite, with , but is not strictly positive on every nonzero vector. The complex Hermitian matrix is positive definite because whenever . For real symmetric matrices, MIT's lecture 27 relates these tests to the geometry of quadratic forms and their minima.[ref]
https://ocw.mit.edu/courses/18-06sc-linear-algebra-fall-2011/d163e012754258d3b374548504d8a18a_MIT18_06SCF11_Ses3.3sum.pdf. Strang, MIT 18.06SC, lecture 27 summary, Positive definite matrices and minima, pp. 1–2, 2011
A fundamental source of positive operators is for a bounded linear operator :For the complex matrix used above,Its eigenvalues are and , both positive. Positivity organizes the squared error minimized by a least-squares solution: the same occurs in the normal equation for a linear inverse problem. The positive square root of an operator gives the converse squared-length description, for every bounded positive operator .
On a real Hilbert space, nonnegativity of alone does not imply self-adjointness, so the Hermitian operator hypothesis is part of the definition. On a complex Hilbert space, real-valuedness of all these expressions already forces self-adjointness by the polarization identity.
Strict positivity and a uniform positive lower bound also differ in infinite dimension. On the l2 sequence space, the bounded Hermitian operator satisfiesConsequently no satisfies for every . Compare the positive definite symmetric operator and uniformly positive definite symmetric operator conditions.
For bounded positive operators , membership gives . Both terms are nonnegative, so both vanish. The Cauchy-Schwarz inequality for a positive operator makes . The reverse inclusion is immediate. Commutativity is not required.
For a bounded positive operator , the form is positive semidefinite and Hermitian. Expanding for complex and minimizing its quadratic polynomial gives the displayed inequality. If , varying along the negative direction of forces , covering the degenerate case. In particular implies , by testing against arbitrary .
A symmetric operator on a real Hilbert space is strictly positive definite when for every nonzero in its operator domain. Strict positivity proves uniqueness of a solution of , but in infinite dimension it need not prove existence for every . A uniformly positive definite symmetric operator additionally has a positive lower bound independent of . Symmetry must be stated separately when positivity is defined only by a real quadratic expression: adding a skew-symmetric matrix changes the operator without changing that expression.
For a symmetric bounded bilinear form that is strictly positive on nonzero vectors, and a bounded linear functional , set . The first variation is . A weak solution of satisfies , so it is the unique minimizer. Conversely, any minimizer solves the weak equation. A coercive bilinear form gives existence by the Lax-Milgram theorem; strict positivity alone does not.
For a bounded operator on a real Hilbert space, with . Consequently the Euler-Lagrange equation of is . It is only when is self-adjoint or when the relevant solution also annihilates the skew part. For example has positive quadratic form , but that form contains no information about its skew part.
A positive definite symmetric operator is uniformly positive definite if for some . For a bounded linear operator defined on a whole Hilbert space, this is a symmetric coercive operator, and the Lax-Milgram theorem supplies a unique solution to for every . For a differential operator, the corresponding bounded coercive form is used on its form space.
A positive contraction is a positive operator whose operator norm is at most one. Its spectrum lies in , so the spectral theorem for normal operators gives . Positive contractions are the effects in a POVM and occur in the pure-state gentle measurement bound.
The support of a positive operator is the closed subspaceThe range identity follows from image-kernel orthogonality for an adjoint and . Equivalently, it is the range of the spectral projection . In finite dimension, it is the span of eigenvectors with positive eigenvalues, and the range is already closed.
Closure matters in infinite dimension. On the l2 sequence space, has zero kernel and hence full support, but is not onto: is square-summable and its formal inverse image is not. In quantum relative entropy, containment of the first state's support in the reference state's support is necessary for finiteness; finite dimensions also make it sufficient, while infinite dimensions can have divergence despite that containment.
Every bounded positive operator on a Hilbert space has a unique bounded positive operator with . The continuous functional calculus applies the continuous scalar function on the compact spectrum of , givingIt also gives and .
For uniqueness, any positive with commutes with and hence with uniform polynomial approximations to . Thus commutes with , and commuting positive operators with equal squares are equal applies. Positivity is essential: has both and as self-adjoint square roots, but only the former is positive. For , the positive square root is .
For bounded positive operators satisfying the hypotheses, , so vanishes on the closure of the range of . The kernel of a sum of positive operators identity shows that it also vanishes on . Self-adjointness and the Hilbert projection theorem decompose the whole Hilbert space as that closed range plus that kernel. Thus . This proves the commuting case of uniqueness of a positive square root without using a spectral representation.
For a bounded Hermitian operator , the continuous functional calculus defines andBoth parts are positive operators, commute, and satisfy , so their supports are orthogonal. They apply the scalar functions and to the spectrum. The spectral theorem gives the same definitions even with continuous spectrum; a finite eigenvalue sum is only a special case.
If is additionally a trace-class operator, . Consequently implies . The trace-class assumption is essential: a general bounded Hermitian operator need not have a finite trace or trace norm. For , the two parts are and .
For a bounded Hermitian operator , its negative part of a Hermitian operator isIt is a positive operator, despite the word negative: its eigenvalues in a finite-dimensional diagonalization are the magnitudes of the negative eigenvalues of . For such a diagonalization, . The spectral-integral formula also covers a continuous spectrum without eigenvectors. Its support is orthogonal to the support of the positive part of a Hermitian operator, and .
For a bounded Hermitian operator , its positive part of a Hermitian operator isIt is a positive operator whose support is the positive spectral subspace. In a finite-dimensional spectral decomposition, ; the integral definition remains valid without an eigenbasis. Together with the negative part of a Hermitian operator, it obeys , and .
For Hermitian operators and , one writes when is positive semidefinite. This partial order is also called the semidefinite order.
For bounded positive operators in the Löwner order, one has even if and do not commute. A useful proof usesFor , put to obtain before the factor . The zero case is immediate. The continuous functional calculus therefore gives the corresponding operator integral.
For , inverse order yields . This follows by conjugating by and inverting its spectrum. ConsequentlyIntegration with positive weight proves the result. The integral converges in operator norm: near zero the integrand is bounded by , and near infinity by , or the corresponding bound for .
For a subspace of a finite-dimensional inner-product space,This follows directly by moving across the inner product and using .
A bounded operator on a real or complex Hilbert space is a normal operator when . Equivalently,The forward direction follows by taking inner products. Conversely the norm equality makes for every , so the polarization identity forces the self-adjoint difference to vanish.
Every Hermitian operator and every unitary operator is normal. The complex diagonal matrix is normal but neither self-adjoint nor unitary. In contrast, the unilateral shift on the l2 sequence space has and , so it is an isometry but not normal. For unbounded operators, normality additionally involves closedness and matching operator domains; the everywhere-defined bounded criterion above cannot be applied while ignoring them.
For a normal operator on a real or complex inner product space and a polynomial , the adjoint operator identity givesEvery power of commutes with every power of , since . Expanding finite sums consequently proves that commutes with , so it is a normal operator. The bar on means coefficientwise complex conjugation, and is essential for complex coefficients.
For a normal operator on a real or complex inner product space, equality of and gives . If , put . Then forces , and the adjoint operator identity givesThus and . Applying this implication successively to provesIn particular a nilpotent normal operator is zero. The argument uses no spectral theorem.
For a normal bounded operator, . If is bounded below, its range is closed; normality makes its adjoint injective, so the range is also dense and the operator is invertible.
The spectral theorem represents a normal operator by multiplication by its spectral parameter. In a finite-dimensional inner product space, this is diagonalization in an orthonormal basis; for a self-adjoint operator on a Hilbert space, it uses a projection-valued measure. Unbounded operators additionally require their natural square-integrability domain.
Every normal operator on a finite-dimensional complex inner-product space has an orthonormal basis of eigenvectors. An eigenvector is also an eigenvector of the adjoint with conjugate eigenvalue, so its orthogonal complement reduces the operator and induction applies.
A finite family of pairwise commuting normal operators on a finite-dimensional complex inner product space, commuting also with each other's adjoints, generates a commutative unital C-star algebra. Its finite maximal ideal space identifies the algebra with functions on finitely many points. The characteristic functions of those points pull back under the Gelfand transform to mutually orthogonal self-adjoint linear projections summing to the identity. Their ranges are the simultaneous eigenspaces, giving simultaneous diagonalization without choosing separate incompatible eigenbases.
For a Hermitian matrix with unitary diagonalization , define for a function on its finite spectrum. This acts on eigenvalues, unlike an entrywise matrix function . For example, the spectral square of a matrix is its ordinary matrix product with itself, whereas its entrywise square is a Hadamard power.
A normal operator on a separable complex Hilbert space has a unique normalized projection-valued measure such thatFor a bounded normal operator, is supported on its compact spectrum and the integral defines an everywhere-defined bounded operator. For a closed densely defined normal operator, its operator domain isThe scalar measure in this expression is positive and has total mass .
A spectral integral can exist without a single eigenvector. On , is bounded and self-adjoint, and . An eigenvector for would have to be supported on , a set of Lebesgue measure zero, so it would vanish in the L2 space. The spectral measure therefore extends orthonormal diagonalization to continuous spectral parameters rather than claiming an eigenbasis always exists.
Every self-adjoint operator on a separable complex Hilbert space is diagonal modulo a compact operator. For a bounded operator the compact self-adjoint error can have arbitrarily small Hilbert-Schmidt norm. In a cyclic spectral representation, partition a bounded real interval into intervals of width at most and successively bisect. Orthogonal step-function differences give basis vectors supported in the parent intervals. A diagonal value chosen in each support interval gives total squared column error at most . Since , this tends to zero. Apply the construction to countably many cyclic summands with summable squared error bounds. For an unbounded operator first split into bounded spectral intervals, obtaining a bounded compact error on the same domain.
Bounded self-adjoint operators on separable infinite-dimensional complex Hilbert spaces are unitarily equivalent modulo compact operators exactly when they have the same essential spectrum in the Calkin algebra. The Weyl-von Neumann theorem diagonalizes each modulo a compact error. Their diagonal cluster sets are the essential spectra; cluster-set matching of bounded sequences pairs the two diagonal lists with differences tending to zero, giving a compact diagonal difference. The separability assumption is essential: projections with countable and uncountable range dimensions can have the same essential spectrum but cannot differ by a compact operator after unitary conjugation.
A projection-valued measure assigns an orthogonal projection to each measurable set, with , , , and countable additivity in the strong operator topology on pairwise disjoint sets.
The continuous functional calculus of a bounded normal operator is a representation of , with . The Riesz-Markov-Kakutani representation theorem represents by a positive regular measure . The assignment is isometric for the norm, and extends to a unitary map onto the closed cyclic subspace it generates. Multiplication by and represent and . Orthogonal decomposition into cyclic reducing subspaces, allowing an uncountable family, constructs the Borel functional calculus for a normal operator without requiring separability.
For a unital star-representation , the Riesz-Markov-Kakutani representation theorem gives regular measures representing . Define by . Positivity and make these positive contractions. The identities imply commutation with and ; hence . Thus each is an orthogonal projection. Scalar countable additivity and orthogonality give countable additivity in the strong operator topology, constructing the normalized regular projection-valued measure.
If is a commutative C-star algebra of bounded operators on a Hilbert space containing its identity, there is a unique regular projection-valued measure on the character space such that and for every , where is the Gelfand transform.
For a faithful representation of by integration against a regular projection-valued measure , every nonempty open subset satisfies . Choose a nonzero continuous function supported inside , using compact Hausdorff normality. If , its spectral integral vanishes, contradicting faithfulness. Regularity is understood through the scalar spectral measures.
A spectral projector is the value of the projection-valued measure of a normal operator on a measurable part of its spectrum. In finite dimension, the spectral projector onto an eigenspace extracts the corresponding eigenvector component.
For a normal operator with projection-valued measure and vectors , its scalar spectral measure of a normal operator isThis is generally a complex measure, while is positive with total mass . For every bounded Borel function , the Borel functional calculus for a normal operator satisfiesFor an unbounded , one must additionally have , equivalently . This domain condition makes the pairing and scalar integral well-defined; an arbitrary unbounded function cannot be inserted without checking it.
For a self-adjoint operator , its projection-valued spectral measure of a normal operator assigns each Borel set an orthogonal projection . For a normalized quantum state , the probability of a measurement result in is . In finite dimensions, is the sum of the orthogonal projections onto eigenspaces whose eigenvalues lie in . This generalizes an individual eigenvector probability to degenerate eigenvalues and continuous spectra. The functional-calculus notation uses the indicator function of the set.
For a bounded normal operator with spectral measure of a normal operator , any projection commutes with and by the Borel functional calculus for a normal operator. Its range is a closed reducing subspace. If the spectrum has two points, choose disjoint nonempty relative open sets around them. Full support of a faithful spectral measure makes both projections nonzero and their product zero, so either range is nonzero and proper, in particular an invariant subspace.
For a projection-valued measure and vectors , the scalar spectral measure is the complex measure . The positive measure has total mass .
A sequence of scalar spectral measures converges weakly to if its integrals against every bounded continuous function converge. A concrete sufficient setting is that all underlying normal-operator spectra lie in one fixed compact set andThe Stone-Weierstrass theorem approximates any continuous function on uniformly by polynomials in . Also : apply Cauchy-Schwarz to the disjoint spectral projections in any finite partition. These uniform total-variation bounds control the approximation errors, so the mixed polynomial tests imply all continuous tests.
On a noncompact real spectrum, a uniform second-moment bound gives uniform tightness for positive scalar spectral measures, but does not alone justify the polynomial approximation step. High-degree polynomial tails require additional uniform integrability, and identifying a limit from all moments needs an additional hypothesis ensuring that these determine the measure. Merely asserting tightness does not supply either missing condition.
For a self-adjoint operator , fix the resolvent convention . For real , Stone formula states the strong-operator limitEquivalently the integrand is . The endpoint halves are necessary when there are eigenvalues at the endpoints; if neither endpoint has a spectral atom, the limit is simply the interval spectral projection.[ref]
Mathematical Methods in Quantum Mechanics by Gerald Teschl. first edition, §4.1, Theorem4.3, printed p114
To see the limit, use the spectral theorem to replace the resolvent difference by the scalar kernelIt lies in . Its arctangent primitive tends to one inside , zero outside , and one half at each endpoint. For every vector , dominated convergence applied to proves convergence in norm after applying the operators to . This explains both the strong rather than necessarily operator-norm limit and the half weights.
If in a Hilbert space, some subsequence has Cesàro means converging in norm to . For , choose the subsequence so successive vectors have summably small inner products; expansion of the squared norm of its averages then proves convergence.
Thus Hilbert spaces have the weak form of the Banach-Saks property.
For a Hilbertian basis , a sequence converges weakly if and only if is norm bounded and every coordinate sequence converges.
Every bounded sequence in a separable Hilbert space has a weakly convergent subsequence. Successive subsequences make each basis coordinate converge, and a diagonal subsequence converges in every coordinate; the coordinate criterion for weak convergence in a separable Hilbert space finishes the proof.
If in a Hilbert space, thenFor , take the inner product with and apply Cauchy-Schwarz before passing to the lower limit.
Suppose in a separable Hilbert space with Hilbertian basis . Then in norm exactly whenThe condition prevents norm from escaping to successively higher basis coordinates.
Every orthonormal sequence in a Hilbert space converges weakly to zero, by Bessel inequality, but no subsequence converges strongly because distinct terms remain distance apart.
If converges weakly to in a Banach space, there are convex combinations of each tail that converge in norm to .
Articles were limited to the first 100 out of 106 total.
Articles by others on the same topic
A Hilbert space is a fundamental concept in mathematics and quantum mechanics, named after the mathematician David Hilbert. It is a complete inner product space, which is a vector space equipped with an inner product that allows for the measurement of angles and lengths.
Key for quantum mechanics, see: mathematical formulation of quantum mechanics, the most important example by far being .



