In linear algebra, a real symmetric matrix represents a self-adjoint operator represented in an orthonormal basis over a realinner product space. The corresponding object for a complex inner product space is a Hermitian matrix with complex-valued entries, which is equal to its conjugate transpose. Therefore, in linear algebra over the complex numbers, it is often assumed that a symmetric matrix refers to one which has real-valued entries. Symmetric matrices appear naturally in a variety of applications, and typical numerical linear algebra software makes special accommodations for them.
The following matrix is symmetric:
The sum and difference of two symmetric matrices is symmetric.
This is not always true for the product: given symmetric matrices and , then is symmetric if and only if and commute, i.e., if .
For any integer , is symmetric if is symmetric.
If exists, it is symmetric if and only if is symmetric.
Rank of a symmetric matrix is equal to the number of non-zero eigenvalues of .
Decomposition into symmetric and skew-symmetric
Any square matrix can uniquely be written as sum of a symmetric and a skew-symmetric matrix. This decomposition is known as the Toeplitz decomposition. Let denote the space of matrices. If denotes the space of symmetric matrices and the space of skew-symmetric matrices then and , i.e.
A symmetric matrix is determined by scalars (the number of entries on or above the main diagonal). Similarly, a skew-symmetric matrix is determined by scalars (the number of entries above the main diagonal).
Matrix congruent to a symmetric matrix
Any matrix congruent to a symmetric matrix is again symmetric: if is a symmetric matrix, then so is for any matrix .
If and are real symmetric matrices that commute, then they can be simultaneously diagonalized: there exists a basis of such that every element of the basis is an eigenvector for both and .
Every real symmetric matrix is Hermitian, and therefore all its eigenvalues are real. (In fact, the eigenvalues are the entries in the diagonal matrix (above), and therefore is uniquely determined by up to the order of its entries.) Essentially, the property of being symmetric for real matrices corresponds to the property of being Hermitian for complex matrices.
Complex symmetric matrices
A complex symmetric matrix can be 'diagonalized' using a unitary matrix: thus if is a complex symmetric matrix, there is a unitary matrix such that is a real diagonal matrix with non-negative entries. This result is referred to as the Autonne–Takagi factorization. It was originally proved by Léon Autonne (1915) and Teiji Takagi (1925) and rediscovered with different proofs by several other mathematicians. In fact, the matrix is Hermitian and positive semi-definite, so there is a unitary matrix such that is diagonal with non-negative real entries. Thus is complex symmetric with real. Writing with and real symmetric matrices, . Thus . Since and commute, there is a real orthogonal matrix such that both and are diagonal. Setting (a unitary matrix), the matrix is complex diagonal. Pre-multiplying by a suitable diagonal unitary matrix (which preserves unitarity of ), the diagonal entries of can be made to be real and non-negative as desired. To construct this matrix, we express the diagonal matrix as . The matrix we seek is simply given by . Clearly as desired, so we make the modification . Since their squares are the eigenvalues of , they coincide with the singular values of . (Note, about the eigen-decomposition of a complex symmetric matrix , the Jordan normal form of may not be diagonal, therefore may not be diagonalized by any similarity transformation.)
Using the Jordan normal form, one can prove that every square real matrix can be written as a product of two real symmetric matrices, and every square complex matrix can be written as a product of two complex symmetric matrices.
Cholesky decomposition states that every real positive-definite symmetric matrix is a product of a lower-triangular matrix and its transpose,
If the matrix is symmetric indefinite, it may be still decomposed as where is a permutation matrix (arising from the need to pivot), a lower unit triangular matrix, and [relevant?] is a direct sum of symmetric and blocks, which is called Bunch–Kaufman decomposition 
A general (complex) symmetric matrix may be defective and thus not be diagonalizable. If is diagonalizable it may be decomposed as
where is an orthogonal matrix , and is a diagonal matrix of the eigenvalues of . In the special case that is real symmetric, then and are also real. To see orthogonality, suppose and are eigenvectors corresponding to distinct eigenvalues , . Then
Since and are distinct, we have .
Symmetric matrices of real functions appear as the Hessians of twice differentiable functions of real variables (the continuity of the second derivative is not needed, despite common belief to the opposite).
Every quadratic form on can be uniquely written in the form with a symmetric matrix . Because of the above spectral theorem, one can then say that every quadratic form, up to the choice of an orthonormal basis of , "looks like"
with real numbers . This considerably simplifies the study of quadratic forms, as well as the study of the level sets which are generalizations of conic sections.
This is important partly because the second-order behavior of every smooth multi-variable function is described by the quadratic form belonging to the function's Hessian; this is a consequence of Taylor's theorem.
An matrix is said to be symmetrizable if there exists an invertible diagonal matrix and symmetric matrix such that
The transpose of a symmetrizable matrix is symmetrizable, since and is symmetric. A matrix is symmetrizable if and only if the following conditions are met:
implies for all
for any finite sequence
Other types of symmetry or pattern in square matrices have special names; see for example: