Author: Eiko
Tags: matrix, matrix analysis, eigenvalues, singular values, p-adic
Time: 2024-11-29 17:41:45 - 2024-12-06 22:20:59 (UTC)
Reference: p-adic Differential Equations by Kiran S. Kedlaya
Matrix With Complex Numbers
For complex numbers we use the norm for vectors and the operator norm for matrices.
We always arrange the eigenvalues of in absolute value decreasing order.
Singular Values
It is familiar by inner product and orthogonalization that real symmetric matrices have real eigenvalues. Semi-positive definite matrices have non-negative eigenvalues, which also holds for complex matrices.
Definition. The singular values of are the square roots of the eigenvalues of .
For a real symmetric matrix these are basically the absolute values of the eigenvalues.
Singular Value Decomposition
There are unitary such that
is a diagonal matrix, where are the singular values of .
Singular value decomposition preserves metric on both sides. So it is a good measurement of the size of the matrix.
Changing notation ,we can write it as , which basically means you can find orthogonal basis such that .
For any vector we have .
For any two dimensional subspace there is a non-zero vector such that . This generalizes to higher dimensions.
is the smallest number such that for every dimensional subspace you can find a non-zero vector such that . i.e.
Using this formula it is obvious that all singular value does not exceed operator norm, . Actually . So you can consider singular values as generalizations of operator norm! Instead of one number, you have a sequence of numbers describing the size of the matrix in different dimensions.
Remarks. Singular values are the invariants of a linear map between two inner product spaces , actually for matrices, here are given basis and equipped with standard inner product. The structures of two inner products on source and target spaces are preserved.
Interaction With Exterior Algebra
For an inner product space with orthonormal basis , the exterior space has inner product structure and an orthonormal basis for .
Since we can map
Singular Values Bounds Minors
In fact this observation gives another interpretation of singular values in terms of exterior product and orthonormal basis,
In particular this means all minors of are bounded by the partial products of singular values, if then
From and we see that . We have
Applying above to we obtain Weyl’s inequality
which is an equality when .
Weyl’s inequality has a converse, if you have and such that for all then there is a matrix such that are the singular values and are the eigenvalues.
Remarks. Unlike eigenvalues, singular values do not behave well under polynomial compositions, since the singular value decomposition essentially assumes the source and target spaces be different spaces. But they do work under inverses, from the singular value decomposition you can see the singular values of are .
Perturbations
The singular values behave better with additive perturbations,
This can be easily proved by the curious formula of in sup inf, and the two facts that .
The eigenvalues cannot easily be controlled by additive perturbations, but we can say about the characteristic polynomials.
This can be seen as the term is the sum of all -principal minors, we have
For multiplicative perturbation, if is invertible we have
This can be seen using the formula
A similar formula also holds, you can see this by applying transpose (also called dual or adjoint) and observe that singular values are invariant under transpose (dual) operation.
Matrix With -adic Numbers
For -adic fields (or actually, any other non-archimedean valued fields) the infinity norm on vectors is more appropriate
The operator norm is defined as usual using this norm.
As a corollary we have
This means .
Note also that , therefore
Matrix in preserves the vector norm, . This can be seen by
Hodge Polygon
For a sequence its associated polygon is the path joined by . The are understood as successive slopes, and if are non-decreasing, the polygon is convex.
Given a matrix , let be a sequence whose partial sum satisfy
are called elementary divisors or invariant factors of .
The Hodge polygon of is the associated polygon of this sequence, and the singular values of are given by , here should be the base used in the absolute value .
Singular Values Computes Maximal Minors
As a corollary of the definition,
These are are clearly invariant under the action of , as a result the singular values and Hodges polygons are also invariant.
Since we are in a valuation ring, through the same process of putting matrix with PID coefficients into normal form, we have the smith normal form of as the diagonal matrix, where and . This definition of singular value make sense because the vector norms on source and target spaces are preserved by .
The process guarantees that which means , i.e. Hodge polygon is convex.
Similarly we have the subspace characterization of singular values
Remarks. Here preserves vector norm, so it is natural choice for action. Unlike the complex case we do not try to preserve the inner product structure but we choose to preserve the vector norm. This is because the -adic norm is not obtained from an inner product. The singular values are the invariants of a linear map between two normed spaces .
Newton Polygon Associated To Eigenvalues
For a matrix with eigenvalues the Newton polygon is the associated polygon of the sequence , so that .
As with eigenvalues, Newton polygon is invariant under conjugation by .
Weyl’s inequality holds as well
with equality when . This means the Newton polygon lies above the Hodge polygon, with the same starting and ending points.
Hodge-Newton Decomposition
If the Hodge and Newton polygons of a matrix ‘breaks into two parts’, i.e. for some we have
the matrix has a Hodge-Newton decomposition at , there is an integral matrix such that
If moreover then can be chosen to be zero.
Perturbations Of -adic Matrices
A lot of the perturbation results for complex matrices have a -adic analog.
Non-Archimedean Property For Vectors
Recall that for numbers we have , with equality when .
There is a vector non-archimedean property (for the -norm) for vectors , since , we have
If , there exists at least on such that , at this time the equality holds.
Perturbations
If , then for all .
This is the direct consequence of the vector non-archimedean property and the sup-inf property of singular values, as for the vector that attains the sup-inf of , we have therefore .
Let and be the characteristic polynomials. Then
Here the coefficients appearing in the complex version of the statement are replaced by , due to the archimedean property of -adic numbers.
For multiplicative perturbation, the result is the same