Questions tagged [numerical-linear-algebra]
{numerical-linear-algebra} questions involving algorithms for linear algebra computations.
101 questions with no upvoted or accepted answers
22
votes
0
answers
445
views
Why does the random shift in the QR eigenvalue algorithm work in the non-symmetric case over the complex field
I tried to implement the QR algorithm for non-symmetric matrices with complex entries to show to my students. The main part of the implementation was standard: the Householder reduction to the ...
- 61.9k
13
votes
0
answers
591
views
What are the difficulties in proving almost-everywhere stability of Gaussian elimination?
It is well known that Gaussian elimination without pivoting is numerically unstable, and in practice Gaussian elimination is done with row pivoting (partial pivoting). A theorem of Wilkinson states ...
11
votes
0
answers
764
views
Fast computation of matrix product $AXA^T$ with fixed $A$?
Suppose we have two $n$-by-$n$ matrices $X$ and $A$, where $A$ is known and $X$ may change in different invocations, and we want to compute $AXA^T$. Is there an algorithm that beats the naive one of ...
- 75
11
votes
1
answer
2k
views
Quantifying the failure of the Cholesky factorization test for indefinite matrices
The Cholesky factorization is the classic test to check if a matrix is positive definite. In infinite precision it is also an exact test: A matrix has a Cholesky factorization iff it is positive ...
- 3,217
8
votes
0
answers
481
views
Problems where Conjugate gradient works much better than GMRES
I am interested in cases where Conjugate gradient works much better than GMRES method.
In general, CG is preferable choice in many cases of SPD because it requires less storage and theoretical bound ...
- 489
7
votes
0
answers
116
views
Are there attempts to numerically finding algebraic structures over finite-dimensional vector spaces?
By "algebraic structure" I mean a finite set of linear operators between tensor products of copies of one (or more) finite-dimensional (complex or real) vector spaces, fulfilling a set of ...
- 3,001
7
votes
0
answers
197
views
A special eigenvalue problem
For my research I need to solve a generalised eigenvalue problem
$Ax=\lambda B x$, where $A$, $B$ are general matrices, and selectively find only eigen-pairs $\lambda, x$ such that $\lambda\in \mathbb{...
- 492
7
votes
0
answers
120
views
How do I find elements of an algebra which generate an algebra contained in a fixed subspace?
Suppose $V$ is a linear subspace of a finite dimensional $C^*$-algebra $A$. (Feel free to assume $A$ is a multi-matrix algebra over $\mathbb C$.)
I would like to find $x \in V$ such that $\mathbb C [...
- 7,800
6
votes
0
answers
141
views
Algorithm to check a conjectural value for the rank of a large matrix
Feel free to suggest a different title, I'm not sure how to phrase this. I'm in the following somewhat specific situation:
I'm checking a conjecture which at the end of the day boils down to the ...
- 8,524
5
votes
0
answers
202
views
Difficulty of solving $Ax=b$ in terms of limiting spectral density of $A$?
Suppose $A$ is a random real-valued $n\times n$ matrix and we want to know the difficulty of solving $Ax=b$ when entries of $b$ are sampled IID from Normal$(0,1)$.
Can we say anything about the ...
- 2,442
5
votes
0
answers
392
views
Preconditioner for finding the smallest eigenpairs of a large, but structured, matrix
I'm trying to find the eigenvector corresponding to the second smallest eigenvalue of a large $(4,000,000 \times 4,000,000)$ matrix $L$. $L$ is a graph Laplacian, with the following structure: $L = D -...
- 500
5
votes
0
answers
112
views
Inverses of the sums of all possible subsets of a set of symmetric and positive definite matrices
I have a set of $c$ matrices $A_1 ... A_c$ which are all symmetric and positive definite. I would like to calculate the inverses of all the possible sums, i.e.
$(A_1+A_2)^{-1},(A_1+A_3)^{-1},(A_2+A_3)...
- 71
5
votes
0
answers
160
views
reference for perturbation of projection result
Let $A$ and $B$ have the same rank and dimensions. If $P_A$ denotes the projection onto the range space of $A$, then
$$
\|P_A - P_B\|_2 \leq \|A - B\| \cdot \min (\|A^\dagger\|_2, \|B^\dagger\|_2).
$$
...
- 922
5
votes
1
answer
124
views
Is Sun's spectral variation bound for normal matrices optimal?
In On the variation of the spectrum of a normal matrix, Sun proves the following result (Corollary 1.2):
Let $A$ be an $n$-square normal matrix and $B$ an arbitrary $n$-square matrix. Then $$ \min_{\...
- 283
4
votes
0
answers
82
views
Is there a fast way to compute the lowest eigenvalue of this symmetric PD matrix in this specific scenario?
Consider
$$C = A^H D A + M$$
where
$A$ is a $m \times m$ unitary matrix.
$D$ is a $m \times m$ diagonal matrix with entries either $0$ or $1$.
The number of $1$'s is $n \ll m$.
$M$ is a $m \times ...
- 698
4
votes
0
answers
539
views
Using Linear Programming as an iterative procedure
Suppose, we have a linear program and an optimal solution to it. Suppose now, we get a new constraint. We want to obtain an optimal solution to the given linear program extended by that new constraint....
- 391
4
votes
0
answers
957
views
Generate non-negative linear combinations of non-negative vectors with different supports
(I will not be surprised if this problem has been solved and/or has a trivial solution – I just do not know the right terminology to google for it.)
So the problem is as follows. I have an $m \...
- 367
4
votes
0
answers
149
views
Zero diagonal nonsymmetric block checkerboard matrix: orbits and numerical ranges
Let $A \in \mathbb{R}^{m \times m}$ be a nonsymmetric zero diagonal matrix with a zero/non-zero pattern which is symmetric and persymmetric (i.e. symmetric in the northeast-to-southwest diagonal).
If ...
- 323
4
votes
0
answers
2k
views
What is the time complexity of the largest singular value and its vectors?
Full zero-error SVD on an $m \times n$ matrix $A$ would cost $O(\min(m^2n,mn^2))$. What is the time complexity if we need only the largest singular value and its corresponding vectors? I think it is $...
- 123
4
votes
0
answers
578
views
Determining whether a Schur complement is invertible
Consider the symmetric matrix
$$M = \begin{bmatrix}
A & B \\
B^T & -C
\end{bmatrix}$$
where $A \in \cal{R}^{n \times n}$ and $C \in \cal{R}^{m\times m}$ are symmetric, ...
- 41
4
votes
0
answers
233
views
Determining Nullspace Basis such that only one column is deleted or added as row is added or deleted, and remaining columns of basis stay the same
I would like to compute, in MATLAB, the basis Z for the nullspace of an m by n matrix A, such that if one row of A is added (resulting in A_a), the basis for A_a is n-m-1 of the n-m columns of Z, i.e.,...
- 1,517
4
votes
0
answers
381
views
Efficiently factorize a KKT system with block diagonal upper corner
I have a system resulting from a quadratic energy minimization with linear equality constraints enforced with Lagrange multipliers which has the form:
\begin{equation}
A =
\left[\begin{array}{c|c}
\...
- 723
3
votes
0
answers
259
views
Efficient way to calculate Smith Normal Form of large integer matrices
I am interested in calculating the Smith Normal Form for Laplacian matrices of hypercube graphs. Using the elementary divisors method from SAGE, I was able calculate up to the 11-cube (which has a $2^{...
3
votes
0
answers
374
views
efficient numerical algorithm for matrix determinant
It appears that in numerical analysis the question of computing the determinant $\det A$ of a real or a complex $n\times n$ matrix $A$ is not well-studied, and a usual recommendation is to use matrix ...
- 14k
3
votes
0
answers
173
views
Can the Jordan decomposition of a matrix be computed in a backwards stable way?
Let $PJP^{-1}$ denote the Jordan decomposition of $M$. The matrix $J$ is a direct sum of Jordan blocks; it is unique up to permutation of the Jordan blocks. The matrix $P$ is not unique.
There are two ...
- 4,943
3
votes
0
answers
122
views
Preconditioners for $Ax=y$ that rely on hierarchical statistical modeling
Solving $Ax=y$ exactly can be done as:
fit a linear autoregressive model by treating rows of $A$ as data
apply this model to $A^T y$
Imperfect predictive model corresponds to an approximate inverse ...
- 2,442
3
votes
0
answers
147
views
Convolution integral and its matrix representation
My background is chemistry and I was exploring some one dimensional deconvolution problems i.e., resolution of two or more overlapping peaks. A lot of excellent work was done in the 1970-80s. However, ...
- 879
3
votes
0
answers
95
views
Sparse perturbation
Let $x, x_0\in\mathbb{R}^n$ be two vectors satisfying $$\frac{\|x\|_1}{\|x\|_2}\leq\frac{\|x_0\|_1}{\|x_0\|_2}.$$
$\| \cdot\|_1$ and $\| \cdot\|_2$ are the $\ell_1$ and $\ell_2$ norm in $\mathbb{R}^n$,...
- 31
3
votes
0
answers
243
views
An inequality concerning the solution of a Lyapunov equation
Let $A$ be an Hurwitz stable matrix (i.e. the real part of the eigenvalues of $A$ lie in the left-half plane) and $Q$ be a positive semidefinite matrix ($Q\ge 0$, for short). Let $P>0$ be the ...
- 2,712
3
votes
0
answers
60
views
Quasi-M matrices?
Does any body know a reference on lower triangular matrices with negative entries everywhere except for the diagonal and subdiagonal where entries are positive (when all entries are negative with ...
- 131
3
votes
0
answers
256
views
What would be a better method for numerical diagonalization of a certain Vandermonde-like matrix?
For the fractional iteration of the $\exp()$-function Hellmuth Kneser had 1942 proposed an analytic solution valid on the real numbers; there is a technical implementation for Pari/GP of this method ...
- 5,342
3
votes
0
answers
130
views
Computing the norm of the columns of an implicitly defined matrix
I have an $n \times n$ matrix $M = \Sigma W$ where $\Sigma$ is diagonal and $W$ orthogonal. $W$ is implicitly defined, i.e. I can only perform matrix-vector products (but I also have access to $W^T$).
...
3
votes
0
answers
220
views
Could SVD be used to optimize the partial inner-products?
Suppose a set $N$ of $n$ distinct points in $m-$dimensional space is given in $X\in\mathbb{R}^{n\times m}$. Also, suppose a subset $L\subset N$, $|L|=l<m<n$, with
$m-$dimensional coordinates in ...
- 131
3
votes
0
answers
1k
views
How many iterations are required for the Lanczos algorithm to converge?
I am trying to find the n smallest eigenvalues and eigenvectors of a NxN SPD matrix using Lanczos method. What is the number of iterations usually required? I mean, does it scale as $O(N)$ or $O(\sqrt{...
- 31
2
votes
0
answers
67
views
Characteristics of conjugate gradients' iterations for a matrix with clustered spectrum
I am interested in solving
\begin{equation}
Ax = b
\end{equation}
for a large sparse linear symmetric positive definite matrix $A$ by Conjugate Gradients method. (These systems usually come as ...
- 121
2
votes
0
answers
27
views
Solve sparse system with nested inverse
What is the most efficient way to solve an equation
\begin{align*}
(A\,E^{-1}\,C) x = b, \qquad A\in \mathbb{R}^{m\times n}, \, E \in \mathbb{R}^{n\times n}, \, C\in \mathbb{R}^{n\times m}
\end{align*}...
- 21
2
votes
0
answers
121
views
Proving some properties of the Landweber–Fridman iterates
$\newcommand\norm[1]{\lVert#1\rVert}$Let $B\in \mathbb R^{n\times n}$ be a symmetric and positive definite matrix. Assume that $x\in \mathbb R^n$ is the solution of $Bx=w$ for some given $w\in \mathbb ...
- 21
2
votes
0
answers
194
views
Maximize the product of Hadamard matrix and a vector
Let $X$ be an $n \times n$ Hadamard matrix (i.e. entries are in $\{-1,1\}$ and rows are orthogonal). For my application, we can assume $n=2^k$.
Given a vector $\bf{w} \in R^n$, I want to find the $X^*$...
- 21
2
votes
0
answers
75
views
Dense matrix vs sparse matrix, when they have same number of nonzero elements
I came across a new way in the literature to solve PDE problems numerically, which is called 'Patch Reconstruction'. One example paper is: Li, R., Sun, Z., Yang, F., & Yang, Z. (2019). A finite ...
- 21
2
votes
0
answers
47
views
Transforming a symmetric matrix into pentadiagonal form
Given a symmetric matrix $A$, which has complex values in the diagonal, but whose all other entries are real, I am interested in finding an orthonormal transformation $Q$ such that $Q^tAQ$ is a ...
- 251
2
votes
0
answers
616
views
block diagonal approximation of (SPD) matrix
I am interested in approximating a symmetric matrix in a block diagonal form, i.e. compute just some entries of the matrix located in blocks around the diagonal. Are there any theoretical guarantees ...
- 335
2
votes
1
answer
301
views
Best approximation with tensors of rank $\ge2$
Let $k\in\mathbb N$, $H_i$ be a (finite-dimensional, if necessary) $\mathbb R$-Hilbert space for $i\in I:=\{1,\ldots,k\}$, $H:=\bigotimes_{i\in I}H_i$ denote the tensor product$^1$ of $(H_i)_{i\in I}$ ...
- 167
2
votes
0
answers
52
views
Large-scale projected minimum-eigenvalue computations
I am interested in efficient numerical procedures for solving large-scale instances of the following projected minimum-eigenvalue problem:
$$\mu := \min_{v \in \mbox{ker}(A)} \frac{v^T H v}{\lVert v \...
- 35
2
votes
0
answers
33
views
Discrete maximum priniciple for parabolic operators
While reading a paper on the topic 'Numerical solutions for generalized Black-Scholes equation', It is given that their numerical scheme can be executed explicitly by solving a linear system $\mathbf ...
- 121
2
votes
0
answers
43
views
Partitioning $n$-space based on linear combinations
I'm trying to figure out the approximate number of areas the positive $n$-space will be divided into if we partition it as follows: we have $k$ linear functions $F_1$, $F_2$, ..., $F_k$ on $n$ ...
- 21
2
votes
0
answers
112
views
Coarse grid correction
Let $A_h \in \mathbb{R}^{n \times n}$ be the matrix corresponding to a finite element discretization of some nonselfadjoint, bounded and indefinite bilinear form corresponding to a second order ...
- 323
2
votes
0
answers
1k
views
Condition number of the product of two matrices
Consider two matrices $A$ and $B$ that are non-square in general and may not be full rank. Assuming their shapes are such that the product $A\cdot B$ is well-defined, what is the relationship between ...
- 171
2
votes
0
answers
272
views
Difficult integral - speed up numerical integration using a trick?
I have the following integral that I need to solve (with high precision) millions of times in my simulations. This is time consuming and is prohibiting me from proceeding from proceeding forward.
I ...
2
votes
1
answer
275
views
Computing 3-term connection coefficients for wavelets
I am trying to calculate the three-term connection coefficients
$$
Λ_{l,m}^{d_1,d_2,d_3} = \int_{-\infty}^\infty \varphi^{(d_1)}(x) \varphi^{(d_2)}_l(x) \varphi^{(d_3)}_m(x) dx
$$
for Daubechies ...
- 31
2
votes
0
answers
131
views
SVD when only off-diagonal terms are known
I have a real matrix $A \in \mathbb{R}^{n\times n}$ such that:
$A$ is symmetric
All the off-diagonal terms are known and positive
Has rank $k<n$
Unfortunately I don't know the values of the ...
