Numerical ranges of the powers of an operator (Q2267467): Difference between revisions

From MaRDI portal
Added link to MaRDI item.
ReferenceBot (talk | contribs)
Changed an Item
 
(2 intermediate revisions by 2 users not shown)
Property / MaRDI profile type
 
Property / MaRDI profile type: MaRDI publication profile / rank
 
Normal rank
Property / OpenAlex ID
 
Property / OpenAlex ID: W2079330335 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5673950 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Operator radii and unitary operators / rank
 
Normal rank
Property / cites work
 
Property / cites work: On the numerical radius of matrices and its application to iterative solution methods / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3910969 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Numerical ranges and dilations<sup>∗</sup> / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4533180 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4769499 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Some mapping theorems for the numerical range / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3874765 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3116905 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Minimal range theorems for operators with thin spectra / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3998482 / rank
 
Normal rank

Latest revision as of 11:45, 2 July 2024

scientific article
Language Label Description Also known as
English
Numerical ranges of the powers of an operator
scientific article

    Statements

    Numerical ranges of the powers of an operator (English)
    0 references
    0 references
    0 references
    1 March 2010
    0 references
    For a bounded linear operator \(A\) on a complex Hilbert space \(H\) with inner product \(\langle\cdot,\cdot\rangle\), its numerical range \(W(A)\), numerical radius \(w(A)\) and Crawford number \(c(A)\) are defined by \[ W(A)= \{\langle Ax,x\rangle: x\in H,\;\| x\|= 1\},\quad w(A)= \sup\{|z|: z\in W(A)\} \] and \[ c(A)= \text{inf}\{|z|: z\in W(A)\}, \] respectively. In this paper, results are derived which relate \(W(A^k)\), \(w(A^k)\) and \(c(A^k)\) (for \(k\) a positive integer or a negative integer if \(A\) is invertible) to \(W(A)\), \(w(A)\) and \(c(A)\). More specifically, it is shown for \(k\geq 1\) that {\parindent=6mm \begin{itemize}\item[(1)] \(W(A^k)\) is contained in \(W(\widetilde A)^k\), where \(\widetilde A= A\oplus uA\oplus\cdots\oplus u^{k-1} A\), \(u= \exp(2\pi i/k)\) and \(W(\widetilde A)^k\) denotes the convex set \(\{z^k: z\in W(\widetilde A)\}\), \item[(2)] if \(W(A)\) is contained in the circular segment \(\{z\in\mathbb{C}:|z|\leq 1,\,\text{Re\,}z\geq \cos\varphi\}\) \((0<\varphi\leq\pi/2)\), then \(W(A^k)\) is contained in \(\{z\in\mathbb{C}: |z|\leq 1\), \(\text{Re\,}z\geq \cos(k\varphi)\}\) for \(\lceil k/2\rceil\leq \pi/(2\varphi)\), \item[(3)] if \(c(A)= w(A)\cos\varphi\) \((0< \varphi\leq \pi/2)\), then \(c(A^k)\geq w(A^k)\cos(k\varphi)\) for \(k\leq \pi/(2\varphi)\), and \item[(4)] if \(c(A)\geq w(A)\cos(\pi/(2k))\), then \(c(A)^k\geq c(A^k)\). \end{itemize}} In the case of \(A\) invertible, the authors obtain \(c(A^{-1})\geq c(A)/w(A)^2\) and \(w(A^{-k})\geq w(A)^{-k}\) for all \(k\geq 1\). In the latter, the equality holds if and only if \(A\) is a nonzero multiple of a unitary operator. This is a generalization of a result of Stampfli that if \(A\) is invertible such that \(w(A)= w(A^{-1})= 1\), then \(A\) is unitary.
    0 references
    numerical range
    0 references
    numerical radius
    0 references
    Crawford number
    0 references
    bounded linear operator
    0 references
    Hilbert space
    0 references

    Identifiers