Searcharxiv⌕ Search

arXiv subjects

Anna Pelczar-Barwacz

Publications and source records attributed to Anna Pelczar-Barwacz.

14 recordsLinked to original sources

Geometric duality, perfect graphs, and the Sierpiński space

In their classical paper \emph{On the stopping time Banach space}, Bang and Odell, among a plethora of results concerning the dyadic stopping time space and its dual, presented the first non-trivial example of the \emph{duality phenomenon} between combinatorial Banach spaces. We give a full characterization of such pairs $(\mc{F}_0, \mc{F}_1)$ of families of finite sets: This duality holds iff there is a perfect graph $G$ on $\NN$ such that $\mc{F}_0$ consists of all finite cliques of $G$ and $\mc{F}_1$ consists of all finite anti-cliques of $G$. As it turns out, Lovász' famous perfect graph theorem is an immediate corollary of this result. Among the many examples of such pairs of families, we investigate a particularly interesting one, when $G$ is the Sierpiński graph, and study general methods of embedding combinatorial and classical sequence spaces in the generated space, including the Schreier and $\ell_p$ spaces.

math.FA↗

Banach spaces with arbitrary finite Baire order

We investigate intrinsic Baire classes of Banach spaces defined by Argyros, Godefroy and Rosenthal (2003). We introduce a construction, for any Banach space $X$ with a basis, of an $\ell_1$-saturated separable Banach space $Y$ such that for any $α\leqslant ω_1$ we have $Y^{**}_{1+α} \cong Y \oplus X^{**}_α$, where $X^{**}_α$ denotes the $α$-th intrinsic Baire class of $X$. We apply this construction to answer two open problems by Argyros, Godefroy and Rosenthal (2003), namely we build separable Banach spaces of any Baire order less or equal to $ω$, and a non-universal separable Banach space of order $ω_1$. Finally, we apply the construction to show an analogue of a result of Lindenstrauss (1971) by constructing, for any Banach space $X$ with a basis and any $n \in \mathbb{N}$, a Banach space $Y$ such that $Y^{**}_n \cong Y^{**}_{n-1} \oplus X$, showing that any such $X$ can appear as the space of functionals in a bidual Banach space $Y^{**}$ that are of $n$-th intrinsic Baire class but not of $(n-1)$-th intrinsic Baire class.

math.FA↗

On the spaces dual to combinatorial Banach spaces

We present quasi-Banach spaces which are closely related to the duals of combinatorial Banach spaces. More precisely, for a compact family $\mathcal{F}$ of finite subsets of $ω$ we define a quasi-norm $\lVert \cdot \rVert^\mathcal{F}$ whose Banach envelope is the dual norm for the combinatorial space generated by $\mathcal{F}$. Such quasi-norms seem to be much easier to handle than the dual norms and yet the quasi-Banach spaces induced by them share many properties with the dual spaces. We show that the quasi-Banach spaces induced by large families (in the sense of Lopez-Abad and Todorcevic) are $\ell_1$-saturated and do not have the Schur property. In particular, this holds for the Schreier families.

math.FA↗

The ZOO of combinatorial Banach spaces

We study Banach spaces induced by families of finite sets in the most natural (Schreier-like) way, that is, we consider the completion $X_\mc{F}$ of $c_{00}$ with respect to the norm $\sup\{\sum_{k\in F}|x(k)|:F\in\mc{F}\}$ where $\mc{F}$ is an arbitrary (not necessarily compact) family of finite sets covering $\mbb{N}$. Among other results, we discuss the following: (1) Structure theorems bonding the combinatorics of $\mc{F}$ and the geometry of $X_\mc{F}$ including possible characterizations and variants of the Schur property, $\ell_1$-saturation, and the lack of copies of $c_0$ in $X_\mc{F}$. (2) A plethora of examples including a relatively simple $\ell_1$-saturated combinatorial space which does not satisfy the Schur property, as well as a new presentation of Pełczyński's universal space. (3) The complexity of the family $\{H\subseteq\NN:X_{\mc{F}\upharpoonright H}$ does not contain $c_0\}$.

math.FA↗

Reflexive Calkin algebras

For a Banach space $X$ denote by $\mathcal{L}(X)$ the algebra of bounded linear operators on $X$, by $\mathcal{K}(X)$ the compact operator ideal on $X$, and by $Cal(X) = \mathcal{L}(X)/\mathcal{K}(X)$ the Calkin algebra of $X$. We prove that $Cal(X)$ can be an infinite-dimensional reflexive Banach space, even isomorphic to a Hilbert space. More precisely, for every Banach space $U$ with a normalized unconditional basis not having a $c_0$ asymptotic version we construct a Banach space $\mathfrak{X}_U$ and a sequence of mutually annihilating projections $(I_s)_{s=1}^\infty$ on $\mathfrak{X}_U$, i.e., $I_sI_t = 0$, for $s\neq t$, such that $\mathcal{L}(\mathfrak{X}_U) = \mathcal{K}(\mathfrak{X}_U)\oplus[(I_s)_{s=1}^\infty]\oplus\mathbb{C}I$ and $(I_s)_{s=1}^\infty$ is equivalent to $(u_s)_{s=1}^\infty$. In particular, $Cal(\mathfrak{X}_U)$ is isomorphic, as a Banach algebra, to the unitization of $U$ with coordinate-wise multiplication. Banach spaces $U$ meeting these criteria include $\ell_p$ and $(\oplus_n\ell_\infty^n)_p$, $1\leq p<\infty$, with their unit vector bases, $L_p$, $1 <p<\infty$, with the Haar system, the asymptotic-$\ell_1$ Tsirelson space and Schlumprecht space with their usual bases, and many others.

math.FA↗

Small operator ideals on the Schlumprecht and Schreier spaces

We present a method of building operators on a Banach space $X$ that generate distinct operator ideals in the algebra $\mathscr{B}(X)$ of bounded linear operators on $X$. We show that there are exactly $2^\mathfrak{c}$ distinct small closed operator ideals on the Schlumprecht space and there is a chain of cardinality $\mathfrak{c}$ of small closed operator ideals on any Schreier space of finite order.

math.FA↗

Unconditionally saturated Banach space with the scalar-plus-compact property

We construct a Bourgain-Delbaen $\mathscr{L}_\infty$-space $\mathfrak{X}_{Kus}$ with strongly heterogenous structure: any bounded operator on $\mathfrak{X}_{Kus}$ is a compact perturbation of a multiple of the identity, whereas the space $\mathfrak{X}_{Kus}$ is saturated with unconditional basic sequences.

math.FA↗

Operators in tight by support Banach spaces

We answer the question of W.T. Gowers, giving an example of a bounded operator on a subspace of Gowers unconditional space which is not a strictly singular perturbation of a restriction of a diagonal operator. We make some observations on operators in arbitrary tight by support Banach space, showing in particular that in such space no two isomorphic infinitely dimensional subspaces form a direct sum.

math.FA↗

A type (4) space in (FR)-classification

We present a reflexive Banach space with an unconditional basis which is quasi-minimal and tight by range, i.e. of type (4) in Ferenczi-Rosendal list within the framework of Gowers' classification program of Banach spaces. The space is an unconditional variant of the Gowers Hereditarily Indecomposable space with asymptotically unconditional basis.

math.FA↗

Strictly singular non-compact operators on a class of HI spaces

We present a method for constructing bounded strictly singular non-compact operators on mixed Tsirelson spaces defined either by the families (A_n) or (S_n) of a certain class, as well as on spaces built on them, including hereditarily indecomposable spaces.

math.FA↗

Strictly singular operators in asymptotic $\ell_p$ Banach spaces

We present condition on higher order asymptotic behaviour of basic sequences in a Banach space ensuring the existence of bounded non-compact strictly singular operator on a subspace. We apply it in asymptotic $\ell_p$ spaces, $1\leq p<\infty$, in particular in convexified mixed Tsirelson spaces and related asymptotic $\ell_p$ HI spaces.

math.FA↗

Isomorphisms and strictly singular operators in mixed Tsirelson spaces

We study the family of isomorphisms and strictly singular operators in mixed Tsirelson spaces and their modified versions setting. We show sequential minimality of modified mixed Tsirelson spaces $T_M[(\mc{S}_n,θ_n)]$ satisfying some regularity conditions and present results on existence of strictly singular non-compact operators on subspaces of mixed Tsirelson spaces defined by the families $(\mc{A}_n)_n$ and $(\mc{S}_n)_n$.

math.FA↗