SearcharxivSearch

arXiv · 1504.01790

Generalizations of two cardinal inequalities of Hajnal and Juh\'asz

Abstract

A non-empty subset $A$ of a topological space $X$ is called \emph{finitely non-Hausdorff} if for every non-empty finite subset $F$ of $A$ and every family $\{U_x:x\in F\}$ of open neighborhoods $U_x$ of $x\in F$, $\cap\{U_x:x\in F\}\ne\emptyset$ and \emph{the non-Hausdorff number $nh(X)$ of $X$} is defined as follows: $nh(X):=1+\sup\{|A|:A\subset X$ is finitely non-Hausdorff$\}$. Clearly, if $X$ is a Hausdorff space then $nh(X)=2$. We define the \emph{non-Urysohn number of $X$ with respect to the singletons}, $nu_s(X)$, as follows: $nu_s(X):=1+\sup\{\mathrm{cl}_\theta(\{x\}):x\in X\}$. In 1967 Hajnal and Juh\'asz proved that if $X$ is a Hausdorff space then: (1) $|X|\le 2^{c(X)\chi(X)}$; and (2) $|X|\le 2^{2^{s(X)}}$; where $c(X)$ is the cellularity, $\chi(X)$ is the character and $s(X)$ is the spread of $X$. In this paper we generalize (1) by showing that if $X$ is a topological space then $|X|\le nh(X)^{c(X)\chi(X)}$. Immediate corollary of this result is that (1) holds true for every space $X$ for which $nh(X)\le 2^\omega$ (and even for spaces with $nh(X)\le 2^{c(X)\chi(X)}$). This gives an affirmative answer to a question posed by M. Bonanzinga in 2013. A simple example of a $T_1$, first countable, ccc-space $X$ is given such that $|X|>2^\omega$ and $|X|=nh(X)^\omega=nh(X)$. This example shows that the upper bound in our inequality is exact and that $nh(X)$ cannot be omitted (in particular, $nh(X)$ cannot always be replaced by $2$ even for $T_1$-spaces). In this paper we also generalize (2) by showing that if $X$ is a $T_1$-space then $|X|\le 2^{nu_s(X)\cdot 2^{s(X)}}$. It follows from our result that (2) is true for every $T_1$-space for which $nu_s(X)\le 2^{s(X)}$. A simple example shows that the presence of the cardinal function $nu_s(X)$ in our inequality is essential.

Explore related subjects

Keep this discovery

BibTeXRIS

Ivan S. Gotchev. 2015-04-08. Generalizations of two cardinal inequalities of Hajnal and Juh\'asz. https://arxiv.org/abs/1504.01790

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

Maximal Center of Distances of Finite Ultrametric Spaces and Perfect Binary Trees

We investigate the finite ultrametric spaces $(X,d)$ that have a given cardinality of the center of distances and a minimal cardinality of the set $X$. It is shown that such spaces are isometric if and only if their centers of distances are the same. The representing trees of these spaces are characterized up to isomorphism.

math.GN

A continuous $3$-distributive frame that is not $\omega$-distributive

We give a negative answer to the question, posed by Ern\'e, whether every $3$-distributive lattice is $\omega$-distributive. More precisely, we exhibit a continuous frame that is $\kappa$-distributive for every integer $\kappa\geq 2$, but is not a wide coframe. The frame is the open-set lattice of a compact, locally compact, countably based $T_0$ topological meet-semilattice, obtained from Lawson's construction in the logarithmic form described by Goubault-Larrecq. The failure of $\omega$-distributivity is witnessed by an explicit matrix with countably many nonempty finite rows: all row joins are the same nonzero element, whereas every choice of one entry from each row has meet zero. The same space answers negatively Ern\'e's accompanying question whether every $4$-web space is a wide web space. All properties of the construction needed for these conclusions are proved directly.

math.GN

An overlooked weakening of perfect normality: Perfect regularity in spaces and locales

We introduce the notion of perfect regularity as an appropriate weakening of perfect normality, both for spaces and locales. Various characterizations are given, using Dedekind-MacNeille completions, injective hulls, and sublocales. We place the new class of perfectly regular frames among various well-studied classes of frames. We also introduce the construction of perfect regularization of a completely regular frame, compare it to Isbell's well-known booleanization construction, and argue that it is at least as important as the latter.

math.GN