Groupoid metrization theory. With applications to analysis on quasi-metric spaces and functional analysis (Q1928116)

The aim of the present book is to prove some very general metrization results showing that apparently unrelated results, from topology, harmonic analysis and functional analysis, have in fact a common background. The idea is to prove metrization results for semigroupoids and groupoids which extend and unify results from topology (Alexandrov-Urysohn metrization theorem for uniform spaces), harmonic analysis (Macías-Segovia metrization theorem for quasimetric spaces), functional analysis (the Aoki-Rolewicz normability theorem) and topological group theory (the Birkhoff-Kakutani metrization theorem). A semigroupoid is a nonempty set \(G\) equipped with a partially defined binary operation \(*\) satisfying the associative law: if \(a*b\) and \(b*c\) are defined, then \((a*b)*c\) and \(a*(b*c)\) are defined and equal. A groupoid is a semigroupoid \((G,*)\) endowed with a further unary operation \((\cdot)^{-1}:G\to G\) such that \(a*a^{-1}\) and \(a^{-1}*a\) are defined for all \(a\in G,\) and \(a*b*b^{-1}=a,\; a^{-1}*a*b=b,\) provided \(a*b\) is defined. The algebraic properties of semigroupoids and groupoids are studied in the second chapter of the book, where topological semigroupoids and groupoids are considered as well. The first chapter, ``Introduction'', contains the motivation for the general method proposed by the authors as well as a discussion about some relevant applications. The general metrizability theorems (GMT) are proved in the third chapter ``Quantitative metrization theory'', first for a groupoid \((G,*,(\cdot)^{-1})\) endowed with a function \(\psi:G\to [0,\infty)\) satisfying some conditions related to the algebraic structure of \(G\), such as \(\psi(a*b)\leq C_1\max\{\psi(a),\psi(b)\}, \) and \(\psi(A^{-1})\leq C_0\psi(a), \) for all \(a,b\in G\). The main result here is Theorem 3.26 whose statement is six pages long and which contains very general quantitative metrizability results, inspired by Macías-Segovia's one, and which extends many known results. The more general case of semigroupoids is treated in Theorem 3.28. One considers also the case of quasimetric spaces and one shows how the metrization results mentioned above fit in this general scheme. In Chapter 4, ``Applications to analysis on quasimetric spaces'', the general results obtained in the third chapter are specialized to quasimetric spaces. A quasimetric on a set \(X\) is a function \(d:X\times X\to [0,\infty)\), satisfying all the axioms of a metric, except for the triangle inequality which is replaced with \(d(x,z)\leq c(d(x,y)+d(y,z)),\) for some \(c\geq 1\). Similarly, a quasi-norm \(\|\cdot\|\) on a vector space \(X\) satisfies \(\|x+y\|\leq c(\|x\|+\|y\|)\) instead of the triangle inequality. (Notice that there is another meaning of the term quasimetric, namely that of an asymmetric metric \(d\), i.e., \(d\) satisfies all the axioms of a metric except for symmetry; it is possible that \(d(x,y)\neq d(y,x)\) for some \(x,y\in X\), see, e.g., [\textit{P. Fletcher} and \textit{W. F. Lindgren}, Quasi-uniform spaces. New York etc.: Marcel Dekker (1982; Zbl 0501.54018)].) Among the topics treated in this Chapter, we mention: the extension of Hölder functions defined on quasimetric spaces and the properties of the spaces of these functions (separation and density results; in contrast to the metric case, the space of Lipschitz functions on a quasimetric space could contain only the null function -- note that the dual of the quasi-normed space \(L^p[0,1]\) is \(\{0\}\) for \(0<p<1\)), the study of the regularized distance function, Whitney-like partitions of unity via Hölder functions, smoothness indices of quasimetric spaces, distribution theory on quasimetric spaces, Hardy spaces on quasimetric spaces, the Pompeiu-Hausdorff distance, and others. The goal of the fifth Chapter ``Nonlocally convex analysis'', is to establish completeness and separability criteria of topological vector spaces that are typically nonlocally convex such as Lebesgue-like spaces (\(L^p\)-spaces with \(0<p<1\)), Lorentz spaces, Orlicz spaces, discrete Lizorkin-Triebel and Besov spaces, etc. The treatment is based on the theory of the so-called capacitary functions, introduced and studied by the first three authors and \textit{E. Ziadé} in [J. Funct. Anal. 262, No. 11, 4766--4830 (2012; Zbl 1281.46029)]. The aim of the last chapter of the book, Chapter 6, ``Functional analysis on quasi-pseudonormed groups'', is to show that a significant portion of functional analysis can be carried out on groups equipped with topologies that are only partially compatible with their algebraic structure, called quasi-pseudonormed groups. More specifically, one supposes that the topology on a group \(G\) is generated by a nonnegative function \(\psi\) satisfying some properties similar to, but weaker than, those of a norm on a vector space. One proves within this framework quantitative versions of the open mapping theorem, of the closed graph theorem and of the uniform boundedness principle, and one shows how all these three fundamental principles of functional analysis can be obtained in a unified way. The book is very well organized -- there are four detailed indices at the end of the book, a Symbol Index, a Subject Index, a Theorem Index and an Author Index. The bibliography contains 130 items. This research monograph, based mainly on the original contributions of the authors, proposes a very general approach to some results in topology, harmonic analysis and functional analysis, all revolving around the idea of metrizability. It will be of interest to researchers in these areas, as well as to those interested in the abstract approach proposed by the authors.