Weightable quasi-uniformities (Q452809)

A quasi-uniformity on a set \(X\) is a filter \(\mathfrak U\) on \(X\times X\) such that every member of \(\mathfrak U\) is a reflexive relation on \(X,\) and if \(U\in \mathfrak U,\) then \(V\circ V\subseteq U\) for some \(V\in\mathfrak U.\) A quasi-uniformity \(\mathfrak U\) on a set \(X\) is called totally bounded provided that for each \(U\in\mathfrak U\vee\mathfrak U^{-1}\) there is a finite set \(F\subseteq X\) such that \(U(F)=X.\) Let us call a quasi-uniformity \(\mathfrak U\) on a set \(X\) \(t\)-symmetrizable if there is a totally bounded quasi-uniformity \(\mathcal Z\) on \(X\) such that \(\mathfrak U\vee \mathcal Z\) is a uniformity. The following two theorems seem to be the main results of the paper: Theorem 1. Let \(\mathfrak U\) and \(\mathcal V\) be quasi-uniformities on a set \(X\) such that there is a totally bounded quasi-uniformity \(\mathcal Z\) on \(X\) with \[ \mathfrak U\subseteq \mathcal V \vee \mathcal Z. \] Then the inclusion \[ \mathfrak U\subseteq \mathcal V \vee (\mathfrak U\vee\mathcal V^{-1} )_{w} \] holds where \((\mathfrak U\vee\mathcal V^{-1} )_{w}\) is the finest totally bounded quasi-uniformity coarser than \(\mathfrak U\vee\mathcal V^{-1}.\) Theorem 2. Let \(\mathfrak U\) be a \(t\)-symmetrizable quasi-uniformity on a set \(X.\) Then \(\mathfrak U\) can be written as the supremum of a family of quasi-pseudometric quasi-uniformities \(\mathfrak U_{q}\) where \(q\) is a quasi-pseudometric on \(X\) such that the quasi-uniformity \(\mathfrak U_{q}\) is \(t\)-symmetrizable. Recall that a function \(d: X\times X\rightarrow [0,\infty)\) is a quasi-pseudometric on \(X\) if \(d(x,x)=0\) whenever \(x\in X\) and \(d(x,z)\leq d(x,y)+d(y,z)\) whenever \(x,y,z\in X.\) In the first part of the paper the authors study also the quasi-uniformities that are generated by the so-called weightable quasi-pseudometrics.