Dimensions \(\mathcal K\)-Ind and \(\mathcal L\)-Ind. Some answers (Q554407): Difference between revisions

In previous papers the author developed new inductive dimension functions \({\mathcal K}\)-Ind and \({\mathcal L}\)-Ind, where \({\mathcal K}\) is a non-empty set of finite simplicial complexes and \({\mathcal L}\) is a non-empty set of compact polyhedra. In [Quest. Answers Gen. Topology 28, No. 2, 109--132 (2010; Zbl 1222.54036)] the transfinite variants tr-\({\mathcal K}\)-Ind and tr-\({\mathcal L}\)-Ind were introduced for normal spaces and questions were asked, some of which are answered here. Before the definition we need some auxiliary notions. For a space \(X\) the set of all closed subsets of a space \(X\) is called \(\exp X\) and we let \(\text{Fins}(\exp X)\) the set of finite sequences of members of \(\exp X\). Complexes \(K\) are finite complete simplicial complexes. We put \(\text{Exp}_K(X)= \{\Phi\in\text{Fin}_s(\exp X)\): there is an embedding of the nerve \(N(\Phi)\subset K\}\). For \(\Phi= (F_1,\dots, F_m)\in \text{Fin}_s(\exp X)\) a sequence \(u= (U_1,\dots,U_k)\), \(k\geq m\), is called a \(K\)-neighbourhood of \(\Phi\) if \(F_j\subset U_j\) and there is an embedding of the nerve \(N(u)\subset K\). A set \(P\subset X\) is said to be a \(K\)-partition of \(\Phi\) (notation: \(P\in\text{Part}(\Phi,K)\)) if \(P= X-\bigcup u\) for some \(K\)-neighbourhood \(u\) of \(\Phi\). The dimension tr-\({\mathcal K}\)-Ind is defined as follows. Let \({\mathcal K}\) be a non-empty set of complexes. To every space \(X\) one assigns the transfinite dimension tr-\({\mathcal K}\)-Ind \(X\) which is \(-1\), an ordinal number \(\alpha\), or \(\infty\). tr-\({\mathcal K}\)-Ind \(=-1\) iff \(X= \varnothing\); tr-\({\mathcal K}\)-\(\text{Ind\,}X\leq\alpha\), for some ordinal \(\alpha\), if for every \(K\in{\mathcal K}\) and \(\Phi\in \text{Exp}_K(X)\) there exists a partition \(P\in\text{Part}(\Phi,K)\) such that tr-\({\mathcal K}\)-\(\text{Ind\,}P< \alpha\); tr-\({\mathcal K}\)-\(\text{Ind}=\infty\) iff tr-\({\mathcal K}\)-\(\text{Ind\,}X> \alpha\), for every ordinal \(\alpha\). We call a mapping \(f\) from a closed subset of a space \(X\) to a polyhedron \(L\) a partial mapping. A closed subset \(P\subset X\) is called a partition of a partial mapping \(f\) (notation: \(P\in\text{Part}(f,L)\)), if \(P\) is disjoint from \(\text{dom}(f)\) and \(f\) can be extended over \(X- P\). Then when we let \(L\) be a non-empty set of polyhedra, the definition of tr-\({\mathcal L}\)-Ind is like tr-\({\mathcal K}\)-Ind when we substitute partitions \(P\in\text{Part}(\Phi, K)\) by \(P\in\text{Part}(f,L)\). Using results from the earlier papers it is proven that when \({\mathcal L}_1\) is homotopically dominated by \({\mathcal L}_2\), then tr-\({\mathcal L}_1\)-\(\text{Ind\,}X\leq\)-tr-\({\mathcal L}_2\)-\(\text{Ind\,}X\) for every space \(X\); extending this result outside the class of hereditarily normal spaces to general spaces. Therefore, when \({\mathcal L}_1\) and \({\mathcal L}_1\) are homotopy equivalent, the equality tr-\({\mathcal L}_1\)-\(\text{Ind\,}X=\text{tr-}{\mathcal L}_2\text{-Ind\,}X\) holds. Another result is that tr-\({\mathcal L}\text{-Ind\,}X=\text{tr-Ind\,} X\) holds for every space \(X\) if and only if \({\mathcal L}\) contains a disconnected polyhedron. The results for the transfinite dimensions tr-\({\mathcal K}\)-Ind and tr-\({\mathcal L}\)-Ind hold simultaneously for the finite dimension functions \({\mathcal K}\)-Ind and \({\mathcal L}\)-Ind. A complex \(K\) is said to be an \(I(\psi)\)-complex (tr-\(I(\psi)\)-complex) (notation: \(K\in I(\psi)\)-\(c\) (\(K\in \text{tr}\)-\(I(\psi)\)-\(c)\)) when for any space \(X\in\psi\) we have: if \(K\)-\(\text{Ind\,}X<\infty\) then also \(\text{Ind\,}X<\infty\) (if tr-\(K\text{Ind\,}X<\infty\) then also tr-\(\text{Ind\,}X<\infty\)). Accordingly, the classes of polyhedra tr-\(I(\psi)\)-\(p\) and \(I(\psi)\)-\(p\) are defined using tr-\(L\)-\(\text{Ind\,}X\) and \(L\)-\(\text{Ind\,}X\). For the class \(\psi\) of all spaces we abbreviate to (tr-)\(I\)-\(c\) and (tr-)\(I\)-\(p\). Then \(I\)-\(c=\text{tr-}I\)-\(c= \{K: K\) is a disconnected complex\} and \(I\)-\(p=\text{tr}\)-\(I\)-\(p = \{K: K\) is a disconnected polyhedron\}. Under the assumption of the Continuum Hypothesis (CH) it is proven that there exists a perfectly normal compact space \(X\) such that \(\dim X= 1\) and tr-\(\text{Ind\,}X=\infty\). Therefore under CH for the class of all perfectly normal compact spaces we have (tr-)\(I\)-\(c = (\text{tr}\)-)\(I(pn\text{\,Comp})\)-\(c\) and (tr-)\(I\)-\(p= (\text{tr}\)-)\(I(pn\, \text{Comp})\)-\(p\).