Principle \(\mathrm{S}_1(\mathcal{P}, \mathcal{R})\): ideals and functions (Q1738946): Difference between revisions

In what follows, an \textit{ideal} on a set \(M\) is a family \(\mathcal{I} \subseteq \mathcal{P}(M)\) which is hereditary, i.e., \(B \in \mathcal{I}\) whenever \(B \subseteq A \in \mathcal{I}\), closed under finite unions, contains all finite subsets of \(M\) and \(M \notin \mathcal{I}\). Given a topological space \(X\), \(C_p(X)\) denotes the family of all real valued continuous functions defined on \(X\), endowed with the inherited topology from the Tychonoff product topology of\, \(^X \mathbb{R}\) (that is, the topology of pointwise convergence). Let \(\mathcal{I}\) be an ideal on \(\omega\). A sequence \(\langle x_n: n \in \omega\rangle\) of elements of a topological space \(X\) is said to be \textit{\(\mathcal{I}\)-convergent} to \(x \in X\) if the set \(\{n \in \omega: x_n \notin U\} \in \mathcal{I}\) for every neighbourhood \(U\) of \(x\). A sequence \(\langle f_n: n \in \omega\rangle\) of real functions on \(X\) is said to be a \textit{\(\mathcal{I}\)-monotone sequence} if \(\{n \in \omega: f_n \not\leqslant f_m\} \in \mathcal{I}\) for every \(m \in \omega\). Let \(\mathcal{P}\) and \(\mathcal{R}\) be families of sets. The selection principle \(S_1(\mathcal{P},\mathcal{R})\) asserts that for any sequence \(\langle \mathcal{U}_n: n \in \omega \rangle\) of elements of \(\mathcal{P}\) there is a set \(U_n \in \mathcal{U}_n\) for each \(n \in \omega\) such that \(\langle U_n: n \in \omega \rangle\) is a member of \(\mathcal{R}\). In the paper under review, the authors introduce ideal versions of some classes of topological spaces previously defined by Scheepers and Arhangel'skii (namely, Scheepers' \(S_1(\Gamma,\Gamma)\)-spaces, Arhangel'skii's \(\alpha_4\)-spaces and Scheepers' monotonic sequence selection property). For instance, given ideals \(\mathcal{I}\) and \(\mathcal{J}\) on \(\omega\) and a topological space \(X\), the ideal version of Scheepers' monotonic sequence selection property is given by \(S_1(\mathcal{I}-\Gamma_\mathbf{0}^m,\mathcal{J}-\Gamma_\mathbf{0})\), where \(\mathbf{0}\) denotes the constant zero-valued function on \(X\) and the families of sequences of functions \(\mathcal{J}-\Gamma_\mathbf{0}\) and \(\mathcal{I}-\Gamma_\mathbf{0}^m\) are given by \[ \mathcal{J}-\Gamma_\mathbf{0} = \{A \in\,^\omega(C_p(X) \setminus \{\mathbf{0}\}): A \,\,\text{is}\,\, \mathcal{J}\text{-convergent to}\,\,\mathbf{0}\} \] and \[ \mathcal{I}-\Gamma_\mathbf{0}^m = \{A \in\,^\omega(C_p(X) \setminus \{\mathbf{0}\}): A \,\,\text{is}\,\,\mathcal{I}\text{-monotone and}\,\, \mathcal{I}\text{-convergent to}\,\,\mathbf{0}\}. \] A large number of results on spaces satisfying the above mentioned ideal versions (and also on cardinal invariants related to them) are established in the paper.