Counting cocomponents of a topological space (Q702204): Difference between revisions

Motivated by the duality between clone and coclone theory in universal algebra a topological space \(X\) is said to have at most \(n\) cocomponents if for every continuous mapping \(f : X^{n+1} \to X\) there exists an \(I \in [\{1,\dots,n+1\}]^{\leq n}\) such that \(f(x) = f(y)\) whenever \(x,y\in X^{n+1}\) and \(x_i = y_i\) for each \(i\in I\), i.e. if every continuous \(f : X^{n+1} \to X\) depends on at most \(n\) coordinates. Topological spaces with at most one cocomponent were called coconnected in [\textit{V. Trnková}, Serdica Math. J. 24, No. 1, 25--36 (1998; Zbl 0940.54028)]. Here the following results are proved. Theorem 1. For any natural number \(n\) and for any Hausdorff space \(X\) the following properties are equivalent: (i) \(X\) has at most \(n\) cocomponents; (ii) there exists an \(m > n\) such that every continuous mapping \(f : X^m \to X\) depends on at most \(n\) coordinates; (iii) for every natural number \(m\), every continuous mapping \(f : X^m \to X\) depends on at most \(n\) coordinates. Theorem 2. If a Hausdorff space \(X\) has at most \(n\) cocomponents, then, for every natural number \(m\), the space \(X^m\) has at most \(n\) cocomponents. Example 1. There exists a metrizable space \(X\) which does not have finitely many cocomponents but every continuous mapping \(f : X^\omega\to X\) depends on only finitely many coordinates. Example 2. There exist a metrizable space \(X\) of the form \(X = A \times B\) with \(A\) and \(B\) rigid and a continuous mapping \(f : X^\omega\to X\) which does not depend on only finitely many coordinates. Example 3. For every natural number \(n\) there exists a metrizable space \(X_n\) which has precisely \(n\) cocomponents but cannot be expressed as \(A\times B\) with \(|A| > 1\) and \(|B| > 1\).