Separating sets by upper semicontinuous and Darboux upper semicontinuous functions (Q1928436): Difference between revisions

The pairs of sets \(A_0, A_1\subset\mathbb R\) which can be expressed as \(A_0=f^{-1}(\{0\}), A_1=f^{-1}(\{1\})\), or as \(A_0=f^{-1}((-\infty,0)), A_1=f^{-1}((1,\infty))\) respectively, for some upper semicontinuous Darboux function \(f:\mathbb R\to\mathbb R\) are characterized. The proofs of the necessity of the characterizing conditions use results of \textit{A.~Maliszewski} [Fundam. Math. 175, No. 3, 271--283 (2002; Zbl 1017.26002)], the construction of separating functions uses transformation of a bilaterally \(\mathfrak c\)-dense set to a set open with respect to the density topology by \textit{W. J.~Gorman III} [Proc. Am. Math. Soc. 17, 825--830 (1966; Zbl 0147.04405)] and the existence of upper semincontinuous approximately continuous functions separating a clopen set from its complement. Characterizations of pairs which can be separated, i.e., there is an upper semicontinuous Darboux function \(f\) with inclusions instead of the above equalities, are derived. Corresponding results concerning general upper semicontinuous functions are also included.