A remark on invariant operator ranges (Q803516): Difference between revisions

Foias has proved the following theorem: Let A be a weakly closed subalgebra of B(H) which contains an m.a.s.a. S. Let \(L=tH\) be an invariant operator range of A with \(t\in S\), \(t\geq 0\). Finally, let \(\{e_{\alpha}\}_{\alpha \geq 0}\) be the spectral scale of t. Then there exists a constant C, \(0<C<+\infty\) such that \[ \bigvee_{a\in A}a(1-e_{\alpha})H\subset (1-e_{\alpha /C})H \] for all \(\alpha\geq 0.\) The author proves the reverse of Foias' theorem as follows: Theorem. Let \(A\subset B(H)\) be an algebra and \(t\in B(H)\), \(t\geq 0\). Let \(\{e_{\alpha}\}_{\alpha \geq 0}\) be the spectral scale of t. If there is a \(C\geq 0\) such that \(\bigvee_{a\in A}a(1-e_{\alpha})H\subset (1- e_{\alpha /C})H,\) then \(t^{1/2}H\in Lat_{}(A).\) (Lat\({}_{}(A)\) denotes the collection of all operator ranges which are invariant for every element of A).