On a theorem of F. Riesz (Q5892374): Difference between revisions

The author studies a well-known theorem of F. Riesz for Lorentz spaces generated by concave functions. More precisely, the following theorem is proved. Let \(\Phi :[0,+\infty)\to [0,+\infty)\) be a nondecreasing concave function such that \(\Phi (0)=0\), \(\mu \) a positive measure on a \(\sigma \)-algebra of subsets and \(N_{\Phi}(S)\) the Banach space of all measurable functions \(f\) with the norm \[ \| f\|_{N_{\Phi}}=\int_0^{\infty}\Phi (\lambda_f(t))\,dt, \text{ where } \lambda_f(t)=\mu \{x\in S:| f(x)| >t\}. \] Let \(\mu (S)<\infty \) or \(\Phi \) be unbounded. If \(f,f_m\in N_{\Phi}(S)\), \(\{f_m\}\to f\) a.e. and \(\| f_m\|_{N_{\Phi}}\to \| f\|_{N_{\Phi}}\), then \(\lim_{m\to \infty}\| f_m-f\|_{N_{\Phi}}=0\). Examples show that all the assumptions of the theorem are essential.