On a new integral-type operator from the Bloch space to Bloch-type spaces on the unit ball (Q1018297)

Let \(\mathbb{B}\) be the unit ball of \(\mathbb{C}^n\) and \(\mu : [0,1) \to [0, \infty)\) be a continuous function satisfying the following normal conditions: there exist \(0<a<b\) and \(\delta \in [0,1)\) such that, as \(r \rightarrow 1\) on the interval \([\delta, 1)\), the function \(\frac{\mu(r)}{(1-r)^\gamma}\) decreases to \(0\) for \(\gamma=a\) and increases to \(+\infty\) for \(\gamma=b\), respectively. The Bloch-type space \(\mathcal{B}_\mu(\mathbb{B})\) is the Banach space of analytic functions \(f:\mathbb{B} \to \mathbb{C}\) such that \[ \| f \|_{\mathcal{B}_\mu} = |f(0)| + \sup_{z \in \mathbb{B}} \mu(|z|) \big| \mathfrak{R}f(z) \big| < \infty, \] where \(\mathfrak{R}f\) is the radial derivative of \(f\). There is a corresponding little Bloch-type space \(\mathcal{B}_{\mu,0}(\mathbb{B})\) with the same norm, for which additionally \(\lim_{|z| \to 1} \mu(|z|) \big| \mathfrak{R}f(z) \big| = 0\). The standard Bloch spaces \(\mathcal{B}\), \(\mathcal{B}_0\) are the special case \(\mu(r) = 1-r^2\) (although, in this special case, the equivalent norm with \(\nabla f\) is used instead of \(\mathfrak{R}f\)). Consider linear integral operators of the form \[ P^g_{\varphi}(f)(z) = \int_{t=0}^1 f(\varphi(tz)) g(tz) \frac{dt}{t} \qquad (z \in \mathbb{B}), \] where \(f,g:\mathbb{B} \to \mathbb{C}\) and \(\varphi : \mathbb{B} \to \mathbb{B}\) are all analytic, with \(g(0)=0\). These and similar operators (and some important special cases) between various spaces have been considered in many papers, by the author and others. In the present paper, complete results are obtained for boundedness, compactness and essential norms for all cases in which \(P^g_{\varphi}\) maps from \(\mathcal{B}\) or \(\mathcal{B}_0\) into \(\mathcal{B}_\mu\) or \(\mathcal{B}_{\mu,0}\). For example, \[ \| P^g_{\varphi} \|_{\mathcal{B} \to \mathcal{B}_\mu} = \| P^g_{\varphi} \|_{\mathcal{B}_0 \to \mathcal{B}_\mu} = \sup_{z \in \mathbb{B}} \mu(|z|) \big| g(z) \big| \max\left\{ 1, \frac{1}{2}\log \frac{1+|\varphi(z)|}{1-|\varphi(z)|} \right\}. \] There appears to be a very minor unimportant misprint, which does not affect the results and proofs: it seems that \(\lim\) should be replaced by \(\limsup\) in (22), (23) and just after (25).