Some remarks on a class of meromorphic starlike functions (Q810667): Difference between revisions

Let S(p) denote the class of functions f(z) analytic in the unit disk \(D=\{z:| z| <1\}\) except for a single pole at \(z=p\), \(0<p<1\) and normalized by \(f(0)=f'(0)-1=0\). \textit{J. E. Miller} [Proc. Am. Math. Soc. 80, 607-613 (1980; Zbl 0451.30008)] introduced the starlike subclass \(S^*(p,w_ 0)\) of S(p) as the set of all functions f(z)\(\in S(p)\) satisfying the inequality \[ Re\{\frac{zf'(z)}{f(z)-w_ 0}\}<0 \text{ for } \rho <| z| <1\text{ and for some } \rho,\quad p<\rho <1. \] Miller also shows the necessary and sufficient condition for \(f(z)=S^*(p,w_ 0)\) to be \[ Re\{\psi (f,z,p,w_ 0)\}=Re\{\frac{zf'(z)}{f(z)-w_ 0}+\frac{p}{z-p}-\frac{pz}{1- pz}\}<0\text{ for } | z| <1. \] The authors define a starlike subclass \(\sigma^*(p,w_ 0)\) of S(p) geometrically. A function f(z)\(\in S(p)\) is said to be a member of the class \(\sigma^*(p,w_ 0)\) if C-f(D) is the starlike set with respect to \(w_ 0\). The authors provide two types of necessary and sufficient conditions for functions to belong to \(\sigma^*(p,w_ 0)\) as follows: Theorem. The following are equivalent: a) \(f(z)\in \sigma^*(p,w_ 0);\) b) f(z)\(\in S(p)\) and satisfy the inequality \(Re\{\psi (f(z),z,p,w_ 0)\}<0\) (z\(\in D);\) c) \(f(z)=w_ 0+\frac{pw_ 0}{(z-p)(1-pz)}\exp \{\int_{| x| =1}2\log (1-xz)d\mu (x)\}\) here \(\mu\) is a probability measure on D, \(\mu\) and \(w_ 0\) satisfy the equation \(w_ 0=-\{p+(1/p)-2\int_{| x| =1}x d\mu (x)\}^{-1}.\) The general extremal problem for the family \(ST(p)=\cup_{w_ 0}\sigma^*(p,w_ 0)\) is obtained by using Golusin's variational formulas and standard technique of proofs, for example see \textit{B. Pinchuk} [Duke Math. J. 35, 724-734 (1968; Zbl 0167.361)]: Let \(J(f)=Re F(f(\zeta),f'(\zeta),...,f^{(n)}(\zeta),w_ 0,\zeta)\), \(f(z)\in \sigma^*(p,w_ 0)\). Then the functional J(f) attains its maximum (or minimum) over the class ST(p) only for the following functions: \[ f(z)=w_ 0+\frac{pw_ 0}{(z-p)(1- pz)}\sum^{m}_{j=1}(1-x_ jz)^{2\lambda_ j},\quad w_ 0=- \{p+\frac{1}{p}-2\sum^{m}_{j=1}\lambda_ jx_ j\}^{-1}, \] when \(m\leq n+2\), \(| x_ j| =1\), \(\lambda_ j\geq 0\) and \(\sum^{m}_{j=1}\lambda_ j=1\). In other words, the measure \(\omega\) corresponding to the extremal functions are a step one with at mot \(n+2\) jump parts.