The MacLane class and complex differential equations in the unit disk (Q837598): Difference between revisions

The authors study the interaction between the analytic or meromorphic coefficients and solutions of the linear differential equation \[ f^{(k)}(z)+a_{k-1}(z)f^{(k-1)}(z)+\cdots + a_1(z)f'(z)+ a_0(z)f(z)=0 \tag{*} \] in the unit disc \(\mathbb{D}=\{z\in\mathbb{C}:|z|<1\}\) of the complex plane \(\mathbb{C}\). In addition, Riccati differential equations are also studied. The main results are Theorems~3, 5 and 6. Theorems~3 and 5 concern the connection between the growth of solutions and the growth of coefficients of (*). Theorem~3 consists of the following two statements: (i) Suppose all solutions \(f\) of (*) satisfy \[ \int_0^1\log^+\log^+M(r,f)\,dr<\infty, \tag{2} \] where \(M(r,f)=\max_{|z|=r}|f(z)|\). Then for \(j=0,\ldots,k-1\), each analytic coefficient \(a_j(z)\) satisfies \[ \int_0^1\log^+\log^+M(r,a_j)\,dr<\infty. \] (ii) Suppose all analytic coefficients \(a_j(z)\), \(j=0,\dots,k-1\), of (*) satisfy \[ \int_0^1M(r,a_j)\,dr<\infty. \] Then all solutions \(f\) of (*) satisfy (2). Statements (i) and (ii) are related to the MacLane class \(\mathcal{A}\) because the condition (2) is known to imply \(f\in\mathcal{A}\). Recall that the MacLane class consists of nonconstant analytic functions in \(\mathbb{D}\) with asymptotic values at each point of a dense set on the unit circle \(\mathbb{T}=\{z\in\mathbb{C}:|z|=1\}\). Theorem~5 is of the same nature as Theorem~3, but it concerns the equation (*) with meromorphic coefficients and solutions with few zeros only, and is related to the meromorphic counterpart of the MacLane class. Theorem~6 concerns Riccati differential equations and reads as follows: Let \(G\) be a domain in the finite \(z\)-plane and let \(f(z)\) be analytic in \(G\). Let \(b\not=\infty\) be a boundary point of \(G\), and let \(a_0(z)\), \(a_1(z)\) and \(a_2(z)\) be given functions, analytic in some disk \(\Delta_0={z:|z-b|<r}\). Let \(\Gamma:z=\psi(u)\), \(0\leq u\leq1\), be a continuous curve such that \(\psi(1)=b\) and \(\Gamma-\psi(1)\subset G\cap\Delta_0=G_0\). Define \[ \phi(z)=a_0(z)+a_1(z)f(z)+a_2(z)f^2(z)-f'(z). \tag{3} \] Consider the differential equation \[ w'(z)-a_1(z)w(z)=a_2(z)w^2(z)+(a_0(z)-\phi(z)). \tag{4} \] Suppose that \(g(z)\) is a solution of (4) such that \(g(z)\) is differentiable in \(\Delta_0\). If (3) with \(z\in G_0\) satisfies \(\phi(\psi(u))\to\lambda\not=\infty\) as \(u\to1^-\), then \(f\) has an asymptotic value on \(\gamma\), either finite or infinite.