Badly approximable points on planar curves and winning (Q1684659)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Badly approximable points on planar curves and winning
scientific article

    Statements

    Badly approximable points on planar curves and winning (English)
    0 references
    0 references
    0 references
    0 references
    12 December 2017
    0 references
    Let \(i, j > 0\) with \(i+j=1\). A pair of real numbers \((x,y) \in\mathbb{R}^2\) is said to be \((i,j)\)-badly approximable if there is a constant \(c(x,y) > 0\) such that for any \(q \in\mathbb{N}\), \[ \max\{ \Vert qx \Vert^{1/i}, \Vert qy \Vert^{1/j}\} > c(x,y) q^{-1}. \] The set of all such pairs is denoted by \(\mathrm{Bad}(i,j)\). An inhomogeneous analogue is defined in the following way: Let \(\theta=(\gamma, \delta) \in\mathbb{R}^2\) and let \(\mathrm{Bad}_\theta(i,j)\) denote the set of \((x,y) \in\mathbb{R}^2\) for which constant \(c(x,y) > 0\) such that for any \(q \in\mathbb{N}\), \[ \max\{ \Vert qx - \gamma \Vert^{1/i}, \Vert qy - \delta \Vert^{1/j}\} > c(x,y) q^{-1}. \] It is known that these sets have maximal Hausdorff dimension. The intersection properties of these sets have been a subject of interest in recent years. One way of guaranteeing good intersection properties is by proving that the sets in question are winning for the so-called Schmidt game. In the paper under review, the authors show this to be the case for \(\mathrm{Bad}_\theta(i,j) \cap C_f\), whenever \(C_f = \{(x, f(x)) : x \in I\}\) for some interval \(I\) and some function \(f \in C^2(I)\) with \(f''(x) \neq 0\) for all \(x \in I\). In the special case when \(\theta = (0,0)\), this extends results of \textit{D. Badziahin} and \textit{S. Velani} [Math. Ann. 359, No. 3--4, 969--1023 (2014; Zbl 1360.11081)] and \textit{V. Beresnevich} [Invent. Math. 202, No. 3, 1199--1240 (2015; Zbl 1338.11060)]. Additionally, it is shown that if \(C_f\) is replaced by a straight line satisfying certain Diophantine properties, the same conclusion holds true. This extends another result of Badziahin and Velani [loc. cit.]. Finally, using variations on the same method, it is shown that \(\mathrm{Bad}_\theta (i,j)\) is itself a winning subset of \(\mathbb{R}^2\). The proof of the results depend on first establishing the homogeneous (\(\theta = (0,0)\)) result using arithmetical, combinatorial and geometrical methods previously developed for the study of the sets \(\mathrm{Bad}(i,j)\). Subsequently, these are transferred to the inhomogeneous setup.
    0 references
    0 references
    0 references
    0 references
    0 references
    inhomogeneous Diophantine approximation
    0 references
    non-degenerate curves
    0 references
    Schmidt games
    0 references
    0 references
    0 references