Counting numbers that are divisible by the product of their digits

Qizheng He,Carlo Sanna
2024-04-03
Abstract:Let $b \geq 3$ be a positive integer. A natural number is said to be a base-$b$ Zuckerman number if it is divisible by the product of its base-$b$ digits. Let $\mathcal{Z}_b(x)$ be the set of base-$b$ Zuckerman numbers that do not exceed $x$, and assume that $x \to +\infty$. First, we prove an upper bound of the form $|\mathcal{Z}_b(x)| < x^{z_b^{+} + o(1)}$, where $z_b^{+} \in (0,1)$ is an effectively computable constant. In particular, we have that $z_{10}^+ = 0.665{\scriptstyle\ldots}$, which improves upon the previous upper bound $|\mathcal{Z}_{10}(x)| < x^{0.717}$ due to Sanna. Moreover, we prove that $|\mathcal{Z}_{10}(x)| > x^{0.204}$, which improves upon the previous lower bound $|\mathcal{Z}_{10}(x)| > x^{0.122}$, due to De Koninck and Luca. Second, we provide a heuristic suggesting that $|\mathcal{Z}_b(x)| = x^{z_b + o(1)}$, where $z_b \in (0,1)$ is an effectively computable constant. In particular, we have that $z_{10} = 0.419{\scriptstyle\ldots}$. Third, we provide algorithms to count, respectively enumerate, the elements of $\mathcal{Z}_b(x)$, and we determine their complexities. Implementing one of such counting algorithms, we computed $|\mathcal{Z}_b(x)|$ for $b=3,\dots,12$ and large values of $x$ (depending on $b$), and we showed that the results are consistent with our heuristic.
Number Theory
What problem does this paper attempt to address?
This paper attempts to solve the counting problems related to Zuckerman numbers. Specifically, the author studies natural numbers that satisfy specific arithmetic constraints, especially those natural numbers that are divisible by the product of their digits when represented in a certain base (i.e., Zuckerman numbers). The main problems include: 1. **Upper bound estimation**: Prove that when \(x \to +\infty\), the upper bound of the number \(|Z_b(x)|\) of base - \(b\) Zuckerman numbers is \(x^{z^+_b + o(1)}\), where \(z^+_b\in(0, 1)\) is a computable constant. In particular, for \(b = 10\), the previous upper bound \(|Z_{10}(x)|<x^{0.717}\) is improved, and the new upper bound is \(z^+_{10}=0.665...\) 2. **Lower bound estimation**: Prove that the lower bound of \(|Z_b(x)|\) is \(x^{0.204}\), which is better than the previous result \(|Z_{10}(x)|>x^{0.122}\) 3. **Heuristic hypothesis**: Propose a heuristic hypothesis that \(|Z_b(x)|=x^{z_b + o(1)}\), where \(z_b\in(0, 1)\) is a computable constant. For \(b = 10\), \(z_{10}=0.419...\) 4. **Algorithm design**: Provide algorithms for counting and enumerating the elements of \(Z_b(x)\) and determine the time complexity of these algorithms. The time complexity of the optimal general - purpose counting algorithm is \(x^{z^*_b+o(1)}\), where \(z^*_b\in(0, 1)\) is a computable constant. In summary, the main objective of this paper is to make more accurate estimates of the number of Zuckerman numbers through rigorous mathematical analysis and algorithm design and to provide effective computational methods.