On some polynomial values of repdigit numbers (Q485508): Difference between revisions

Let \[ f_{k,m}(x)=\frac{x(x+1)\ldots(x+k-2)((m-2)x+k+2-m)}{k!} \] be the \(m\)th order \(k\)-dimensional polygonal number, where \(k\geq 2\) and \(m\geq 3\) are fixed integers. As special cases for \(f_{k,3}\) we get the binomial coefficient \(\binom{x+k-1}{k}\), for \(f_{2,m}(x)\) and \(f_{3,m}(x)\) we have the corresponding polygonal and pyramidal numbers, respectively. Another important class of combinatorial numbers is the numbers of the form \(d\cdot \frac{10^n-1}{10-1}\), \(1\leq d\leq 9\). They are called repdigits and for \(d=1\), repunits. These numbers above have already been investigated from several aspects and therefore have a rich literature. In this paper the authors studied the equal values of repdigits and the \(k\)-dimensional polygonal numbers. They stated some effective finiteness theorems and for small parameter values they completely solved the corresponding equations. A common generalization of repdigits and generalization repunits are numbers of the form \[ d\cdot \frac{b^n-1}{b-1}, \] i.e. taking repdigits with repeating digit \(d\) in the number system of base \(b\), where \(1\leq d<b\) and \(b\geq 2\) are integers. Consider the equation \[ d\cdot \frac{b^n-1}{b-1}=f_{k,m}(x). \] One of their main effective results is the following theorem: Theorem [Theorem 2.1] Suppose that \(k\geq 3\) or \(k=2\) and \(m=4\) or \(m>13\). Then the equation above has only finitely many integer solutions in \(x\) and \(n\), further, \[ \max(| x| ,n)<c, \] where \(c\) is an effectively computable constant depending on \(k\), \(m\), \(b\) and \(d\). For \(k=2\) and \(m\in\{3,5,6,7,8,9,10,11,12\}\) the equation above has infinitely many solutions for infinitely many values of the parameters \(b\), \(d\). For other interesting results see the paper in details.