Continued fractions and Gauss class number problem for real quadratic fields (Q447814): Difference between revisions

The authors put forward some numerical data to support the truth of Gauss' conjecture that there are infinitely many real quadratic fields of class number one. Let \(d\) be a square free positive integer and let \(\varepsilon=\frac{t+u\sqrt{d}}{2}\) be the fundamental unit of the quadratic field of discriminant \(d\). Then the authors define the Yokoi invariant as the positive integer \(m=[\frac{u^2}{t}]\). Let \(l\) be the period length of the continued fraction expansion of the quadratic surd \(w\) associated to \(d\) (\(w\) is either \(\sqrt{d}\) or \(\frac{\sqrt{d}+1}{2}\)). Note that the authors generalize the definition of the Yokoi invariant to general \(d\), not necessarily square free that we omit here. In one of their first results (Theorem 2.1), they prove inequalities connecting the Yokoi invariant to the fundamental unit that enables them to conclude that when the Yokoi invariant is large, then so is the fundamental unit. It follows then by Siegel's result that the class number \(h\) is small. They work with ``discriminants \(d\) of minimal type'' the definition of which uses the partial quotients of the continued fraction of \(w\), rather technical to give here. They show that the Yokoi invariants of positive integers that are not of minimal type is less than or equal to \(3\). Moreover under certain conditions, the Yokoi invariant of \(d\) of minimal type is large. These \(d\) then by the results mentioned above have small class number. In their experimental data they group discriminants \(d\) with the same value of \(l\) and pick out the minimal element \(d_l\) for each \(l\). One of their conjectures is that if \(l\neq 7, 11, 49, 225, 299\) or \(1032\), then \(d_l\) is square free with class number \(1\). They give other similar conjectures by looking at the list corresponding to the exceptional values above and picking out values of minimal type that have class number \(1\).