On the approximation of numbers, connected with elliptic functions (Q1813397): Difference between revisions
From MaRDI portal
Created a new Item |
Added link to MaRDI item. |
||
links / mardi / name | links / mardi / name | ||
Revision as of 00:21, 30 January 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | On the approximation of numbers, connected with elliptic functions |
scientific article |
Statements
On the approximation of numbers, connected with elliptic functions (English)
0 references
25 June 1992
0 references
Let \(\wp(z)\) be the Weierstrass elliptic function; let \(g_ 2\), \(g_ 3\) be its invariants; let \(2\omega_ 1\), \(2\omega_ 2\) be an arbitrary fixed pair of its fundamental periods; let \(\zeta(z)\) be the Weierstrass \(\zeta\)-function connected with \(\wp(z)\); and let \(\eta_ k=\zeta(\omega_ k)\), \(k=1,2\). We denote by \(\xi_ i\), \(i=1,2,3\), approximating algebraic numbers; by \(L_ i\) and \(n_ i\) the length and the degree of the number \(\xi_ i\), and by \(\mathbb{A}\) the set of all algebraic numbers. We assume that \(\xi_ i\neq 0\). If among the numbers \(\omega_ 1\), \(\omega_ 2\), \(\eta_ 1\), \(\eta_ 2\) there exists an algebraic one, then at least one of the invariants of \(\wp(z)\) is a transcendental number. An estimate of simultaneous approximation of the invariants of the function \(\wp(z)\) with algebraic periods has been obtained in the author's paper [Diophantine approximations 2, Collect. Artic., Moskva 1986, 114-121 (1986; Zbl 0654.10033)]. In this paper we prove the following Theorem: Let \(\eta_ 2\in\mathbb{A}\), \(n=\deg \mathbb{Q}(\xi_ 1,\xi_ 2,\xi_ 3)\), \[ \begin{aligned} M&= n\left(\min(n_ 2,n_ 3) \left( {{\ln L_ 2} \over {n_ 2}} + {{\ln L_ 3} \over {n_ 3}} +1\right)+ {{\ln L_ 1} \over {n_ 1}} +\ln n\right),\\ N&= n\left( \min(n_ 2,n_ 3) \left( {{\ln L_ 2} \over {n_ 2}} + {{\ln L_ 3} \over {n_ 3}} +1\right)+\ln \left(n\left(1+ {{\ln L_ 1} \over {n_ 1}}\right) \right)\right).\end{aligned} \] If there exists \(C>0\) such that for all \(t,x,y\in\mathbb{Z}\), \(x^ 2+y^ 2\neq 0\), \(| x|, | y| <t\), we have \[ \min| x\eta_ 1+y\eta_ 2|>\exp (-Ct^ 3),\tag{\(*\)} \] then there exists an effective constant \(\Lambda=\Lambda(\omega_ 1,\omega_ 2)>0\) such that \[ |\eta_ 1- \xi_ 1|+ | g_ 2-\xi_ 2|+ | g_ 3-\xi_ 3|> \exp(-\Lambda MN). \] We mention that condition \((*)\) does not follow from the properties of the Weierstrass \(\zeta\)-function (for example, one of the numbers \(\eta_ 1\), \(\eta_ 2\) may be equal to zero). In the case when \(\eta_ 1,\eta_ 2\in \mathbb{A}\), from the above formulated theorem we obtain the following result: Corollary. Let \(\eta_ 1,\eta_ 2\in \mathbb{A}\), \(n'=\deg \mathbb{Q}(\xi_ 2,\xi_ 3)\), \(N_ 1=n'(\min(n_ 2,n_ 3) ({{\ln L_ 2} \over {n_ 2}}+{{\ln L_ 3} \over {n_ 3}}+1)+\ln n')\). If for all \(x,y\in\mathbb{Z}\), \(x^ 2+y^ 2\neq 0\), we have \(x\eta_ 1+y\eta_ 2\neq 0\), then there exists an effective constant \(\Lambda_ 1= \Lambda_ 1(\omega_ 1,\omega_ 2)>0\), such that \[ | g_ 2-\xi_ 2| + | g_ 3-\xi_ 3|>\exp(- \Lambda_ 1 N^ 2_ 1). \]
0 references
transcendence
0 references
Weierstrass elliptic function
0 references
invariants
0 references
fundamental periods
0 references
Weierstrass \(\zeta\)-function
0 references
simultaneous approximation
0 references