Géza Freud and lacunary Fourier series (Q1089864): Difference between revisions

From MaRDI portal
Added link to MaRDI item.
RedirectionBot (talk | contribs)
Removed claim: reviewed by (P1447): Item:Q411449
Property / reviewed by
 
Property / reviewed by: Q748816 / rank
Normal rank
 

Revision as of 20:06, 14 February 2024

scientific article
Language Label Description Also known as
English
Géza Freud and lacunary Fourier series
scientific article

    Statements

    Géza Freud and lacunary Fourier series (English)
    0 references
    0 references
    1986
    0 references
    The author considers lacunary Fourier series with Hadamard gaps \(f(x)=\sum^{\infty}_{k=0}r_ k(\cos (n_ kX+\phi_ k),\) where f is bounded on an interval, \(r_ k\geq 0\), \(n_ 0=0\), \(n_ 1\geq 1\) and \(\inf_{k\geq 1}(n_{k+1}/n_ k)=q>1\). He discusses the connection between the behavior of f at a point \(x_ 0\) and the asymptotic behavior of \(n_ kr_ k\) as \(k\to \infty.\) The following result is due to G. Freud: If \(f\) is differentiable at one point, then \(r_ kn_ k\to 0\) as \(k\to \infty\). A converse result says: If \(r_ kn_ k\to 0\) as \(k\to \infty\), \(f\) is differentiable on a dense set. The author states the following precision: Suppose that \(0<\alpha \leq r_ kn_ k\leq \beta <\infty\) and \(n_ k\leq Q^ k\) \((k=1,2,...).\) Then with \(Df(x,h):=f(x+h)-f(x),\) \(f\) is nowhere differentiable, \(\limsup_{h\to 0}| Df(x,h)/h| <\infty\) for a dense set of x,\(Df(x,h)=O(| h| \log 1/| h|)\) and \(Df(x,h)+Df(x,- h)=O(| h|)\) (h\(\to 0)\) uniformly on [-\(\pi\),\(\pi\) ], \(\lim\sup_{h\to 0}| Df(x,h)/(| h| \log 1/| h|))| >0\) for every \(x\) except on a set of the first category of Baire, \[ \limsup_{h\to 0+}| Df(x,h)| /(h\sqrt(\log 1h \log\log\log 1/h))<\infty \] for almost every x. Further, he gives a measure theoretical interpretation of the theorem and explains connections between lacunary and random series.
    0 references
    lacunary Fourier series
    0 references
    Hadamard gaps
    0 references
    random series
    0 references

    Identifiers