An inequality for sums of binary digits, with application to Takagi functions (Q542846): Difference between revisions
From MaRDI portal
Added link to MaRDI item. |
Removed claim: author (P16): Item:Q294285 |
||
Property / author | |||
Property / author: Pieter C. Allaart / rank | |||
Revision as of 17:56, 12 February 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | An inequality for sums of binary digits, with application to Takagi functions |
scientific article |
Statements
An inequality for sums of binary digits, with application to Takagi functions (English)
0 references
20 June 2011
0 references
Let \(\phi(x)=2\inf\{|x-n|\;:\;n\in \mathbb{Z}\}\). For \(\alpha>0\) let \(f_\alpha(x)=\sum\limits_{j=0}^\infty2^{-\alpha j}\phi(2^jx)\). In particular, \(\frac{1}{2}f_1\) is Takagi's continuous nowhere differentiable function. \textit{J. Tabor} and \textit{J. Tabor} [J. Math. Anal. Appl. 356, No. 2, 729--737 (2009; Zbl 1188.26008)] proved that \[ f_\alpha\Big(\frac{x+y}{2}\Big)\leq\frac{f_\alpha(x)+f_\alpha(y)}{2}+|x-y|^\alpha, \] for all \(\alpha\in[1,2]\) and \(x,y\in[0,1]\). By developing an explicit expression for \(f_\alpha\) at dyadic rational points, the author shows that the above inequality can be reduced to a simple inequality for weighted sums of binary digits. This simple inequality (which is also of independent interest) is used to give an alternative proof of the result of Tabor and Tabor, which provides more insight into the essential structure of the function \(f_\alpha\).
0 references
Takagi function
0 references
approximate convexity
0 references
digital sum inequality
0 references