Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated


Statement

Problem 4.17 (p. 76, quoted). "It is shown by Salem and Zygmund [697] that there exist positive constants C3,C4C_3,C_4 such that for every positive ε\varepsilon, we have

(C3−ε)(nlog⁡n)12<max⁡∣z∣=1∣P(z)∣<(C4+ε)(nlog⁡n)12(C_3-\varepsilon)\big(n\log n\big)^{\frac12}<\max_{|z|=1}|P(z)|<(C_4+\varepsilon)\big(n\log n\big)^{\frac12}

apart from o(2n)o(2^n) polynomials P(z)P(z). Is this result true with C3=C4C_3=C_4 and if so, what is the common value?"

The polynomials are those of Problem 4.13 (p. 75), P(z)=∑k=1nεkzkP(z)=\sum_{k=1}^n\varepsilon_kz^k with each εk=∓1\varepsilon_k=\mp1, of which there are 2n2^n. The book's [697] is R. Salem and A. Zygmund, Some properties of trigonometric series whose terms have random signs, Acta Math. 91 (1954), 245--301. The problem carries no attribution line, and Table 2 (p. 253) lists it among the problems of the 1967 edition.

Update 4.17 (p. 76). The update says Halász (the book's [362]: G. Halász, On a result of Salem and Zygmund concerning random polynomials, Studia Sci. Math. Hungar. 8 (1973), 369--377) proved the conjecture with C3=C4=1C_3=C_4=1, together with the analogous result for trigonometric polynomials.

Source. W. K. Hayman and E. F. Lingham, Research Problems in Function Theory, arXiv:1809.07200v2 (21 September 2018), Chapter 4, p. 76. The edition read is identified on the source card.

Read depth. Claims checked: the problem, its update and the two cited reference entries were read clause by clause on the printed page. The book proves nothing; it poses and reports.

Proof pointer

None; a problem. Halász's paper is on the Halász card.

Dependencies

None.

Bears on

  • Problem 523: Problem 4.17 asks, in the form "apart from o(2n)o(2^n) polynomials", for a common constant in the nlog⁡n\sqrt{n\log n} size of the maximum modulus on the circle; #523 asks for it almost surely, with the sum from k=0k=0. Update 4.17 credits Halász with the common value 11 in the book's form; the problem page records what his paper proves for the almost-sure form.