← All problems

Erdős Problem #304 — Egyptian fraction lengths with distinct denominators

45 years old · documented by

For integers b≥2b\ge 2 and 1≤a<b1\le a<b, let N(a,b)N(a,b) be the least positive integer kk for which there are integers n1,…,nkn_1,\ldots,n_k satisfying 1<n1<⋯<nk1<n_1<\cdots<n_k and

ab=∑i=1k1ni.\frac{a}{b}=\sum_{i=1}^{k}\frac{1}{n_i}.

Define

N(b):=max⁡1≤a<bN(a,b),N(b):=\max_{1\le a<b}N(a,b),

where the maximum is over integer aa. The question is to estimate N(b)N(b) as bb grows, in particular whether there are absolute constants C>0C>0 and b0b_0 such that

N(b)≤Clog⁡log⁡bN(b)\le C\log\log b

for every integer b≥b0b\ge b_0. Erdős Problems #304

Background

These representations are Egyptian fractions: sums of distinct unit fractions. The practice is ancient; the Rhind Papyrus, dating to around 1650 B.C., includes examples of such decompositions. Erdős’s 1950 paper made the maximal length needed to represent a rational with fixed denominator into a quantitative problem and proposed the order of growth asked about here. Graham, Paul Erdős and Egyptian Fractions

The question is listed as Erdős–Graham, 1980, p. 37. Its close connection with the separate problem of finding unit-fraction decompositions of 11 with prescribed denominator features is also noted in later work on that topic; in particular, the boundary case N(b−1,b)N(b-1,b) is singled out. Erdős Problems #304 The smallest denominator not contained in a unit fraction decomposition of 1 with fixed length

References

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.