Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Source. Croot, published paper, p. 233, Lemma 1 and equation (1). Croot attributes this to Lemma 3.1 of Canfield, Erdős and Pomerance, “On a problem of Oppenheim concerning ‘factorisatio numerorum’,” Journal of Number Theory 17 (1983), 1–28. That paper numbers no Lemma 3.1. Its Section 3 prints an unnumbered lemma (p. 9), Theorem 3.1 (p. 10, a lower bound only), and an unnumbered two-sided Corollary (p. 15), which the paper itself cites on p. 7 as the Corollary to Theorem 3.1. Lemma 1 follows from that Corollary with , for which . The canonical published source and its uniform smooth-number corollary are already filed separately. Their analytic proof remains external here.
Define, for and ,
and
Exact external input. For each fixed ,
Equivalently, for every , the count eventually lies between and . No uniformity in an arbitrary varying is asserted. The lower-bound construction uses monotonicity between fixed nearby values of to handle its particular varying parameters.
Proof scope. This page records the precise analytic theorem used by Croot. The external Canfield–Erdős–Pomerance proof has not been reconstructed here. The prime-power variant following (1) is a separate, complete relative deduction.
Source corrections. Both displayed definitions of and on p. 233 print where is required. Taken literally, they would not depend on and could not satisfy (1). The arXiv v1 and the author manuscript date reference [1] to 1980; the published bibliography gives 1983.
Bears on. Theorem 1, the construction, and Problem 202.