n + k usually divides C(2n, n) (Pomerance, Theorem 2) #
C. Pomerance, Divisors of the middle binomial coefficient, Amer. Math. Monthly 122 (2015) 636–644,
Theorem 2: for every positive integer k, the set of n with n + k ∣ C(2n, n) has asymptotic
density 1. For primes p ≥ 2k the power of p in n + k always divides C(2n, n) (inequality (4),
MiddleBinomial.padicValNat_add_le); for each of the finitely many primes p < 2k, the n where it
fails are few, because C(2n, n) has many factors p for most n (Lemma 2 of the paper: few n ≤ x
have few carries in base p) and few n ≤ x have p^j ∣ n + k with j large.
Main results #
MiddleBinomial.tendsto_density_add_dvd_centralBinom: for k >= 1 the proportion of n < x with n + k dividing C(2n, n) tends to 1 (Theorem 2)
References #
- C. Pomerance, Divisors of the middle binomial coefficient, Amer. Math. Monthly 122 (2015) 636–644, doi:10.4169/amer.math.monthly.122.7.636.
Counting residues in blocks #
Large digits #
Analysis #
Carries #
Per-prime bound #
theorem
MiddleBinomial.tendsto_density_add_dvd_centralBinom
(k : ℕ)
(hk : 1 ≤ k)
:
Filter.Tendsto (fun (x : ℕ) => ↑{n ∈ Finset.range x | n + k ∣ n.centralBinom}.card / ↑x) Filter.atTop (nhds 1)
Pomerance, Theorem 2. For every k ≥ 1, the proportion of n < x with n + k ∣ C(2n, n)
tends to 1.