REVIEW 4 major objections 5 minor 1 references
Factoring Catalan numbers
T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read For the n-th Catalan number, an odd prime p lies in the k-th Legendre layer precisely when n mod p^k is in the open interval (p^k/2, p^k−1), so the p-adic valuation v_p(Cat(n)) is a count of such k.
desk verdict Theorem 4.6 is correct but elementary, a direct restatement of Legendre's formula; the paper's advertised Segment method rests on an unproved and partly false interval generalization. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing mechanism is the layer indicator $\lfloor 2n/p^k\rfloor-\lfloor n/p^k\rfloor-\lfloor (n+1)/p^k\rfloor$ for odd $p$: it is $1$ precisely when the fractional part of $n/p^k$ is at least $1/2$ and $p^k$ does not divide $n+1$, which is exactly the interval condition $n \bmod p^k \in (p^k/2, p^k-1)$. Around this, the paper builds Chebyshev's Segments, prime intervals of the form $p((n+1)/t,\, 2n/(2t-1))$ in the single layer, generalized to all layers, which select groups of primes from interval boundaries rather than by testing each prime individually. The residue theorem then supplies an individual prime test that avoids enumerating empty Segments and avoids extracting roots.
What would settle it
For $n=100$ and $p=3$, evaluate the Legendre sum directly from factorials, $v_3(200!)-v_3(100!)-v_3(101!)=(66+22+7+2)-(33+11+3+1)-(33+11+3+1)=1$; if the number of $k$ satisfying $3^k/2 < 100 \bmod 3^k < 3^k-1$ is not $1$, the central theorem is wrong.
Extended reading notes
Core claim
On the paper's own terms, the discovery is a layer-by-layer residue criterion for the odd part of a Catalan number. Legendre's formula writes $v_p(\mathrm{Cat}(n))$ as $\sum_{k\ge 1}(\lfloor 2n/p^k\rfloor-\lfloor n/p^k\rfloor-\lfloor (n+1)/p^k\rfloor)$, and the paper shows each summand is $1$ for an odd prime $p$ exactly when $n \bmod p^k$ lies in $(p^k/2, p^k-1)$. Hence $v_p(\mathrm{Cat}(n))$ equals the number of $k$ satisfying that one residue inequality, and a prime appears as a repeated factor precisely by satisfying the inequality in several layers. The argument achieves this by comparing fractional parts, in effect modifying Kummer's carry rule, and the same layer picture is presented geometrically through Chebyshev's Segments.
Load-bearing premise
The load-bearing premise for the segmentation algorithm is the assertion that the generalized Chebyshev Segments partition each Legendre layer exactly; in Section 2 the corresponding interval description is only exact when $p^2 > 2n$, and the paper states the general version without proof.
Editorial extensions
If this is right
- For any odd prime $p \le 2n$, the exponent $v_p(\mathrm{Cat}(n))$ can be obtained by checking $k=1,2,\dots$ against one residue inequality, with no binomial coefficient or Catalan number written out.
- A repeated factor is visible as duplicate layer membership: a prime divides $\mathrm{Cat}(n)$ with exponent $m$ exactly when the inequality holds for exactly $m$ values of $k$.
- Small-prime divisibility tests become residue tests; for instance $5$ enters the single layer exactly when $n \equiv 3 \pmod 5$, with higher layers adding powers of $5$.
- Factorization can be split into a coarse pass that harvests large primes from Chebyshev Segments and a fine pass that checks individual small primes, and the paper implements this pipeline up to $n = 10^8$.
Reading between the lines
- A corollary the paper does not spell out is a squarefree test: $\mathrm{Cat}(n)$ has a squared odd prime factor exactly when the inequality holds for some $k \ge 2$, so repeated odd factors can be detected by checking primes only up to roughly $\sqrt{2n}$.
- Because the criterion uses only $n \bmod p^k$, it suggests a sieve-like computation of valuations across many $n$ for a fixed prime, with no per-$n$ factorization.
- The layer-residue form is close to Kummer's carry count, which raises a testable question: whether the same interval test continues to hold for related quotients such as $(1/(n+a))\binom{2n}{n}$ or other generalized Catalan families.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes a layer-by-layer scheme for factoring Catalan numbers. It partitions odd primes into Legendre layers by the exponent k such that p^k divides the Catalan number, and subdivides each layer into intervals called Chebyshev Segments whose endpoints are explicit rational/algebraic functions of n and t. The central new result is Theorem 4.6: an odd prime p falls into the k-th Legendre layer iff n mod p^k lies in the open interval (p^k/2, p^k - 1). The paper also gives algorithms to locate a prime inside a layer via segment enumeration, an example for n=9999 and n=10^8, and advertises an online HTML factoring service for n up to 10^8.
Significance. If the segmentation machinery were fully justified, the paper would offer a practical, memory-lean way to enumerate the prime factors of very large Catalan numbers without computing the integer itself. The layer criterion in Theorem 4.6 is genuine, elementary, and exactly checkable; it is a crisp reformulation of Legendre's formula with a clean residue-interval characterization, and it deserves to be recorded. The paper ships explicit worked examples, a standalone derivation of the valuation formula, and a self-contained proof route for Theorem 4.6. However, the advertised computational method rests on the segment decomposition whose soundness is not established here, so the significance of the full package is conditional.
major comments (4)
- [Section 2, around (2.2)] The assertion that A_t = p(2n/(2t+1), 2n/(2t-1)) is stated for primes with v_p(A)=t, i.e. v_p(1·3·...·(2n-1))=t. This is false in general. For n=10, p=3, A=1·3·...·19 and v_3(A)=4, but 3 lies in the interval (20/9, 20/7)≈(2.22,2.86), not in the asserted A_4. The error occurs precisely when p^2≤2n, so the segmentation formulas (2.2a) and (2.2b), while they may still identify membership in the single layer L^{(1)}, are not derived from the stated premise.
- [Theorem 3.1] The displayed Segment formula S_t^(k)(n)=p(((n+1)/t)^(1/k), (2n/(2t-1))^(1/k)) is the load-bearing generalization of the segmentation to arbitrary Legendre layers, but it is stated without proof. The supporting text only points at the k=1 and k=2 cases, and the k=2 formula is itself written with missing radical notation in the manuscript. Since Algorithm 3.3 and the claimed factorizing software in Section 5 both depend on this theorem, the central computational claim—that the k-th layer is exactly the union of these segment intervals—is not rigorously supported.
- [Example 2.4] The last line of the example reports the check for p=29 and then prints 'we get 23 ∈ S435.' This is a copy-paste error (should reference 29) and, together with the p=23 line, leaves the example inconsistent. A reader cannot tell which prime the last check refers to.
- [Example 4.1a] The notation 'v2 (Cat (10^8)) ≥ 2' is wrong on two counts: the layer membership test for p=13 in the SQUARE-layer shows that 13^2 divides the Catalan number, so the valuation of 13—not of 2—is at least 2. The manuscript also defines v_p as the exponent of p in the factorization, so 'v2(Cat(...)) ≥ 2' is simply mislabeled.
minor comments (5)
- [Abstract and Section 5] The phrase 'index up to 108' and 'the 108-th Catalan number' omit the exponent notation; it should be 10^8.
- [Example 4.7] The modular residues are written as '37', '57', '77', '117', '137' where the intended values are p^7; the computations are otherwise traceable, but the typography obscures the verification.
- [Example 3.2] For the SQUARE-layer segment S_4^(2)(9999)=p(50;53.44) the displayed set is {51, 53}; since 51 is not prime, this is a transcription error.
- [General exposition] The manuscript uses both 'Chebyshev’s Segments' and informal terminology ('Black Holes', 'fishing') that is not standard; the informal passages do not affect the mathematics but make the exposition hard to follow for the general reader.
- [Section 1.2] The claim that primes in p(n+1; 2n) satisfy p^2 ∤ Cat(n) is true for p>n+1, but the remark immediately above it ('prime factors of Cat(n) are less than 2n') would benefit from quoting the standard binomial-coefficient argument rather than stating it.
Circularity Check
No circularity: Theorem 4.6 is derived algebraically from Legendre's formula, and the self-citations are not load-bearing.
full rationale
The central derivation is self-contained against the standard Legendre formula. Starting from (3.2), the k-th summand for v_p(Cat(n)) is simplified to floor(2r/q)-floor((r+1)/q) with q=p^k and r=n mod q; this expression equals 1 exactly when q/2 < r < q-1, which is Theorem 4.6. This is an algebraic rewrite of an external standard result, not a conclusion defined into the premise, and no fitted parameters are used. The citations [Er16] and [Er19] are historical or supplementary and do not carry the proof. The unproved status and apparent error of the A_t interval in Section 2 and Theorem 3.1 are substantive correctness gaps in the segmentation method, but they do not exhibit the specific reduction pattern required for circularity. Although Theorem 4.6 can be described as a modular reformulation of Legendre's formula, the paper explicitly proves the equivalence, so this is legitimate derivation rather than circular reasoning.
Assumptions & free parameters
assumptions (3)
- standard math Legendre's formula v_p(n!) = sum over k of floor(n/p^k)
- ad hoc to paper The set A_t = p(2n/(2t+1), 2n/(2t-1)) equals primes with v_p(A)=t
- ad hoc to paper Theorem 3.1: the k-th layer equals the union of intervals S_t^(k)(n) = p(((n+1)/t)^(1/k), (2n/(2t-1))^(1/k))
Cite this review
Pith. "Pith review of Factoring Catalan numbers." pith.science (2026). https://pith.science/paper/BXP2B6VJ
@misc{pith2026190803752,
author = {Pith},
title = {Pith review of: Factoring Catalan numbers},
year = {2026},
howpublished = {\url{https://pith.science/paper/BXP2B6VJ}},
note = {Machine review of arXiv:1908.03752}
}
abstract
The paper describes a prime factorization of the Catalan numbers. Odd prime factors are distributed in layers in accordance with Legendre's formula. The content of each layer is a network of the intervals, Chebyshev's Segments. The primes of Segment are not calculated and are selected on the basis of its bounds. Layers contain non-repeated primes. Repeated factors are formed when primes are duplicated among different layers. The paper slightly modifies Kummer's theorem for the selection of individual prime factors, also starting from the boundaries of Segments. In conclusion, the reader is offered a software service for factorization of the Catalan number with index up to $10^8$
Reference graph
Works this paper leans on
-
[1]
[A108] Neil J. A. Sloane. The On-Line Encyclopedia of Integer Sequences. https://oeis.org/A000108 [DS18] Emeric Deutsch and Bruce E. Sagan. Congruences for Catalan and Motzkin numbers and related sequences. https://arxiv.org/pdf/math/0407326.pdf [Er16] Gennady Eremin. Multilayer Factorization of Catalan Numbers. https://studyres.com/doc/13013409 [Er19] Ge...
work page Pith review arXiv 2015
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.