The non-kk-gonal pyramidal number formula for k9k\geq 9

From papers

Let k9k\geq 9, and let the nn-th non-kk-gonal pyramidal number be the nn-th positive integer that is not a kk-gonal pyramidal number. The preceding formula defines this sequence in terms of the integer cube root of 6n/(k2)6n/(k-2). Non-kk-gonal pyramidal number formula. For k9k\geq 9, the nn-th non-kk-gonal pyramidal number is given by the formula in Eq.. The claim extends the stated formula beyond the range 3k83\leq k\leq 8 established earlier in the paper; the supplied text gives no proof or resolution for k9k\geq 9.

Progress summary

Open

The proposed formula for numbers excluded from the kk-gonal pyramidal sequence remains unproved when there are nine or more sides.

The claim extends a formula already proved for 3k83\leq k\leq 8 to every k9k\geq 9. The source explicitly labels this extension as Conjecture 1, not as a theorem.

Known results

  • Theorem 9 proves the formula for 3k83\leq k\leq 8.
  • For other values of kk, the relevant formulas are established only for sufficiently large nn, namely nn0(k)n\geq n_0(k) for some n0(k)>0n_0(k)>0.

Current evidence

The retrieved literature continues to present the k9k\geq 9 statement as a conjecture; no independently documented proof, counterexample, verification, or model-based solution claim was found.

Current status (as of August 2026): The formula is proved for 3k83\leq k\leq 8 and eventually valid for each fixed kk in the broader range, but its validity for every nn when k9k\geq 9 remains open.

Sources
Sources & referencesView supporting material

Primary source

Chai Wah Wu, “Algorithms for complementary sequences”, arXiv:2409.05844 (2025).

Solutions 1

Proof

Proof, valid for every k3k\geq 3

Let

Pk(m)=m(m+1)(m(k2)(k5))6P_k(m)=\frac{m(m+1)\bigl(m(k-2)-(k-5)\bigr)}{6}

be the mm-th kk-gonal-pyramidal number, and let ak(n)a_k(n) be the nn-th positive integer not among these numbers. Set

h=6nk23.h=\left\lfloor\sqrt[3]{\frac{6n}{k-2}}\right\rfloor.

I will prove the following three-branch formula from Equation (5):

  • ak(n)=n+h+1a_k(n)=n+h+1 if
6n(k2)h3+3(k1)h2+(2k1)h+6;6n\geq (k-2)h^3+3(k-1)h^2+(2k-1)h+6;
  • ak(n)=n+h1a_k(n)=n+h-1 if
6nh(h1)((k2)h+k+1);6n\leq h(h-1)\bigl((k-2)h+k+1\bigr);
  • ak(n)=n+ha_k(n)=n+h otherwise.

Write c=k21c=k-2\geq1, and put Qm=Pk(m)mQ_m=P_k(m)-m, with the auxiliary values Pk(0)=Q0=0P_k(0)=Q_0=0. Then

6Pk(m)=m(m+1)(c(m1)+3).6P_k(m)=m(m+1)\bigl(c(m-1)+3\bigr).

Its first difference is

6(Pk(m+1)Pk(m))=3(m+1)(cm+2)>0,6\bigl(P_k(m+1)-P_k(m)\bigr) =3(m+1)(cm+2)>0,

so the pyramidal numbers are strictly increasing.

Suppose ak(n)=n+ra_k(n)=n+r. Exactly rr pyramidal numbers precede this value, so

Pk(r)<n+r<Pk(r+1).P_k(r)<n+r<P_k(r+1).

Because all quantities are integers, this is equivalent to

Qr<nQr+1.(1)Q_r<n\leq Q_{r+1}. \qquad (1)

Conversely, (1) says that exactly rr pyramidal and nn complementary values occur through n+rn+r. Thus (1) uniquely characterizes the required offset rr.

The definition of hh gives

ch36n<c(h+1)3.(2)ch^3\leq6n<c(h+1)^3. \qquad (2)

For h1h\geq1, direct expansion gives

ch36Qh1=(c1)h(3h2)+7h6>0.ch^3-6Q_{h-1} =(c-1)h(3h-2)+7h-6>0.

For h0h\geq0, similarly,

6Qh+2c(h+1)3=(h+1)(3ch+5c+3h+6)>0.6Q_{h+2}-c(h+1)^3 =(h+1)(3ch+5c+3h+6)>0.

Together with (2), these inequalities show that

Qh1<n<Qh+2.Q_{h-1}<n<Q_{h+2}.

Consequently, the offset rr can only be h1h-1, hh, or h+1h+1.

It remains to identify which one occurs. The lower expression in the stated formula satisfies

h(h1)(ch+c+3)=6Qh.h(h-1)(ch+c+3)=6Q_h.

Therefore its condition is exactly nQhn\leq Q_h, which, by (1), selects r=h1r=h-1.

The upper expression satisfies

ch3+3(c+1)h2+(2c+3)h+6=6Qh+1+6.ch^3+3(c+1)h^2+(2c+3)h+6 =6Q_{h+1}+6.

Its condition is therefore Qh+1<nQ_{h+1}<n, which selects r=h+1r=h+1.

If neither condition holds, then

Qh<nQh+1,Q_h<n\leq Q_{h+1},

so (1) selects r=hr=h.

Finally, when h=0h=0, the upper threshold is 66, so its condition holds automatically because n1n\geq1; it correctly gives r=1r=1.

Substituting c=k2c=k-2 and ak(n)=n+ra_k(n)=n+r yields exactly the three branches above. Hence Conjecture 1 holds, and its range strengthens from k9k\geq9 to every k3k\geq3.

Source: Chai Wah Wu, “Algorithms for complementary sequences,” Equation (5) and Conjecture 1: https://arxiv.org/abs/2409.05844

0 endorsements
Samuel Schlesinger ·