Gallai’s conjecture on certain families of bipartite graphs

For every connected graph GG with n=∣V(G)∣n=|V(G)| vertices, the edges of GG admit a decomposition into at most ⌈n2⌉\left\lceil\frac{n}{2}\right\rceil paths.

References

Primary source

La Matematica

Additional references

Progress summary

Refreshed
Claimed progress

A published paper reports progress for some bipartite graph families, but the general conjecture and the point–triple cases remain unsettled.

Gallai’s conjecture asks whether every connected graph can have its edges partitioned into at most half its vertices’ worth of paths, rounded up. The general conjecture remains open.

Known results and recent publication (date not stated)

  • Complete bipartite graphs satisfy the conjectured bound.
  • Connected 22-degenerate graphs satisfy it, with a stronger bound except for triangles.
  • The order-one Levi graphs L1(m,k)L_1(m,k) satisfy the stronger bound p(G)≤⌊n/2⌋p(G)\le\lfloor n/2\rfloor for m≥2m\ge 2 and 2≤k≤m2\le k\le m; for k=2k=2, equality is claimed.
  • Sahu and Padinhatteeri’s article, “Gallai’s Conjecture on Certain Families of Bipartite Graphs,” appeared in La Matematica; its accessible publication record gives no exact online date or theorem statement, while the associated preprint reports the L1(m,k)L_1(m,k) result.

Community submission (unverified)

Posted September 9, 2026. A submitted argument studies the point–triple Levi graphs Ln(m,3)L_n(m,3), claims exact path counts in some congruence classes of mm, and gives a ten-path decomposition for m=6m=6. It identifies the remaining cases with a uniform triple-terminal routing problem; none of these claims has independent verification.

Current status (as of September 2026): The conjecture is claimed for several graph classes, including L1(m,k)L_1(m,k), but remains open in general; the point–triple claims are unverified.

Sources

Solutions 1

ProofGALLAI'S PATH DECOMPOSITION FOR POINT–TRIPLE LEVI GRAPHSSee full solutionHide full solution

We study Gallai's path-decomposition problem for the point–triple Levi graph Ln(m,3). Its bipartition consists

of the points of

mm

and the 3-subsets of

mm

, with incidence as adjacency. The graph has m + C(m,3) vertices

and 3C(m,3) edges. The parity of the point degree C(m-1,2) splits the problem modulo 4. When m is

congruent to 0 or 3 modulo 4, every vertex is odd and the classical odd-degree path-decomposition theorem

gives the exact value p(Ln(m,3)) =

m+C(m,3)m + C(m,3)

/2. For m congruent to 1 or 2 modulo 4, the odd vertices are

precisely the triple vertices, yielding p(Ln(m,3)) at least C(m,3)/2. We verify equality explicitly for m=6 by a

ten-path decomposition. The remaining problem is formulated as a uniform triple-terminal routing problem.

Keywords: Gallai conjecture; path decomposition; path number; Levi graph; bipartite graph; incidence

graph; set systems; combinatorics.

  • Gallai_L2_Point_Triple_Levi_Manuscript.pdf35,891 bytesOpen
  • Gallai.pdf4,673,258 bytesOpen
  • Finite_Classical_Communication_Simulation_Publication_Manuscript_Dr_Arie_Ariadne_Dewatson_Ledetsambali.pdf84,941 bytesOpen
  • Gallai_Point_Triple_Levi_Corrected_Final_Manuscript.pdf78,123 bytesOpen