Jump to content

McMullen problem

From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by David Eppstein (talk | contribs) at 08:27, 18 November 2014 (Statement: missing word). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Unsolved problem in mathematics:
For how many points is it always possible to projectively transform the points into convex position?

The McMullen problem is an open problem in discrete geometry named after Peter McMullen.

Statement

In 1972, McMullen has proposed the following problem:[1]

Determine the largest number such that for any given points in general position in affine d-space Rd there is a projective transformation mapping these points into convex position (so they form the vertices of a convex polytope).

Equivalent formulations

Gale transform

Using the Gale transform, this problem can be reformulate as:

Determine the smallest number such that every set of points X = {x1, x2, ..., xμ(d)} in linearly general position on Sd-1 it is possible to choose a set Y = {ε1x1,ε2x2,...,εμ(d)xμ(d)} where εi = ±1 for i = 1, 2, ..., μ(d), such that every open hemisphere of Sd−1 contains at least two members of Y.

The number , are connected by the relationships

Partition into nearly-disjoint hulls

Also, by simple geometric observation, it can be reformulate as:

Determine the smallest number such that for every set X of points in Rd there exists a partition of X into two sets A and B with

The relation between and is

Projective duality

An arrangement of lines dual to the regular pentagon. Every five-line projective arrangement, like this one, has a cell touched by all five lines. However, adding the line at infinity produces a six-line arrangement with six pentagon faces and ten triangle faces; no face is touched by all of the lines. Therefore, the solution to the McMullen problem for d = 2 is ν = 5.

The equivalent projective dual statement to the McMullen problem is to determine the largest number such that every set of hyperplanes in general position in d-dimensional real projective space form an arrangement of hyperplanes in which one of the cells is bounded by all of the hyperplanes.

Results

This problem is still open. However, the bounds of are in the following results:

  • Larman proved that . (1972)[1]
  • Michel Las Vergnas proved that . (1986)[2]
  • Alfonsín proved that . (2001)[3]

The conjecture of this problem is , and it is true for d=2,3,4.[1][4]

References

  1. ^ a b c L. G. Larman(1972), "On Sets Projectively Equivalent to the Vertices of a Convex Polytope", Bull. London Math. Soc. 4, pp.6–12
  2. ^ M. Las Vergnas (1986), "Hamilton Paths in Tournaments and a Problem McMullen on Projective Transformations in Rd", Bull. London Math. Soc. 18, pp.571–572
  3. ^ J. L. Ramírez Alfonsín(2001), "Lawrence Oriented Matroids and a Problem of McMullen on Projective Equivalences of Polytopes", Europ. J. Combinatorics 22, pp.723–731
  4. ^ D. Forge, M. Las Vergnas and P. Schuchert(2001), "A Set of 10 Points in Dimension 4 not Projectively Equivalent to the Vertices of Any Convex Polytope", Europ. J. Combinatorics 22, pp.705–708