Exercise 2.12

Consider a nonempty polyhedron P and suppose that for each variable xi we have either the constraint xi ≥ 0 or xi ≤ 0. Is it true that P has at least one basic feasible solution?

Answers

Yes. This is practically a baby version of Theorem 2.6 (Existence of extreme points). We demonstrate that P cannot contain a line, i.e., for any vector x ∈ P and any nonzero vector d ∈ ℜn we can find some λ ∈ ℜ such that x + λd∉P. To do so, we make use of the fact that the negativity-positivity constraints are linearly independent, i.e., assuming without loss of generality the following order of constraints, we have

[ ±1 … 0 ⋮ ⋱ ⋮ 0 … ±1 a 1,n+1…an,n+1 ⋮ ⋱ ⋮ a 1,m … an,m ] [ x1 ⋮ x n ] ≥ [ 0 ⋮ 0 b n+1 ⋮ b m ] .

Proof. Let x be an element of P and let I = {i∣xi = 0}. If n of the standard vectors ±ei, 1 ≤ i ≤ n, corresponding to the active negativity-positivity constraints are contained in I, then x is, by definition, a basic feasible solution and, therefore, a basic feasible solution exists. If this is not the case, then we can find at least one 1 ≤ j ≤ n such that xj≠0, i.e., either xj > 0 or xj < 0. But then it is obvious that

∀ ⁡d ∈ ℜnfor λ := −2xj dj : x − 2xj djd∉P

since the jth component of this vector xj − 2xj djdj = −xj is the opposite of xj and thus violates the j-th constraint. Thus, P cannot contain a line and must therefore, by Theorem 2.6, possess an extreme point. □

User profile picture
2021-12-01 13:51
Comments