Homepage › Solution manuals › Walter Rudin › Functional Analysis › Exercise 1.2
Exercise 1.2
The convex hull of a set in a vector space is the set of all convex combinations of members of , that is the set of all sums in which , ; is arbitrary. Prove that the convex hull of a set is convex and that is the intersection of all convex sets that contain .
Answers
Proof. The convex hull of a set
will be denoted by
. Remark that
(to see that, take
for each
in
) and that
where
(obvious).
Our proof will directly derive from (i)
(iv) in the following lemma,
Let be a subset of a vector space : Its convex hull is convex and the following statements
- (i)
- is convex;
- (ii)
- for all positive scalar variables ;
- (iii)
- for all positive scalar variables such that ;
- (iv)
are equivalent.
From now on, we skip the trivial case then only consider nonempty sets. To prove the first part, let , range over , so that and for some . Every sum ( ) is then in the convex hull of , since
and
In terms of sets , this reads
which was our fist goal. We now aim at the equivalence (i)
(iv)
(i): An easy proof by induction makes the implication (i)
(ii) directly come from (d) of the above exercise 1, chapter 1. (iii) is a special case of (ii), and the implication (iii)
(iv) derives from the definition of the convex hull. We now close the chain with (iv)
(i), by remarking that
is convex whether
. The lemma being proved, let us establish the second part.
To do so, we start from the convexity of
then set
. We may enrich
as follows,
Note that our initial predicate “[ only encompasses] all convex sets that contain A", is now the special case
In any case, the key ingredient is that implies
Conversely, the next formula
is valid and implies
So ends the proof □