Read up on Spectral Theory
Dym - Applied Functional Analysis
Complete Subspaces
We start with the proposition we ended with in the last lecture.
Proposition
i) Any closed subspace of a Banach/Hilbert space is complete, hence is also a Banach/Hilbert space.
ii) Any complete subspace is closed.
iii) The closure of a subspace is a subspace.
Proof
i) Since the subspace is closed, then every convergent sequence converges inside the set thus is complete.
ii) Since every Cauchy sequence in convergent and is contained in the subspace (and every convergent sequence is Cauchy) thus our space must be closed.
iii) Let , assume that (that is is a limit point and not in the original subspace) then whose limit is respectively. Then, and notice that:
that is it has the limit that is is closed under scalar multiplication and vector addition, that is it indeed is a linear subspace.
Examples of Incomplete subspaces
Definition
We would define to be the space of sequences finite length that is with for
It is clear that but notice that if we take the sequence (of finite sequences)
then however the limit
is not. Thus is not closed (and thus is not complete). However if we take the subspace:
Then is closed (and thus complete) in an incomplete space (that is in a complete space ).
Another example to consider is to take the subspace and consider the following sequence of functions:
It is continuous in however the limit
Is not continuous. Thus is another example of a space that is incomplete.
Definition
We define a Hilbert space to be the smallest complete inner product space containing , with the inner product defined to be: