Topics in Functional Analysis (Ausgewählte Kapitel aus Funktionalanalysis):
Locally Convex Vector Spaces (Lokalkonvexe Vektorräume)
Class Nummer: 250074
Class type: VO (lecture course)
Semester hours: 3
Schedule: Mon, Tue, Thu, 12:00-13:00 Seminarraum 2A310, UZA 2
Start: 6.10.2008
News:
- 2009-01-30: A complete set of my notes can be found on the board
outside my office.
Exams are possible by individual arrangement anytime starting from end of Jannuary.
I will only be away Feb. 18-23 and end of April.
- 2009-01-29: Joint going out on Fr. 30th, 19:30 at 'The Highlander'
- 2009-01-07: The first lecture in Jannuary is Monday 12th.
- 2008-11-11: No lecture on Nov. 13.
- 2008-10-22: There will be no lecture in the first week (1st, 2nd, and 4th)
of December.
- 2008-10-20: There will be no lecture in the first week (3rd, 4th, and 6th)
of November.
Introduction: Generally speaking, functional analysis
is that part of analysis which uses the power of topology to derive
results on function spaces and on operators acting between them.
Clearly such results are vital for applications in wide areas
of analysis, not the least in partial differential equations.
Most treatises start with vector spaces where a topology is induced by a
norm,
giving rise to the theory of Banach spaces. In particular, one derives the
fundamental theorems of Hahn-Banach (on the extension of linear functionals),
Banach-Steinhaus (also called uniform boundedness principle) and the
open mapping as well as the closed graph theorem with Banach's isomorphism theorem
as a corollary (which states that any bijective continuous operator of Banach
spaces already is an isomorphism).
However, many of the function spaces ocurring in analysis cannot be turned into
(complete) normed spaces in a reasonable way. Some of these can at least be
given the structure of a (complete) metric space (called Frechet space, eg.
the spaces of continuous or smooth functions on the real line) which
is characterized by the fact that their topology is induced by a countable family
of semi-norms. Other spaces, in particular those used in
distribution theory
(most prominently the space of test functions, i.e., the space of smooth functions
with compact support), ask for an even more general theory as they cannot be
turned into Frechet spaces.
On the other end of the scale one has topological vector spaces, i.e.,
vector spaces carrying a topology such that the operations (of addition
and scalar multiplication) are continuous. These spaces, however, turn out to
be too general for many purposes, that is to say that they do not allow to prove
the strong statements one is interested in.
As a matter of fact, a very general but still widely useful class is comprised by
those topological vector spaces that possess a fundamental system of neighborhoods of
the origin consisting of convex sets. Alternatively these so-called locally convex vector spaces
can be characterized as those topological vector spaces that allow for their topology to be
generated by a (not necessarily countable) family of semi-norms. The existence of a convex
base of zero-neighbourhoods is strong enough for the Hahn-Banach theorem to hold, yielding
a sufficiently rich theory of continuous linear functionals and allows, in particular,
for a toplogical theory of the function spaces relevant in distribution theory.
In fact, historically, the space of test functions with the appropriate notion of convergence
was introduced by L. Schwartz before a proper topological description was available. It was the
success of distribution theory which stimulated the development of the theory of locally convex vector
spaces and later on both theories evolved in parallel with significant mutual impact.
A slightly extended version of the above text is available
here.
Contents: This course will introduce you to the theory of locally
convex vector spaces. The main topics are
- topological vector spaces
- locally convex vector spaces and vector spaces with semi-norms
- projective and inductive limits
- nuclear spaces
- duality
Literature: I have prepared a short
commented list of books on the topic,
although we will not strictly follow one of these. The most I am influenced
by earlier lecture courses by
Michael Grosser, which themselves were influenced by H. Schaefer's "Topological Vector
Spaces" (2nd ed., Springer GTM, New York, 1999).
What you already should know:
Prerequisites for this course are
- a good knowledge of basic functional analysis---approximately a three-hours course, eg.
Funktionalanalysis 1. (An advanced course on Banach or Hilbert space
theory, in particular spectral theory, is not needed, although you will benefit from some preknowledge
on weak convergence and related matters. In this sense this course is _not_ a direct follow up of summer
term's Functional Analysis 2.)
- a solid basis in general topology---approximately a two-hours course, eg.
Grundbegriffe der Topologie. (Nevertheless we will briefly recall more advanced
topics of topology at the place where they are first needed.)
Finally, as always, a good background in analysis is profitable.
Target audience: This course is primarily designed
for advanced diploma students or master students, but
also Ph.D. students will benefit from it. In particlar, people
knowing distribution theory but who have never been
introduced into the topological aspects of the theory might
be interested.
Related Courses:
Günther Hörmann
and I plan to teach a joint course on
the theory of distributions in summer term. It will not be designed as continuation
of this course but will be quite related.
Administrative Remarks: The course will normally be held 4 hours (i.e., 4x45=180 mins.)
a week but suspended some weeks on (duly) announcement. The language of instruction will
be English (since I expect some foreign students attending).
Position within curriculum:
Diplomstudium Mathematik, 2. Abschnitt, Studienschwerpunkt Analysis.
Exams: oral; by arrangement. Further informations will be given on time.