Lecture 7
What is the main problem?
I. Consider following examples. Do computations according to the instructions, Observe your results and take notes. Your possible questions should be clarified after this course.
,
,
. Fix
compute
and
. Are they equal?
,
. Show that
converges, denote the limit
. Is
continuous? Are
's continuous?
, for
. Compute
, call it
,
. Does
converge to
?
on
. Compute the limit of
for fixed
, call the limit function
. Integrate
from
to
. Compute also integral of
from
to
for fixed
, then take the limit. What is your observation?
These examples are discusson on when two limit processes can be interchanged. More precisely, we ask if the following things are true:
The above examples show that notion of the limit of for fixed
in a metrc space (called pointwise convergence) is too weak to guarantee the limit interchange mentioned here: In words, it does not preserve continuity, it does not preserve limits, it does not preserve integrals. This motivates a new notion of convergence, which is called uniform convergence.
II. Now we discuss these two notions of convergence of a sequence of functions. We should compare uniform with pointwise convergence:
- For pointwise convergence we could first fix a value for
and then choose
. Consequently,
depends on both
and
.
- For uniform convergence
must be uniformly close to
for all
in the domain. Thus
only depends on
but not on
.
Let's illustrate the difference between them graphically:
Figure below left: (pointwise convergence) First fix a value . Then we choose an arbitrary neighborhood around
, which corresponds to a vertical interval centered at
. Finally we pick
so that
intersects the vertical line
inside the interval
Figure below right: (uniform convergence) We draw an -neighborhood around the entire limit function
, which results in an "
-strip" with
in the middle. Now we pick
so that
is completely inside that strip for all
in the domain.
As an immediate consequence of the defition we have the following useful criterion
Theorem. Suppose (
). Let
. Then
uniformly on
if and only if
as
.
Note that constructing a counterexample to show a statement is not true is part of mathematical training.