Exercise 6.2.6

Assume f n f on a set A . Theorem 6.2.6 is an example of a typical type of question which asks whether a trait possessed by each f n is inherited by the limit function. Provide an example to show that all of the following propositions are false if the convergence is only assumed to be pointwise on A . Then go back and decide which are true under the stronger hypothesis of uniform convergence.

(a)
If each f n is uniformly continuous, then f is uniformly continuous.
(b)
If each f n is bounded, then f is bounded.
(c)
If each f n has a finite number of discontinuities, then f has a finite number of discontinuities.
(d)
If each f n has fewer than M discontinuities (where M N is fixed), then f has fewer than M discontinuities.
(e)
If each f n has at most a countable number of discontinuities, then f has at most a countable number of discontinuities.

Answers

(a)
False pointwise when f n ( x ) = 1 1 + n x 2 and f ( x ) = { 1 if  x = 0 0 otherwise

Now suppose ( f n ) converges to f uniformly. We have

| f ( x ) f ( y ) | | f ( x ) f n ( x ) | + | f n ( x ) f n ( y ) | + | f n ( y ) f ( y ) |

By uniform convergence to f , we can fix n large enough so that the first and last terms are both less than 𝜖 3 . Then since f n is uniformly continuous, we can choose δ so that | x y | < δ implies the middle term is less than 𝜖 3 , and that | f ( x ) f ( y ) | < 𝜖 , implying f is uniformly continuous.

(b)
False pointwise when f n ( x ) = { x if  x < n 0 otherwise and f ( x ) = x

Now suppose ( f n ) f uniformly. We want to show f is bounded. Let M n bound f n , ie. | f n ( x ) | < M n for all x A .

Set 𝜖 = 1 and apply the Cauchy Criterion to get N so m n > N implies

| f n ( x ) f m ( x ) | < 1

Setting n = N and rearranging gives

| f m ( x ) | < | f N ( x ) | + 𝜖 < M N + 𝜖

implying f m is bounded by M N when m N .

Now set Ñ > N large enough that m N implies | f ( x ) f m ( x ) | < 1 which, after rearranging gives

| f ( x ) | < 1 + | f m ( x ) | < 1 + M N

Implying f is bounded.

(c)
False for pointwise and uniform convergence. Define f n ( x ) = { 1 m if  x = 1 m  and  m n m N 0 otherwise

and

f ( x ) = { 1 m if  x = 1 m m N 0 otherwise

The reason ( f n ) f uniformly is

| f ( x ) f n ( x ) | = ( { 1 m if  x = 1 m  and  m > n m N 0 otherwise ) < 1 ( n + 1 )

Each f n has exactly n point discontinuities, but f has countably many. Thus this is a counterexample.

Intuitively f n adds on finer and finer details of f , which is why it converges uniformly. But discontinuities can be as small/detailed as we want without screwing up uniform convergence.

(d)
False pointwise since we can have each ( f n ) continuous (zero discontinuities) but have f not be continuous (see (a) for an example).

Now suppose ( f n ) f uniformly. Let x 0 be a discontinuity of f , meaning there exists an 𝜖 0 such that | f ( x 0 ) f ( x ) | > 𝜖 0 no matter how small | x x 0 | is. I’d like to show x 0 is a discontinuity of f n for some n , i.e. that there exists an 𝜖 0 > 0 such that | f n ( x 0 ) f n ( x ) | > 𝜖 0 no matter how small | x x 0 | is.

Pick 𝜖 < 𝜖 0 2 and set N large enough that n N implies | f n ( x ) f ( x ) | < 𝜖 . Applying the three way triangle inequality gives

𝜖 0 < | f ( x 0 ) f ( x ) | | f ( x 0 ) f n ( x 0 ) | + | f n ( x 0 ) f n ( x ) | + | f n ( x ) f ( x ) | < 2 𝜖 + | f n ( x 0 ) f n ( x ) |

Letting 𝜖 0 = 𝜖 0 2 𝜖 > 0 we see | f n ( x 0 ) f n ( x ) | > 𝜖 0 meaning f n is not continuous at x 0 for all n N .

Now given discontinuities of f , applying the above process multiple times shows that eventually f n will have every discontinuity of f , but f n has at most M discontinuities, implying f has at most M discontinuities.

(e)
False pointwise when (using a modified version of Thomae’s function for f n ) f n ( x ) = { 1 if  x = 0 n n + q if  x = p q  (in lowest terms) 0 otherwise and f ( x ) = { 1 if  x Q 0 if  x Q

Now suppose ( f n ) f uniformly. In (d) we showed every discontinuity of f is eventually a discontinuity of f n ; reworded this is saying (where D f is the set of discontinuities of f )

D f n = 1 D f n

Since each D f n is countable, this implies D f is countable.

User profile picture
2022-01-27 00:00
Comments