In the previous sections, we have considered real-number sequences, that is, sequences such that for each . In this section, we consider sequences whose terms are functions. Sequences of functions arise naturally in many applications in physics and engineering. A typical way that sequences of functions arise is in the problem of solving an equation in which the unknown is a function . In many of these types of problems, one is able to generate a sequence of functions through some algorithmic process with the intention that the sequence of functions converges to the solution . Moreover, it would be desirable that the limiting function inherit as many properties possessed by each function such as, for example, continuity, differentiability, or integrability. We will see that this latter issue is rather delicate. In this section, we develop a notion of the limit of a sequence of functions and then investigate if the fundamental properties of boundedness, continuity, integrability, and differentiability are preserved under the limit operation.
Pointwise Convergence
Let be a non-empty subset and suppose that for each we have a function . We then say that is a sequence of functions on .
Let and let for and . Then is a sequence of functions on . As another example, for and let . Then is a sequence of functions on . Or how about
where and .
Let be a sequence of functions on . For each fixed we obtain a sequence of real numbers by simply evaluating each at , that is, . For example, if and we fix then we obtain the sequence . If is fixed we can then easily talk about the convergence of the sequence of numbers in the usual way. This leads to our first definition of convergence of function sequences.
Let be a sequence of functions on . We say that converges pointwise on to the function if for each the sequence converges to the number , that is,
In this case, we call the function the pointwise limit of the sequence .
By uniqueness of limits of sequences of real numbers (Theorem 3.1.12), the pointwise limit of a sequence is unique. Also, when the domain is understood, we will simply say that converges pointwise to .
Consider the sequence defined on by . For fixed we have
Hence, converges pointwise to on . In Figure 8.1, we graph for the values and the function . Notice that and therefore , and for the limit function we have . Hence, the sequence of derivatives converges pointwise to . Also, after some basic computations,
and therefore
On the other hand it is clear that .
Graph of for and
Before considering more examples, we state the following result which is a direct consequence of the definition of the limit of a sequence of numbers and the definition of pointwise convergence.
Let be a sequence of functions on . Then converges pointwise to if and only if for each and each there exists such that for all .
As the following example shows, it is important to note that the in Lemma 8.1.4 depends not only on but in general will also depend on .
Consider the sequence defined on by . For all we have and therefore . On the other hand if then
Therefore, converges pointwise on to the function
In Figure 8.2, we graph for various values of . Consider a fixed . Since it follows that for there exists such that for all . For , in order for we can choose . Notice that clearly depends on both and , and in particular, as get closer to then a larger is needed. We note that each is continuous while is not.
Graph of for
Example 8.1.5 also illustrates a weakness of pointwise convergence, namely, that if is a sequence of continuous functions on and converges pointwise to on then is not necessarily continuous on .
Consider the sequence defined on by . For fixed we have
Hence, converges pointwise on to the function . Notice that each function is continuous on and the pointwise limit is also continuous. After some basic calculations we find that
and exists for each , in other words, is differentiable on . However, is not differentiable on since does not have a derivative at . In Figure 8.3, we graph for various values of .
Graph of for and
Example 8.1.6 illustrates another weakness of pointwise convergence, namely, that if is a sequence of differentiable functions on and converges pointwise to on then is not necessarily differentiable on .
Consider the sequence on defined by . For fixed we find (using l'H\^{o}pital's rule) that
Hence, converges pointwise to on . Consider
and therefore
On the other hand, . Therefore,
or another way to write this is
Examples 8.1.5- 8.1.6 illustrate that the pointwise limit of a sequence of functions does not always inherit the properties of continuity and/or differentiability, and Example 8.1.7 illustrates that unexpected (or surprising) results can be obtained when combining the operations of integration and limits, and in particular, one cannot in general interchange the limit operation with integration.
Exercises
Suppose that is a sequence of functions such that is increasing for each . Suppose that exists for each . Is an increasing function?
Let be a sequence of positive numbers and define as
Find the pointwise limit of the sequence .
Find .
If , find
Recall that is countable and thus there exists a bijection . Define the sequence by letting . Now define as
Find the pointwise limit of the sequence .
Is Riemann integrable? Explain.
Is Riemann integrable? Explain.
Uniform Convergence
In the previous section we saw that pointwise convergence is a rather weak form of convergence since the limiting function will not in general inherit any of the properties possessed by the terms of the sequence. Examining the concept of pointwise convergence one observes that it is a very localized definition of convergence of a sequence of functions; all that is asked for is that converge for each . This allows the possibility that the speed of convergence of may differ wildly as varies in . For example, for the sequence of functions and , convergence of to zero is much faster for values of near than for values of near . What is worse, as convergence of to zero is arbitrarily slow. Specifically, recall in Example 8.1.5 that if and only if . Thus, for a fixed , as we have . Hence, there is no single that will work for all values of , that is, the convergence is not uniform.
Let be a sequence of functions on . We say that converges uniformly on to the function if for any there exists such that if then for all .
Notice that in Definition 8.2.1, the only depends on the given (but fixed) and the inequality holds for all provided . The inequality for all is equivalent to
for all and can therefore be interpreted as saying that the graph of lies in the tube of radius and centered along the graph of , see Figure 8.4.
-tubular neighborhood along the graph of ; if for all then the graph of is within the -tubular neighborhood of
The following result is a direct consequence of the definitions but it is worth stating anyhow.
If converges uniformly to then converges pointwise to .
Let and let be the sequence of functions on defined by . Prove that converges uniformly to .
We compute that
and thus converges pointwise to on . To prove that the convergence is uniform, consider
For any given if is such that then if then for any we have
This proves that converges uniformly to on . Note that a similar argument will not hold if we take .
Show that the sequence of functions converges uniformly to on .
We compute
and therefore if is such that then if then for all . Hence, converges uniformly to on .
On a close examination of the previous examples on uniform convergence, one observes that in proving that converges uniformly to on , we used an inequality of the form:
for some sequence of non-negative numbers such that . It follows that
This observation is worth formalizing.
Let be a sequence of functions. Then converges uniformly to on if and only if there exists a sequence of non-negative numbers converging to zero such that for sufficiently large.
Suppose that converges uniformly on to . There exists such that for all and . Hence, is well-defined for all . Define arbitrarily for . Given an arbitrary , there exists such that if then for all . We can assume that . Therefore, if then . This prove that .
Conversely, suppose that there exists such that and for all . Let be arbitrary. Then there exists such that if then . Hence, if then . This implies that if then for all , and thus converges uniformly to on .
Let be a continuous function on . Prove that there exists a sequence of step functions on that converges uniformly to on .
We end this section by stating and proving a Cauchy criterion for uniform convergence.
The sequence converges uniformly on if and only if for every there exists such that if then for all .
Suppose that uniformly on and let . There exists such that if then for all . Therefore, if then for all we have
To prove the converse, suppose that for every there exists such that if then for all . Therefore, for each the sequence is a Cauchy sequence and therefore converges. Let be defined by . If let be such that for all and . Fix and consider the sequence and thus . Now since then exists and , that is,
Therefore, if then for all .
Exercises
Let be a sequence of functions converging uniformly to . Let be a function and let for each . Under what condition on does the sequence converge uniformly? Prove it. What is the uniform limit of ?
Prove that if converges uniformly to on and converges uniformly to on then converges uniformly to on .
Let be the sequence defined in Exercise 8.1.2. Show that if then converges uniformly.
Let for . Prove that converges uniformly on .
Properties of Uniform Convergence
A sequence on is said to be uniformly bounded on if there exists a constant such that for all and for all .
Suppose that uniformly on . If each is bounded on then the sequence is uniformly bounded on and is bounded on .
By definition, there exists such that
for all and all . Since is bounded, then for all and thus is bounded on with upper bound . Therefore, for all and all . Let be an upper bounded for on for each . Then if then for all and all .
Give an example of a set and a sequence of functions on such that is bounded for each , converges pointwise to but is not uniformly bounded on .
Unlike the case with pointwise convergence, a sequence of continuous functions converging uniformly does so to a continuous function.
Let be a sequence of functions on converging uniformly to on . If each is continuous on then is continuous on .
To prove that is continuous on we must show that is continuous at each . Let be arbitrary. Recall that to prove that is continuous at we must show there exists such that if then . Consider the following:
Since uniformly on , there exists such that for all . Moreover, since is continuous there exists such that if then . Therefore, if then
This proves that is continuous at .
A direct consequence of Theorem 8.3.3 is that if pointwise and each is continuous then if is discontinuous then the convergence cannot be uniform.
Let for and . Each function is clearly continuous. Now and thus . If then
Therefore, converges pointwise to the function
The function is discontinuous and therefore, by Theorem 8.3.3, does not converge uniformly to .
The next property that we can deduce from uniform convergence is that the limit and integration operations can be interchanged. Recall from Example 8.1.7 that if pointwise then it is not necessarily true that
Since , then it in general we can say that
However, when the convergence is uniform we can indeed interchange the limit and integration operations.
Let be a sequence of Riemann integrable functions on . If converges uniformly to on then and
Let be arbitrary. By uniform convergence, there exists such that if then for all we have
or
By assumption, is Riemann integrable and thus if then
By the Squeeze Theorem of Riemann integration (Theorem 7.2.3), is Riemann integrable. Moreover, if then
implies (by monotonicity of integration)
and thus
This proves that the sequence converges to .
The following corollary to Theorem 8.3.5 is worth noting.
Let be a sequence of continuous functions on the interval . If converges uniformly to then and
If each is continuous then and Theorem 8.3.5 applies.
Consider the sequence of functions defined on given by
Draw a typical function .
Prove that converges pointwise.
Use Theorem 8.3.5 to show that the convergence is not uniform.
We now consider how the operation of differentiation behaves under uniform convergence. One would hope, based on the results of Theorem 8.3.3, that if uniformly and each is differentiable then is also differentiable and maybe even that at least pointwise and maybe even uniformly. Unfortunately, the property of differentiability is not generally inherited under uniform convergence. An example of this occurred in Example 8.1.6 where and where for . The convergence in this case is uniform on but although each is differentiable the limit function is not. It turns out that the main assumption needed for all to be well is that the sequence converge uniformly.
Let be a sequence of differentiable functions on . Assume that is Riemann integrable on for each and suppose that converges uniformly to on . Suppose there exists such that converges. Then the sequence converges uniformly on to a differentiable function and .
Let be arbitrary but with . By the Mean Value theorem applied to the differentiable function , there exists in between and such that
or equivalently
Therefore,
Since converges and is uniformly convergent, by the Cauchy criterion, for any there exists such that if then and for all . Therefore, if then
and this holds for all . By the Cauchy criterion for uniform convergence, converges uniformly. Let be the uniform limit of . We now prove that is differentiable and . By the Fundamental theorem of Calculus (FTC), we have that
for each . Since converges to and converges uniformly to we have
Thus and by the FTC we obtain .
Notice that in the statement of Theorem 8.3.8, all that is required is that converge for one . The assumption that converges uniformly then guarantees that in fact converges uniformly.
Consider the sequence defined on by . We compute that and clearly is continuous on for each . Now for all and thus converges pointwise to . To prove that the convergence is uniform we note that
Therefore, converges uniformly to on . Now and thus converges to . By Theorem 8.3.8, converges uniformly to say with and . Now by the FTC, and since then .
Exercises
Give an example of a set and a sequence of functions on such that is bounded for each , converges pointwise to but is not uniformly bounded on .
Suppose that uniformly on and uniformly on . Prove that if and are uniformly bounded on then converges uniformly to on . Then give an example to show that if one of or is not uniformly bounded then the result is false.
Let
for and .
Show that converges pointwise on .
Show that does not converge uniformly on any closed interval containing .
Show that converges uniformly on any closed interval not containing . For instance, take with .
Suppose that has that property that for all and some . Prove that if converges uniformly on to then the sequence converges uniformly to on . Note: and are compositions of functions and not function multiplication.
Let for and where .
Prove directly that the sequence is uniformly Cauchy.
If is the uniform limit of , find without computing .
Consider the sequence of functions on defined as follows:
Prove that converges uniformly to on .
For each fixed , find the improper integral
and show that
The results above seem to contradict Theorem 8.3.5. Explain why there is no contradiction.
Infinite Series of Functions
In this section, we consider series whose terms are functions. You have already encountered such objects when studying power series in Calculus. An example of an infinite series of functions (more specifically a power series) is
In this case, if we set then the above infinite series is . Let us give the general definition.
Let be a non-empty subset of . An infinite series of functions on is a series of the form for each where is a sequence of functions on . The sequence of partial sums generated by the series is the sequence of functions on defined as for each .
Recall that a series of numbers converges if the sequence of partial sums , defined as , converges. Hence, convergence of an infinite series of functions is treated by considering the convergence of the sequence of partial sums (which are functions). For example, to say that the series converges uniformly to a function we mean that the sequence of partial sums converges uniformly to , etc. It is now clear that our previous work in Sections 8.1-8.3 translate essentially directly to infinite series of functions. As an example:
Let be a sequence of functions on and suppose that converges uniformly to . If each is continuous on then is continuous on .
By assumption, the sequence of functions for converges uniformly to . Since each function is continuous, and the sum of continuous functions is continuous, it follows that is continuous. The result now follows by Theorem 8.3.3.
The following translate of Theorem 8.3.5 is worth explicitly writing out.
Let be a sequence of functions on and suppose that converges uniformly to . If each is Riemann integrable on then and
By assumption, the sequence defined as converges uniformly to . Since each is Riemann integrable then is Riemann integrable and therefore is Riemann integrable by Theorem 8.3.5. Also by Theorem 8.3.5, we have
or written another way is
or
or
We now state the derivative theorem (similar to Theorem 8.3.8) for infinite series of functions.
Let be a sequence of differentiable functions on and suppose that converges at some point . Assume further that converges uniformly on and each is continuous. Then converges uniformly to some differentiable function on and .
We now state a useful theorem for uniform convergence of infinite series of functions.
Let be a sequence of functions on and suppose that there exists a sequence of non-negative numbers such that for all , and all . If converges then converges uniformly on .
Let be arbitrary. Let be the sequence of partial sums of the series . By assumption, converges and thus is a Cauchy sequence. Hence, there exists such that for all . Let be the sequence of partial sums of . Then if then for all we have
Hence, the sequence satisfies the Cauchy Criterion for uniform convergence (Theorem 8.2.7) and the proof is complete.
Prove that
For any it holds that
A straightforward application of the Ratio test shows that is a convergent series. Hence, by the -Test, the given series converges uniformly on , and in particular on . By Theorem 8.4.3,
Consider the function whose graph is given in Figure 8.5; one can write down an explicit expression for but the details are unimportant. Consider the series
Since
and converges, then by the -test the above series converges uniformly on any interval . Let be the function defined by the series on . Now, on , the function has only a finite number of discontinuities and thus is Riemann integrable. Therefore, by Theorem 8.3.5, the function is Riemann integrable. The graph of is shown in Figure 8.6. One can show that has discontinuities at the rational points where .
The function
The function
Recall that a power series is a series of the form
where and . Hence, in this case if we write the series as then for each and . In calculus courses, the main problem you were asked to solve is to find the interval of convergence of the given power series. The main tool is to apply the Ratio test (Theorem 3.7.23):
Suppose that exists and is non-zero and (a similar argument can be done when the limit is zero). Then by the Ratio test, the power series converges if , that is, if . The number is called the radius of convergence and the interval is the interval of convergence (if is zero then can be chosen arbitrarily and the argument that follows is applicable). Let and consider the closed interval . Then if then
Now if then by assumption the series converges, and in particular the sequence is bounded, say by . Therefore,
Since , the geometric series converges. Therefore, by the -test, the series converges uniformly on the interval . Let for . Now consider the series of the derivatives
Applying the Ratio test again we conclude that the series of the derivatives converges for each and a similar argument as before shows that the series of derivatives converges uniformly on any interval where . It follows from the Term-by-Term Differentiation theorem that is differentiable and . By the Term-by-Term Integration theorem, we can also integrate the series and
where is any closed and bounded interval.
Consider the power series
Prove that the series converges at every .
Let denote the function defined by the series on the left and let denote the function defined by the series on the right. Justifying each step, show that exists and that .
Similarly, show that exists and .
A Fourier series is a series of the form
where .
Suppose that for a given and , the associated Fourier series converges pointwise on and let be the pointwise limit. Prove that in fact the Fourier series converges on . Hint: For any there exists such that .
Prove that if and are convergent series then the associated Fourier series converges uniformly on .
Suppose that for a given and , the associated Fourier series converges uniformly on and let be the uniform limit. Prove the following:
You will need the following identities:
By an open cover of an interval , we mean a collection of open intervals such that for each and
Here is some set, possibly uncountable. Prove that if is any open cover of then there there exists finitely many such that
Let be continuous and suppose that for all and all . Suppose that converges pointwise to a continuous function . Prove that the convergence is actually uniform. Give an example to show that if is not continuous then we only have pointwise convergence.
We first prove (a). For convenience, write for each and assume without loss of generality that is bounded above. Let and let be such that and
If then we are done because then . By induction, having defined , let be such that and . We claim that for some and thus . To prove the claim, suppose that for all . Then the increasing sequence converges by the Monotone Convergence theorem, say to . Since then for some and thus by convergence there exists such that . However, by definition of we must have that which is a contradiction to the definition of . This completes the proof.
Now we prove (b). First of all since for all it holds that for all and all . Fix and let . By pointwise convergence, there exists such that for all . By continuity of and , there exists such that and for all . Therefore, if then
for all . Hence, converges uniformly to on the interval . It is clear that is an open cover of . Therefore, by part (a), there exists such that . Hence, for arbitrary there exists such that if then for all . If then for all and all . This completes the proof. The sequence on satisfies for all and all , and converges to if and . Since is not continuous on , the convergence is not uniform.
Let be continuous and suppose that for all . Prove that if converges pointwise to a continuous function on then in fact the convergence is uniform.
Exercises
Show that converges uniformly on for every such that . Then show that the given series does not converge uniformly on . Hint: This is an important series and you should know what function the series converges uniformly to.
If prove that converges uniformly on .
Prove, justifying each step, that
For any number let be the function defined as if and if . Let be an enumeration of the rational numbers. Define as
Find the pointwise limit of the sequence . Is the convergence uniform? Explain.