Let (X,dX) be a metric space and (Y,∥⋅∥Y) be a normed vector space. Let fn∈B(X,Y) be a sequence of functions. We say
∑n=1∞fnconverges pointwise if for all x∈X, ∑n=1Nfn(x) converges as N→∞.
∑n=1∞fnconverges uniformly if ∑n=1Nfn converges uniformly as N→∞.
∑n=1∞fnconverges pointwise absolutely if for all x∈X, ∑n=1N∥fn(x)∥Y converges as N→∞.
∑n=1∞fnconverges uniformly absolutely if ∑n=1Nsupx∈X∥fn(x)∥Y converges as N→∞.
Remark.
The main reason we want (Y,∥⋅∥Y) to be a normed vector space instead of just a metric space is because we need to be able to add functions together, and in general metric spaces, we don't have a notion of addition.
Also, in most cases, (Y,∥⋅∥Y)=(R,∣⋅∣).
Theorem (Weierstrass M-test)
Let (X,dX), (Y,∥⋅∥Y), and fn∈B(X,Y) be as above. If Y is complete (with respect to the metric induced by the norm), then (iv) ⟹ (ii).
Since R is complete, the Weierstrass M-test will be our main way of proving uniform convergence of series for most problems.
Example 1.
Let fn(x)=xn, where X=(−1,1) and Y=R. Show that ∑n=1∞xn converges uniformly on any compact subset of (−1,1), but not uniformly on all of (−1,1).
Solution.
Without loss of generality, we may assume that our compact set is [−r,r]. Indeed, given any compact set K⊆(−1,1), we have
K⊆r∈(0,1)⋃(−r,r)
so by compactness, there exist 0<r1<r2<⋯<rn<1 such that
K⊆i=1⋃n(−ri,ri)=(−rn,rn).
Since K is closed, K=K⊆(−rn,rn)=[−rn,rn]. Thus, if the series converges uniformly on [−rn,rn], then it converges uniformly on K also.
since it is a geometric series with ratio ∣r∣<1. By the Weierstrass M-test, we get ∑n=1∞xn converges uniformly on [−r,r], hence on any compact subset of (−1,1).
To show that the convergence is not uniform on all of (−1,1), we proceed by contradiction. Assume that ∑n=1∞xn converges uniformly on (−1,1). Since the pointwise and uniform limits (when they exist) must agree, this means
n=1∑NxnN→∞1−x1
uniformly on (−1,1). But ∑n=1Nxn∈B(X,R) and B(X,R) is complete, i.e., uniform limits of bounded functions are bounded, which implies 1−x1 is bounded on (−1,1), a contradiction. Thus, the series does not converge uniformly on all of (−1,1).
Example 2.
Prove that ∑n=1∞1+n5x2nx converges uniformly on R.
Solution.
Our main tool is the Weierstrass M-test, so let's try to calculate
x∈Rsup∣fn(x)∣wherefn(x)=1+n5x2nx.
We'll try to maximize it using the first derivative test. Since fn is an odd function, to maximize ∣fn∣ on all of R, we can just maximize fn for x≥0:
fn′(x)=(1+n5x2)2n−n6x2=0⟹x=n−25
Also,
fn(n−25)=2n−231>0.
Now we need to prove that this actually gives us the global maximum of fn. Notice that
x→∞limfn(x)=x→∞limx21+n5xn=n50=0.
(We're allowed to use limit laws since the limit of the denominator is non-zero.)
Thus, because fn(n−25)>0, there exists M>0 such that if x≥M, then fn(x)<fn(n−25). The next part of the argument is a little tricky, so let me break it down into steps:
If x≥M, then fn(x) is strictly smaller than fn(n−23), so if fn has a maximum, it must occur on the interval [0,M].
[0,M] is compact and fn is continuous, so fn must have a maximum on [0,M]. In other words, fn has a global maximum.
Since fn is differentiable on [0,M], the maximizer xmax of fn must be a critical point, i.e., fn′(xmax)=0.
The only critical point of fn is n−25, so xmax=n−25.
In summary,
x∈Rsup∣fn(x)∣=fn(n−25)=2n231,
so by the p-series test with p=23>1,
n=1∑∞x∈Rsup∣fn(x)∣=n=1∑∞2n231<∞.
So by the Weierstrass M-test, the series converges uniformly on R.
Example 3.
Show that ∑n=1∞n4+x2n2+x4 converges to a continuous function on R.
Solution.
Warning: It might be tempting to try to prove that the series converges uniformly on R (since that would imply that the series converges to a continuous function), but this series does not converge uniformly on R. Indeed, let f(x)=∑n=1∞n4+x2n2+x4, and because all terms are positive,
(f is larger than its partial sums, so we can drop the absolute value, and a sum of positive numbers is at least one of its terms, e.g., the first one.)
Instead, our strategy is to prove that the series converges uniformly on compact intervals. If we manage to do this, then given any x∈R, there exists R>0 such that x∈(−R,R), so
n=1∑∞n4+x2n2+x4 converges uniformly on [−R,R]⟹f is continuous on [−R,R]⟹f is continuous at x.
This works for any x∈R, so we would show that f is continuous on R. So, we need to show that the sum converges uniformly on [−R,R] for any R>0. As before, our main tool is the Weierstrass M-test, so if x∈[−R,R], then