Taylor Remainder Inequality
If a is in an open interval I and
M≥∣f(k+1)(c)∣ for all c in the interval I, then the Taylor
remainder of degree k centered at x=a satisfies
∣Rkf(x)∣≤(k+1)!M∣x−a∣k+1
for all x in the interval I. The quantity on the right is called the
Taylor bound on the remainder.
3. Convergence Proofs
To prove the Taylor series for f(x) converges to f(x), it suffices
to prove the limit of the Taylor remainders is 0:
k→∞limRkf(x)=0
There are a couple of technical points which need to be discussed before we
start the proof. We need to distinguish between a function f(x) and the
sum of its Taylor series which we denote by Tf(x). We know that the
Taylor series converges to Tf(x) but we do not yet know that it
converges to f(x). Thus, the Taylor series is:
Tf(x)=n=0∑∞n!f(n)(a)(x−a)n
Further, if the Taylor series is approximated by the kth
Taylor polynomial
Tkf(x)=n=0∑kn!f(n)(a)(x−a)n
then the error in the approximation is
Ek=Tf(x)−Tkf(x)=n=k+1∑∞n!f(n)(a)(x−a)n
This needs to be distinguished from the Taylor remainder which is defined
to be:
Rkf(x)=f(x)−Tkf(x)
using f(x) instead of Tf(x). It is Rkf(x) which appears in
the Lemma as well as the Taylor Remainder Formula and the Taylor Remainder
Inequality.
To prove the Taylor series for f(x) converges to f(x), we need
to show the partial sums (the Taylor polynomials) converge to f(x):
k→∞limTkf(x)=f(x)
Subtracting each side from f(x), this is equivalent to:
k→∞lim[f(x)−Tkf(x)]=0
or
k→∞limRkf(x)=0
Prove that the Maclaurin series for f(x)=sinx converges to
sinx for all real numbers x.
The Taylor series is:
Tf(x)=n=0∑∞(2n+1)!(−1)nx2n+1
and the kth partial sum is the Taylor polynomial of degree
2k+1:
T2k+1f(x)=n=0∑k(2n+1)!(−1)nx2n+1
and the Taylor remainder is:
R2k+1f(x)=sinx−n=0∑k(2n+1)!(−1)nx2n+1
To show the remainders converge to 0, we use the Taylor bound on
the remainder. For f(x)=sinx, the (2k+2)nd derivative
is either sinx or −sinx. In either of
these cases, ∣f(2k+2)(x)∣≤1. So we take M=1.
Thus,
∣R2k+1f(x)∣≤(2k+2)!M∣x∣2k+2=(2k+2)!∣x∣2k+2
and so
k→∞lim∣R2k+1f(x)∣≤k→∞lim(2k+2)!∣x∣2k+2=0
We'll show n→∞limn!∣x∣n=0.
Pick an integer N>2∣x∣. Then ∣x∣<2N.
Now if p>N, then p1<N1 and p∣x∣<2NN=21.
We now compute (with n>N)
n→∞limn!∣x∣n=n→∞lim1∣x∣2∣x∣⋯N∣x∣N+1∣x∣⋯n∣x∣=N!∣x∣Nn→∞limN+1∣x∣⋯n∣x∣<N!∣x∣Nn→∞lim(21)n−N=0
To show k→∞limRkf(x)=0, we use
the Taylor bound on the remainder. Let N be a number
>∣x∣. Consider the interval (−N,N).
For f(x)=e2x, we have f(k+1)(x)=2k+1e2x.
So, ∣f(k+1)(x)∣≤2k+1e2N. So we take M=2k+1e2N.
Thus,
∣Rkf(x)∣≤(k+1)!M∣x∣k+1=(k+1)!e2N∣2x∣k+1
and so
k→∞lim∣Rkf(x)∣≤e2Nk→∞lim(k+1)!∣2x∣k+1=0
Placeholder text:
Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum
Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum
Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum
Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum
Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum