A Taylor polynomial stops at some degree n. If you never stop, you get a Taylor series: a power series that, for many functions, equals the function exactly on its interval of convergence. Four of these series come up so often that the AP exam expects you to know them by heart. Once you know them, you can build series for many other functions without taking a single derivative.
When a=0, it’s called a Maclaurin series. Its partial sums are exactly the Taylor polynomials P0,P1,P2,…
For the functions in this course, the Taylor series equals f(x) at every x in its interval of convergence. (For ex, sinx, and cosx you can prove this with the Lagrange error bound, which shrinks to 0 as n grows.)
Inside the interval of convergence, adding more terms makes the partial sums match the function over a wider and wider stretch. For sinx, cosx and ex this keeps going for every x.
The partial sums of the cosine series of degree 4 and 10. The series converges to cosx for every x, so higher degrees fit over wider intervals.
Mixing up the sine and cosine series. Sine has odd powers and starts with x (because sin0=0); cosine has even powers and starts with 1 (because cos0=1). Check by plugging in x=0.
Dividing by the wrong factorial. In the sine series, x5 is divided by 5!, not 3! or (2n)!. The factorial always matches the power: xk goes with k!.
Using the 1/(1 − x) series outside its interval. The series 1+x+x2+⋯ only equals 1−x1 for −1<x<1. At x=2 the function equals −1, but the series diverges.
Writing a general term that doesn’t match the first terms. Always test your general term with n=0,1,2 to see that it reproduces the terms you wrote out, as in Example 2.
Forgetting radians. The series for sinx and cosx are only true in radians. sin30 means 30 radians, not 30∘.
2. (Warm-up) Find the sum of each series by recognizing a known Maclaurin series.
(a) 1+2+2!22+3!23+⋯
(b) 1−2!π2+4!π4−6!π6+⋯
Solution
(a) This is the ex series with x=2, so the sum is e2.
(b) This is the cosx series with x=π, so the sum is cosπ=−1.
3. (Warm-up) What is the sum of 1+0.3+0.32+0.33+⋯? What happens to 1+2+22+23+⋯, even though 1−21=−1?
Solution
The first is the 1−x1 series with x=0.3, which is inside −1<x<1, so the sum is 0.71=710.
The second has x=2, outside the interval of convergence. The partial sums grow without bound, so the series diverges. The formula 1−x1 doesn’t apply there.
4. (Core) Find the Taylor series for ex about x=3. Give the first three terms and the general term.
Solution
Every derivative of ex is ex, so f(n)(3)=e3 for every n:
ex=e3+e3(x−3)+2!e3(x−3)2+⋯=n=0∑∞n!e3(x−3)n
5. (Core) Find the Taylor series for cosx about x=2π. Give the first three nonzero terms and the general term.
Solution
At x=2π, the derivatives cosx,−sinx,−cosx,sinx,cosx,−sinx,… have values 0,−1,0,1,0,−1,… So only odd powers appear:
6. (Core) A function f has f(n)(0)=(−1)n2n for all n≥0. Write the first four terms and the general term of its Maclaurin series. Which familiar function is f?
The general term is n!(−1)n2nxn=n!(−2x)n. That’s the ex series with x replaced by −2x, so f(x)=e−2x.
7. (Core) Find the Taylor series for f(x)=x1 about x=2, and its interval of convergence.
Solution
f(n)(x)=(−1)nn!x−(n+1), so f(n)(2)=2n+1(−1)nn!. Dividing by n!:
x1=21−4x−2+8(x−2)2−⋯=n=0∑∞2n+1(−1)n(x−2)n
This is geometric with ratio −2x−2, so it converges when 2∣x−2∣<1, that is, 0<x<4. At x=0 and x=4 the terms have constant size 21, so they don’t approach 0 and both endpoints diverge. The interval of convergence is 0<x<4.
8. (Challenge) Use the Maclaurin series for ex to find x→0limx2ex−1−x.