Exercises — Term-by-term differentiation and integration of power series
Before we start, one reminder of the vocabulary we will use constantly.

Level 1 — Recognition
L1.1
State the term-by-term derivative of
Recall Solution
WHAT we do: apply the derivative rule — multiply the -th coefficient by , drop the power by one, and start the sum at (the term is a constant, its derivative is ). Here , so : WHY is unchanged: the original series has (it looks like the geometric series slowed by ), and differentiation never changes . So the answer is valid on .
L1.2
State the term-by-term antiderivative of taking the constant so that the antiderivative is at .
Recall Solution
WHAT we do: integrate each brick. Rule: divide the coefficient of by and raise the power to . Here the power is , so : WHAT it is: this is exactly the series (see Example 2 in the parent). The constant is because the definite integral from is at .
Level 2 — Application
L2.1
Starting from the geometric series (), find a power series for
Recall Solution
WHY differentiate? Notice (chain rule on the left). So differentiating the known series hands us the unknown one for free. (Re-index .) So , valid on . Check the first two coefficients: ✓, and the coefficient of is .
L2.2
Find a power series for centred at , and state its radius.
Recall Solution
WHY integrate? Because . So is an antiderivative of . Integrate term by term from to : Constant: , so . Radius: integration keeps .
Level 3 — Analysis
L3.1
The series is , valid on with radius . Examine both endpoints and : does the series converge there? Did convergence get gained or lost compared with the series we integrated?
Recall Solution
At : the series becomes . This is an alternating series whose terms decrease to , so by the alternating-series test it converges (to ). At : plug in — odd powers flip sign, giving , which also converges (to ). Gained or lost? The original series at is , which diverges (terms don't go to ). After integrating, both endpoints now converge — so integration gained convergence at the endpoints. This matches the parent note's rule: is unchanged, but endpoints can improve when you integrate. See Abel's theorem (endpoint behaviour) for why the sum there really equals .
L3.2
Take the series (radius , converges at , diverges at ). Differentiate term by term and check what happens to the endpoint at .
Recall Solution
Differentiate: , radius still . Endpoint : the derivative series is , which diverges (terms don't tend to ). Conclusion: the original converged at , the derivative does not. So differentiation lost convergence at that endpoint — the mirror image of L3.1. This is exactly the "DIRT: Test endpoints again" warning.
Level 4 — Synthesis
L4.1
Find a closed form for and then evaluate .
Recall Solution
WHY start from L2.1? We already know . Our target has one extra factor of , so just multiply both sides by : Numeric evaluation: put (which is inside ):
L4.2
Find a closed form for
Recall Solution
WHY differentiate? The coefficient signals an integral was done — so differentiating should simplify it. Differentiate term by term: Recover by integrating back. Use the partial-fraction split : Constant: and ✓. This is the artanh (inverse hyperbolic tangent) series. Spot check at : .
Level 5 — Mastery
L5.1
Solve the initial value problem using a power series . Identify the resulting function and state .
Recall Solution
WHY a power series? Because term-by-term differentiation turns a differential equation into plain algebra on the coefficients (this is the method of Solving ODEs with power series). Step 1 — differentiate the ansatz. (re-index ). Step 2 — match to . The equation means, coefficient by coefficient of : Step 3 — use the initial condition. . Then , , , and in general Step 4 — read off the function. , with radius (the ratio , so it converges for every ).
L5.2
Using only the known series and one integration, prove the Leibniz series and state clearly which theorem lets you push all the way to the endpoint .
Recall Solution
Step 1 — integrate. From (), integrate from to : Step 2 — endpoint convergence. At the series is , alternating with terms decreasing to , so it converges (L3.1 showed this). Step 3 — why the sum equals . Convergence at the endpoint is not enough on its own; we invoke Abel's theorem (endpoint behaviour): if a power series converges at an endpoint, its sum there equals the limit of the function from inside. Since is continuous and , That is the Leibniz formula.
Recall Fast self-check (fold to test yourself)
Term-by-term derivative of ::: again (it is ). Closed form of ::: . Value of ::: . Closed form of ::: . Theorem that pushes series to ::: Abel's theorem.
Connections
- Parent: 4.3.15
- Power series and radius of convergence
- Cauchy–Hadamard theorem
- Uniform convergence and the Weierstrass M-test
- Taylor and Maclaurin series
- Geometric series
- Abel's theorem (endpoint behaviour)
- Solving ODEs with power series