Tutorials › Calculus II › Undoing Differentiation and Integration

Power Series Operations · Tutorial 768 of 998

Undoing Differentiation and Integration

Follow the coefficients to see exactly when differentiation and integration undo one another—and which constants or polynomial terms require special attention.

Intermediate 10 min read

What You'll Learn

  • Write the coefficient rule for differentiating a power series term by term.
  • Identify the arbitrary constant introduced by termwise integration.
  • Determine why integrating a derivative recovers a series only up to its constant term.
  • Check both orders of differentiation and integration by following coefficients.
  • Track the low-degree polynomial terms affected by repeated operations.

When Do Differentiation and Integration Undo Each Other?

A derivative removes constant terms, while an indefinite integral introduces a choice of constant. For power series, these familiar facts have a precise coefficient-level description: differentiation shifts and rescales coefficients, and integration shifts them back while leaving one coefficient to be chosen.

Keeping track of that coefficient matters whenever you use one operation to reverse the other. Differentiating an antiderivative recovers the original series exactly. Integrating a derivative recovers the original series only after its constant term is restored. Repeating the operations makes the same issue visible for a whole polynomial of low-degree terms.

Coefficient Rules for One Operation

Let a power series centered at \(a\) be

$$ F(x)=\sum_{n=0}^{\infty}c_n(x-a)^n, $$

and suppose its radius of convergence is \(R>0\). The termwise differentiation and integration theorems apply for \(|x-a|<R\). Within that interval, differentiation gives

$$ F'(x)=\sum_{n=1}^{\infty}n c_n(x-a)^{n-1} =\sum_{n=0}^{\infty}(n+1)c_{n+1}(x-a)^n. $$

The second expression reindexes the differentiated series so that it is again written in powers \((x-a)^n\). In particular, the coefficient of \((x-a)^n\) after differentiation is \((n+1)c_{n+1}\), and the original constant coefficient \(c_0\) does not appear.

An antiderivative is obtained by integrating each term and adding an arbitrary constant \(C\):

$$ G(x)=C+\sum_{n=0}^{\infty}\frac{c_n}{n+1}(x-a)^{n+1}. $$

Here the constant term of \(G\) is \(C\), while the coefficient of \((x-a)^{n+1}\) is \(c_n/(n+1)\). This integrated series has the same radius of convergence \(R\); the formulas are being used on its interior, \(|x-a|<R\).

Key coefficient rules: Differentiation sends \(c_n\) to the coefficient \((n+1)c_{n+1}\) in degree \(n\). Integration sends \(c_n\) to \(c_n/(n+1)\) in degree \(n+1\), and adds a freely chosen degree-zero coefficient \(C\).

Why the Two Orders Give Different Results

Let \(D\) denote termwise differentiation and let \(I_C\) denote termwise integration with the constant chosen to be \(C\). Applying differentiation to an antiderivative cancels the factor introduced by integration. Applying integration to a derivative restores every positive-degree coefficient, but it cannot recover the original constant coefficient unless that value is selected.

$$ D(I_C F)=F, \qquad I_C(D F)=F-c_0+C. $$

These identities hold for \(|x-a|<R\). In the second identity, \(c_0=F(a)\), since all positive powers of \(x-a\) vanish at \(x=a\). Thus choosing \(C=c_0\) makes the second composition recover \(F\) exactly; choosing \(C=0\) produces \(F-c_0\).

Theorem: For a power series of positive radius of convergence, termwise differentiation followed by termwise integration returns the original series up to the chosen integration constant. Termwise integration followed by differentiation returns the original series exactly on the interval of convergence.

Why These Identities Hold

The identities follow by comparing coefficients, and the verification also explains why the constant term is the only term lost in a single differentiation.

1
Differentiate the series.
The coefficient of \((x-a)^n\) in \(D F\) is \((n+1)c_{n+1}\). The coefficient \(c_0\) contributes no term because the derivative of a constant is zero.
2
Integrate the differentiated series.
Integrating \((n+1)c_{n+1}(x-a)^n\) gives \(c_{n+1}(x-a)^{n+1}\), since \((n+1)/(n+1)=1\). Therefore all original coefficients \(c_1,c_2,\ldots\) return in their original degrees. The integration constant is \(C\), so the result is \(C+\sum_{n=1}^{\infty}c_n(x-a)^n=F-c_0+C\).
3
Reverse the order.
Integrating \(F\) gives \(C+\sum_{n=0}^{\infty}c_n(x-a)^{n+1}/(n+1)\). Differentiating a term in this sum multiplies its coefficient by \(n+1\) and reduces its degree by one. Each term therefore becomes \(c_n(x-a)^n\), while the constant \(C\) differentiates to zero. The result is exactly \(F\).

The termwise operations are valid inside the original radius of convergence, where the power-series differentiation and integration theorems apply. No endpoint conclusion is needed for these identities.

Worked Example: A Polynomial in Both Orders

Worked Example: A Polynomial in Both Orders

Let \(t=x-2\), and consider \(F(x)=4-3t+5t^2-2t^3\). Differentiate term by term:

$$ F'(x)=0-3+10t-6t^2=-3+10t-6t^2. $$

Integrate this derivative, keeping the arbitrary constant visible:

$$ \int F'(x)\,dx =C-3t+\frac{10}{2}t^2-\frac{6}{3}t^3 =C-3t+5t^2-2t^3. $$

To recover \(F\), choose \(C=4\), the original constant coefficient. The result is \(4-3t+5t^2-2t^3=F(x)\). If instead \(C=0\), the result is \(F(x)-4\).

Check the reverse order by integrating \(F\):

$$ \int F(x)\,dx =C+4t-\frac{3}{2}t^2+\frac{5}{3}t^3-\frac{1}{2}t^4. $$

Differentiate to check:

$$ \frac{d}{dx}\left(C+4t-\frac{3}{2}t^2+\frac{5}{3}t^3-\frac{1}{2}t^4\right) =4-3t+5t^2-2t^3 =F(x), $$

because \(dt/dx=1\). The arbitrary constant disappears, and every term of \(F\) returns.

Worked Example: An Infinite Series

Worked Example: An Infinite Series

Consider the power series \(F(x)=\sum_{n=0}^{\infty}x^n/n!\), which converges for every real \(x\). Differentiate term by term and reindex using \(m=n-1\):

$$ F'(x)=\sum_{n=1}^{\infty}\frac{n x^{n-1}}{n!} =\sum_{n=1}^{\infty}\frac{x^{n-1}}{(n-1)!} =\sum_{m=0}^{\infty}\frac{x^m}{m!} =F(x). $$

Now integrate this derivative term by term, with integration constant \(C\):

$$ \int F'(x)\,dx =C+\sum_{n=0}^{\infty}\frac{x^{n+1}}{(n+1)n!} =C+\sum_{n=0}^{\infty}\frac{x^{n+1}}{(n+1)!}. $$

Reindex the final sum with \(m=n+1\). It contains all terms of \(F\) except the degree-zero term:

$$ \int F'(x)\,dx =C+\sum_{m=1}^{\infty}\frac{x^m}{m!} =C+F(x)-1. $$

Thus \(C=1\) recovers \(F\), whereas \(C=0\) gives \(F-1\). To check the opposite order, integrate \(F\) first:

$$ \int F(x)\,dx =C+\sum_{n=0}^{\infty}\frac{x^{n+1}}{(n+1)!}. $$

Its derivative is \(\sum_{n=0}^{\infty}x^n/n!=F(x)\), as required. The coefficient calculation checks both compositions, including the constant term.

Repeated Operations and Lost Low-Degree Terms

The same principle extends to repeated operations. Differentiating \(k\) times removes all terms of degree less than \(k\), because each such term becomes zero before the \(k\)th derivative. Integrating \(k\) times introduces an arbitrary polynomial of degree at most \(k-1\): each integration contributes a new constant, and later integrations can turn those constants into higher-degree terms.

For \(F(x)=\sum_{n=0}^{\infty}c_n(x-a)^n\), the \(k\)th derivative, for a positive integer \(k\), is

$$ F^{(k)}(x)=\sum_{n=k}^{\infty}\frac{n!}{(n-k)!}c_n(x-a)^{n-k}. $$

If this series is integrated \(k\) times using zero constants at each stage, the result is \(F(x)-\sum_{j=0}^{k-1}c_j(x-a)^j\). The terms of degree \(k\) and above return, but the first \(k\) coefficients do not. Choosing the integration constants can restore those missing terms.

Worked Example: Two Derivatives and Two Integrations

Let \(t=x-1\) and \(F(x)=2+3t-4t^2+5t^3+t^4\). Differentiate twice:

$$ F'(x)=3-8t+15t^2+4t^3, \qquad F''(x)=-8+30t+12t^2. $$

Integrate \(F''\) once, choosing zero constant:

$$ \int F''(x)\,dx=-8t+15t^2+4t^3. $$

Integrate again, again choosing zero constant:

$$ \int\left(-8t+15t^2+4t^3\right)\,dx =-4t^2+5t^3+t^4. $$

This is \(F(x)-2-3t\): the constant and linear terms were lost. Add the polynomial \(2+3t\) to recover \(F\). Check by differentiating the recovered expression twice:

$$ \frac{d}{dx}(2+3t-4t^2+5t^3+t^4) =3-8t+15t^2+4t^3, $$ $$ \frac{d^2}{dx^2}(2+3t-4t^2+5t^3+t^4) =-8+30t+12t^2 =F''(x). $$

The check confirms that two derivatives retain the terms of degree two and higher, while the two integration constants allow the missing degree-zero and degree-one terms to be chosen.

Common Pitfall: Assuming Integration Automatically Restores the Constant

A frequent error is to integrate a derivative and claim that the original function has been recovered without accounting for its constant term. For the polynomial in the first example, setting \(t=x-2\) gives

$$ \frac{d}{dx}(4-3t+5t^2-2t^3)=-3+10t-6t^2. $$

An incorrect calculation then says

$$ \int(-3+10t-6t^2)\,dx=-3t+5t^2-2t^3 =4-3t+5t^2-2t^3. $$

The final equality is false: the expression on the left has constant coefficient \(0\) in powers of \(t\), while the expression on the right has constant coefficient \(4\). The correction is to retain the arbitrary constant:

$$ \int(-3+10t-6t^2)\,dx=C-3t+5t^2-2t^3. $$

Only when \(C=4\) does this equal the original polynomial. The general rule is not that integration reverses differentiation without qualification; it is that integration recovers the differentiated series up to an additive constant.

Why the Coefficient View Is Useful

These rules make it possible to check power-series calculations without relying only on a remembered function formula. Track the degree and coefficient of each term: differentiation lowers the degree by one and multiplies by the former degree; integration raises the degree by one and divides by the new degree. Then account for the constant that differentiation removes or integration leaves undetermined.

For repeated operations, identify how many low-degree coefficients have been removed. After \(k\) derivatives, terms of degrees \(0\) through \(k-1\) are absent. Integrating \(k\) times can restore them only if the resulting polynomial of degree at most \(k-1\) is selected. This distinction is the coefficient-level counterpart of the fact that a derivative cannot distinguish between functions that differ by a constant, or, after repeated differentiation, by a polynomial of the corresponding lower degree.

Check Your Understanding

Use coefficient tracking to answer each question. In questions involving an antiderivative, state what choice of constant or polynomial is needed for exact recovery.

  1. If \(F(x)=\sum_{n=0}^{\infty}c_n(x-a)^n\), what is the coefficient of \((x-a)^n\) in \(F'(x)\)?
  2. What does integrating \(F'\) with constant \(C=0\) give in terms of \(F\) and \(c_0\)?
  3. Why does differentiating an antiderivative of \(F\) recover \(F\) for any choice of integration constant?
  4. After taking two derivatives of a power series, which original degrees have disappeared?
  5. For \(F(x)=7+2(x-a)+(x-a)^2\), what polynomial must be added after integrating \(F''\) twice with zero constants to recover \(F\)?

Want all 1,114 practice problems?

This tutorial covers one topic. Calc Pass covers the full Calculus AB & BC curriculum (the core of Calculus I and II): full review, every practice question, and step-by-step solutions to each one.

QR code to download the Calc Pass app

Get the Calc Pass App

Scan to download and practice on the go, with instant scoring.

Calc Pass Volume 1: 1,114 Practice Questions book cover

Vol. 1: 1,114 Questions

The complete practice question bank, organized by AB & BC topic.

View on Amazon
Calc Pass Volume 2: Solutions Manual book cover

Vol. 2: Solutions Manual

Fully worked solutions to every question in Volume 1.

View on Amazon