What Is A Power Series?
A power series is an infinite series built from rising powers of $(x - a)$, each with its own constant coefficient:
$$f(x) = \sum_{n=0}^{\infty} c_n (x-a)^n = c_0 + c_1(x-a) + c_2(x-a)^2 + c_3(x-a)^3 + \cdots$$
The number $a$ is the center, the constants $c_n$ are the coefficients, and $x$ is the variable. Think of it as a polynomial that never stops. A polynomial has a highest power and then ends; a power series keeps going, so it can match curves that no finite polynomial can.
When $a = 0$ the series takes its simplest shape, powers of $x$ alone:
$$\sum_{n=0}^{\infty} c_n x^n = c_0 + c_1 x + c_2 x^2 + c_3 x^3 + \cdots$$
The one question that governs everything is: for which values of $x$ does the endless sum settle on a finite number? A power series always converges at its center (plug in $x = a$ and every term after the first vanishes), but away from the center it may or may not converge. The set of $x$ where it does is the topic of the next section, and it connects directly to the wider study of sequences and series in calculus.
How Do You Find The Radius And Interval Of Convergence?
For any power series, exactly one of three things happens. It converges only at $x = a$; or it converges for every real $x$; or there is a positive number $R$, the radius of convergence, such that the series converges when $|x - a| < R$ and diverges when $|x - a| > R$.
$$\text{Converges for } |x - a| < R \qquad \text{Diverges for } |x - a| > R$$
The interval of convergence is the actual set of $x$-values that work: the open interval $(a - R, ; a + R)$, plus possibly one or both endpoints, which must be checked one at a time.
The tool for finding $R$ is the ratio test. Form the absolute ratio of consecutive terms and take its limit:
$$L = \lim_{n \to \infty} \left| \frac{c_{n+1}(x-a)^{n+1}}{c_n (x-a)^n} \right|$$
The series converges absolutely when $L < 1$. Solving $L < 1$ for $x$ produces an inequality of the form $|x - a| < R$, which hands you $R$ directly.
Two habits keep this reliable, and both are on the list of documented student errors below: keep the absolute-value bars until the very end, and remember that $R$ is the distance you compare against, not the raw limit you first wrote down.
What Is The Geometric Power Series?
The single most useful power series is the geometric one, powers of $x$ with every coefficient equal to $1$:
$$\sum_{n=0}^{\infty} x^n = 1 + x + x^2 + x^3 + \cdots = \frac{1}{1-x}, \qquad |x| < 1$$
This is worth proving, because the reasoning shows exactly why an interval appears. The partial sum of the first $N + 1$ terms has a closed form (the same one behind a geometric progression):
$$S_N = 1 + x + x^2 + \cdots + x^N = \frac{1 - x^{N+1}}{1 - x}$$
Now let $N \to \infty$. When $|x| < 1$, the power $x^{N+1}$ shrinks to $0$, so $S_N \to \frac{1}{1-x}$. When $|x| \ge 1$, the term $x^{N+1}$ does not shrink to $0$, the partial sums never settle, and the series diverges. That is the whole story of convergence in one line: the terms have to shrink fast enough for the running total to stop moving.
The radius here is $R = 1$. Checking the endpoints: at $x = 1$ the series is $1 + 1 + 1 + \cdots$, which diverges, and at $x = -1$ it is $1 - 1 + 1 - \cdots$, which also diverges. So the interval of convergence is the open interval $(-1, 1)$, both endpoints excluded.
How Do You Differentiate And Integrate A Power Series Term By Term?
Inside its interval of convergence, a power series behaves like a very long polynomial: you may differentiate or integrate it one term at a time, and the new series has the same radius $R$ (the endpoints can change, so they are re-tested). This is what makes power series so powerful for calculus.
Start from the geometric series and differentiate both sides. On the left, use the standard derivative:
$$\frac{d}{dx}\left[ \frac{1}{1-x} \right] = \frac{1}{(1-x)^2}$$
Differentiating the right side one term at a time:
$$\frac{d}{dx}\left[ \sum_{n=0}^{\infty} x^n \right] = \sum_{n=1}^{\infty} n x^{n-1}$$
Setting the two equal gives a brand-new series for free, valid on the same $|x| < 1$:
$$\frac{1}{(1-x)^2} = \sum_{n=1}^{\infty} n x^{n-1} = 1 + 2x + 3x^2 + 4x^3 + \cdots$$
Integration works the same way. Using the integral of the geometric series:
$$\int \frac{1}{1-x} , dx = -\ln(1 - x) + C \qquad \text{and} \qquad \int \sum_{n=0}^{\infty} x^n , dx = \sum_{n=0}^{\infty} \frac{x^{n+1}}{n+1} + C$$
Matching the two at $x = 0$ forces $C = 0$, which yields a power series for the natural logarithm:
$$-\ln(1 - x) = \sum_{n=1}^{\infty} \frac{x^n}{n} = x + \frac{x^2}{2} + \frac{x^3}{3} + \cdots, \qquad |x| < 1$$
Verification: differentiate the result back. $\frac{d}{dx}\left[ -\ln(1-x) \right] = \frac{1}{1-x}$, which is exactly the geometric series we started from, so the term-by-term integration is confirmed.
How Do You Represent A Function As A Power Series?
Once you own the geometric series, you can dress up other functions to match its shape $\frac{1}{1 - (\text{something})}$ and read off their power series. Replace $x$ with $-x^2$ in the geometric series:
$$\frac{1}{1 + x^2} = \frac{1}{1 - (-x^2)} = \sum_{n=0}^{\infty} (-x^2)^n = 1 - x^2 + x^4 - x^6 + \cdots, \qquad |x| < 1$$
Integrating that term by term gives the classic series for the inverse tangent, $\arctan x = x - \frac{x^3}{3} + \frac{x^5}{5} - \cdots$. This table collects the standard representations most worth memorising.
Table: Common functions written as a power series, with their intervals of convergence.
Function | Power series | Interval of convergence |
|---|---|---|
$\dfrac{1}{1-x}$ | $\displaystyle\sum_{n=0}^{\infty} x^n$ | $(-1, 1)$ |
$\dfrac{1}{1+x}$ | $\displaystyle\sum_{n=0}^{\infty} (-1)^n x^n$ | $(-1, 1)$ |
$\ln(1+x)$ | $\displaystyle\sum_{n=1}^{\infty} \frac{(-1)^{n-1} x^n}{n}$ | $(-1, 1]$ |
$\arctan x$ | $\displaystyle\sum_{n=0}^{\infty} \frac{(-1)^n x^{2n+1}}{2n+1}$ | $[-1, 1]$ |
$e^{x}$ | $\displaystyle\sum_{n=0}^{\infty} \frac{x^n}{n \text{ factorial}}$ | $(-\infty, \infty)$ |
The last row is the series a calculator leans on: $e^x = 1 + x + \frac{x^2}{2} + \frac{x^3}{6} + \frac{x^4}{24} + \cdots$, converging for every real $x$. Notice how the interval differs from row to row even when the radius is the same, which is precisely why endpoints must be checked and never assumed.
What Are Some Worked Power Series Examples?
Each example below finds $R$ with the ratio test, then tests both endpoints. They are chosen so that between them every endpoint outcome appears: both excluded, one included, and both included.
Example 1: Find the interval of convergence of $\displaystyle\sum_{n=1}^{\infty} \frac{(x-2)^n}{n}$.
Apply the ratio test, centered at $a = 2$:
$$L = \lim_{n \to \infty} \left| \frac{(x-2)^{n+1}}{n+1} \cdot \frac{n}{(x-2)^n} \right| = \lim_{n \to \infty} |x - 2| \cdot \frac{n}{n+1} = |x - 2|$$
Convergence needs $L < 1$, so $|x - 2| < 1$, giving $R = 1$ and the open interval $(1, 3)$. Test the endpoints in the original series:
At $x = 3$: $\displaystyle\sum \frac{1}{n}$, the harmonic series, which diverges.
At $x = 1$: $\displaystyle\sum \frac{(-1)^n}{n}$, the alternating harmonic series, which converges.
Final answer: the interval of convergence is $[1, 3)$.
Example 2: Find the interval of convergence of $\displaystyle\sum_{n=1}^{\infty} \frac{(x+1)^n}{n^2 , 3^n}$.
Centered at $a = -1$, the ratio test gives:
$$L = \lim_{n \to \infty} \left| \frac{(x+1)^{n+1}}{(n+1)^2 , 3^{n+1}} \cdot \frac{n^2 , 3^n}{(x+1)^n} \right| = \frac{|x+1|}{3} \lim_{n \to \infty} \frac{n^2}{(n+1)^2} = \frac{|x+1|}{3}$$
Convergence needs $\frac{|x+1|}{3} < 1$, so $|x + 1| < 3$, giving $R = 3$ and the open interval $(-4, 2)$. Test the endpoints:
At $x = 2$: $\displaystyle\sum \frac{3^n}{n^2 , 3^n} = \sum \frac{1}{n^2}$, a $p$-series with $p = 2$, which converges.
At $x = -4$: $\displaystyle\sum \frac{(-3)^n}{n^2 , 3^n} = \sum \frac{(-1)^n}{n^2}$, which converges absolutely.
Final answer: the interval of convergence is $[-4, 2]$, both endpoints included.
Example 3: Find the radius of convergence of $\displaystyle\sum_{n=0}^{\infty} \frac{x^n}{n \text{ factorial}}$.
The ratio of consecutive terms collapses because the factorial denominator grows without bound:
$$L = \lim_{n \to \infty} \left| \frac{x^{n+1}}{(n+1) \text{ factorial}} \cdot \frac{n \text{ factorial}}{x^n} \right| = \lim_{n \to \infty} \frac{|x|}{n+1} = 0$$
Since $L = 0 < 1$ for every real $x$, the series converges everywhere.
Final answer: $R = \infty$ and the interval of convergence is $(-\infty, \infty)$.
Why Does A Power Series Converge Only On An Interval?
The interval is not an arbitrary rule; it falls out of a tug-of-war inside each term. Two forces decide whether the endless sum settles:
The powers. As $n$ grows, $(x - a)^n$ shrinks when $|x - a| < 1$ and grows when $|x - a| > 1$. Distance from the center is the dial that controls it.
The coefficients. The $c_n$ can amplify or damp those powers. The ratio test measures the two forces together and reports the exact break-even distance, which is $R$.
Because the deciding quantity is a distance, $|x - a|$, the region of convergence is symmetric about the center: it reaches out the same $R$ in both directions. That symmetry is why the answer is always an interval (or a single point, or the whole line), never a scattered patchwork of $x$-values.
The endpoints sit exactly at the break-even distance, where the ratio test reads $L = 1$ and goes silent. There the two forces are balanced, and the finer details of the coefficients decide the outcome, which is why each endpoint has to be substituted back and judged on its own.
Who Discovered Power Series?
Power series did not begin in Europe. Centuries before Newton, mathematicians in southern India were already summing infinite series to compute $\pi$ and the values of sine and cosine to remarkable accuracy.
Two more figures shaped the modern theory:
Isaac Newton (1643 to 1727, England) treated power series as infinite polynomials he could multiply, divide, and integrate freely, making them a central engine of his calculus.
Augustin-Louis Cauchy (1789 to 1857, France) supplied the rigor, defining the radius of convergence precisely and settling exactly when a power series is allowed to represent a function.
Where Are Power Series Used In The Real World?
The idea of trading a hard function for an easy sum of powers runs quietly through modern technology and science.
Calculators and computers: functions like $\sin x$, $e^x$, and $\ln x$ have no simple formula a chip can evaluate directly, so the hardware sums the first several terms of a power series to the required accuracy.
Physics: the small-angle approximation $\sin \theta \approx \theta$ is just the first term of the sine power series, and it underlies the study of pendulums, optics, and oscillations.
Signal processing: the Z-transform, the backbone of digital filters, is a power series in disguise, with the signal's samples as its coefficients.
Differential equations: many equations that have no closed-form solution can still be solved by assuming the answer is a power series and solving for its coefficients one by one.
Economics and probability: generating functions encode a whole sequence of numbers as the coefficients of one power series, turning counting problems into algebra, a technique close in spirit to the binomial theorem.
One idea, replace a function by rising powers, lets a pocket calculator find a sine, a physicist tame a pendulum, and an engineer design a filter.
What Are The Most Common Mistakes With Power Series?
These four errors account for most lost marks on power series, verified against the Albert.io AP Calculus BC review, university exam review sheets (University of Colorado Math 2300, University of Michigan Math 116), and the num8ers AP Calc BC guide.
Forgetting to test the endpoints.
Where it slips in:
A student finds $R$ with the ratio test, writes the open interval, and stops, never substituting the two endpoints back into the series.
Don't do this:
Do not report $(a - R, ; a + R)$ as the final interval by default. The ratio test is silent at the endpoints, so it has told you nothing about them.
The correct way:
Substitute each endpoint into the original series and test it separately with another tool (alternating series, $p$-series, comparison). Include an endpoint only if that series converges.
Confusing the ratio-test limit with the radius.
Where it slips in:
The ratio test produces something like $\frac{|x - a|}{3} < 1$, and the student writes $R = \frac{1}{3}$ by reading off the fraction.
Don't do this:
Do not treat the coefficient in the limit as $R$. Here $\frac{|x-a|}{3} < 1$ rearranges to $|x - a| < 3$.
The correct way:
Finish solving the inequality $L < 1$ for $|x - a|$. Whatever number sits on the right of $|x - a| < R$ is the radius, so in this case $R = 3$.
Dropping the absolute value in the ratio test.
Where it slips in:
A student cancels the powers of $(x - a)$ without the bars, loses track of a sign, and ends up with a one-sided or negative "radius".
Don't do this:
Do not run the ratio test on signed terms. A radius is a distance and cannot be negative.
The correct way:
Keep the absolute-value bars around the whole ratio until the limit is evaluated, then solve for $|x - a|$.
Using the wrong bracket at an endpoint.
Where it slips in:
After correctly testing the endpoints, the student writes a square bracket where the series diverged, or a round bracket where it converged.
Don't do this:
Do not guess the bracket. A diverging endpoint is never included; a converging one always is.
The correct way:
Match the bracket to the test result: square bracket for a converging endpoint, round bracket for a diverging one. In Example 1 that produced $[1, 3)$.
Practice Problems On Power Series
Find the radius and interval of convergence for each. Answers follow each line.
$\displaystyle\sum_{n=0}^{\infty} \frac{x^n}{2^n}$.
(Answer: $L = \frac{|x|}{2}$, so $R = 2$; both endpoints $x = \pm 2$ give $\sum (\pm 1)^n$, which diverge. Interval $(-2, 2)$.)$\displaystyle\sum_{n=1}^{\infty} \frac{x^n}{n^2}$.
(Answer: $R = 1$; at $x = 1$, $\sum \frac{1}{n^2}$ converges, at $x = -1$, $\sum \frac{(-1)^n}{n^2}$ converges. Interval $[-1, 1]$.)$\displaystyle\sum_{n=0}^{\infty} n , x^n$.
(Answer: $R = 1$; both endpoints give terms that do not shrink to $0$, so both diverge. Interval $(-1, 1)$.)$\displaystyle\sum_{n=1}^{\infty} \frac{(x-5)^n}{n}$.
(Answer: $R = 1$, center $5$; at $x = 6$ harmonic diverges, at $x = 4$ alternating harmonic converges. Interval $[4, 6)$.)$\displaystyle\sum_{n=0}^{\infty} \frac{(2x)^n}{n \text{ factorial}}$.
(Answer: ratio $\frac{2|x|}{n+1} \to 0$ for all $x$, so $R = \infty$. Interval $(-\infty, \infty)$.)Use the geometric series to write $\dfrac{1}{1 - 2x}$ as a power series and give its interval.
(Answer: $\sum_{n=0}^{\infty} (2x)^n = \sum 2^n x^n$, converging when $|2x| < 1$, so $R = \frac{1}{2}$ and the interval is $\left(-\frac{1}{2}, \frac{1}{2}\right)$.)
Where Should You Go Next After Power Series?
Power series sit at the crossroads of sequences, series, and calculus, and several natural doors open from here.
Sequences and series. Firm up the convergence tests (comparison, alternating, $p$-series) that decide the endpoints of every power series.
Geometric progressions. The finite sum $\frac{1 - x^{N+1}}{1-x}$ is the exact engine behind the geometric power series, so revisit where it comes from.
The binomial theorem. Its infinite generalisation, the binomial series, is a power series in its own right and expands $(1 + x)^k$ for any power $k$.
If your child is building toward AP Calculus BC or a first university analysis course, a live Bhanzu trainer teaches power series starting from the "why", the shrinking terms and the balance at the endpoints, in the Bhanzu math program.
Was this article helpful?
Your feedback helps us write better content