Euler's Formula
Why raising e to an imaginary power makes you spin in a circle — and why that's not a coincidence.
What the formula actually says
Three objects that seem to have nothing to do with each other:
\(e \approx 2.718...\) — the base of natural growth. Appears everywhere compounding happens: interest, populations, radioactive decay.
\(i = \sqrt{-1}\). A number that, when squared, gives \(-1\). Lives on a second axis perpendicular to the real number line.
\(\cos\theta\) and \(\sin\theta\) describe positions on a unit circle. Pure geometry — right triangles and angles.
Measured in radians. One full revolution = \(2\pi\). Half turn = \(\pi\). Quarter turn = \(\pi/2\).
Euler's formula says: when you raise \(e\) to the power \(i\theta\), the result is a point on the unit circle at angle \(\theta\). The exponential function and the circular functions are secretly the same thing.
The one-line summary: \(e^{i\theta}\) doesn't make you grow — it makes you rotate. The imaginary unit in the exponent converts "exponential scaling" into "circular motion."
The complex plane — where numbers have two coordinates
A real number lives on a line. A complex number lives on a plane. The horizontal axis is the real part; the vertical axis is the imaginary part.
The complex number \(a + bi\) is just the point \((a, b)\) on this plane. That's it. Nothing more mysterious than that.
The unit circle is the set of all complex numbers with magnitude 1 — all points at distance 1 from the origin. Every point on it can be written as \(\cos\theta + i\sin\theta\) for some angle \(\theta\).
The Pythagorean identity \(\cos^2\theta + \sin^2\theta = 1\) guarantees the magnitude is always 1. So \(\cos\theta + i\sin\theta\) walks around the unit circle as \(\theta\) increases.
What \(e^x\) actually is — the function that grows like itself
Before we can understand \(e^{i\theta}\), we need to deeply understand \(e^x\). Here's the real definition — not a number to memorize, but a property:
\(e^x\) is the unique function that is its own derivative. That is: if \(f(x) = e^x\), then \(f'(x) = e^x\). The rate of change at every point equals the value at that point.
This means: wherever you are, you're growing at exactly the speed of your current height. Double your height, double your speed. It's pure, self-referential growth.
From this one property, we can derive the Taylor series for \(e^x\) — the infinite polynomial that equals \(e^x\) exactly everywhere:
Think of the Taylor series like a recipe card. \(e^x\) is so perfectly self-consistent that you can describe it completely using just the value and all derivatives at a single point — say \(x=0\). Since every derivative of \(e^x\) equals \(e^x\), and \(e^0 = 1\), every coefficient in the recipe is \(\frac{1}{n!}\). The series is the exact, infinite recipe for the function.
Why each term is x^n / n! — the derivation
A Taylor series centered at \(x=0\) is built from the rule: the coefficient of \(x^n\) is \(\frac{f^{(n)}(0)}{n!}\), where \(f^{(n)}\) means the \(n\)-th derivative.
For \(e^x\): every derivative is \(e^x\), and at \(x=0\): \(e^0 = 1\). So every coefficient is \(\frac{1}{n!}\). Therefore:
Since \(0! = 1\) and \(1! = 1\), this simplifies to the familiar form. The factorials grow very fast, making each successive term smaller for any finite \(x\) — guaranteeing the series converges.
The crucial point: this series works for any input — real, complex, imaginary. As long as we can plug something into the series and have it converge, we can compute \(e^{\text{that thing}}\).
What happens when we plug in \(i\theta\)
Let's simply replace \(x\) with \(i\theta\) in the Taylor series. No tricks. Just substitution.
Now we need to simplify the powers of \(i\). This is where the magic lives. Remember \(i^2 = -1\), so:
Powers of \(i\) cycle with period 4: \(1, i, -1, -i, 1, i, -1, -i, \ldots\) It's like a clock with four positions. Multiply by \(i\) once = rotate 90° counterclockwise. Do it four times = full rotation = back to start. This is already a hint that \(i\) and rotation are deeply connected.
Now substitute the powers of \(i\) into the series. Each term \((i\theta)^n = i^n \cdot \theta^n\), so:
Splitting into real and imaginary parts — where \(\cos\) and \(\sin\) appear
Group the terms without \(i\) (real parts) separately from the terms with \(i\) (imaginary parts):
Now: what are the Taylor series for \(\cos\theta\) and \(\sin\theta\)? Let's derive them from scratch.
Applying the Taylor formula \(\sum \frac{f^{(n)}(0)}{n!}\theta^n\):
Compare these with the two halves of \(e^{i\theta}\) we found above. They are identical. Term by term, coefficient by coefficient:
| Term | Power of i | Result | Goes into... |
|---|---|---|---|
| \(\tfrac{(i\theta)^0}{0!} = 1\) | i⁰ = 1 | \(+1\) | cos (real) |
| \(\tfrac{(i\theta)^1}{1!} = i\theta\) | i¹ = i | \(+i\theta\) | sin (imaginary) |
| \(\tfrac{(i\theta)^2}{2!} = -\tfrac{\theta^2}{2!}\) | i² = −1 | \(-\tfrac{\theta^2}{2!}\) | cos (real) |
| \(\tfrac{(i\theta)^3}{3!} = -\tfrac{i\theta^3}{3!}\) | i³ = −i | \(-\tfrac{i\theta^3}{3!}\) | sin (imaginary) |
| \(\tfrac{(i\theta)^4}{4!} = +\tfrac{\theta^4}{4!}\) | i⁴ = +1 | \(+\tfrac{\theta^4}{4!}\) | cos (real) |
| \(\tfrac{(i\theta)^5}{5!} = +\tfrac{i\theta^5}{5!}\) | i⁵ = +i | \(+\tfrac{i\theta^5}{5!}\) | sin (imaginary) |
The even-power terms are all real (because even powers of \(i\) are real: \(\pm 1\)) and they reassemble into \(\cos\theta\). The odd-power terms all carry \(i\) (because odd powers of \(i\) are imaginary: \(\pm i\)) and they reassemble into \(i\sin\theta\). Therefore:
No hand-waving. The cycling of powers of \(i\) (period 4) routes the even terms to cosine and the odd terms to sine. That's the whole mechanism.
Interactive — watch \(e^{i\theta}\) walk around the unit circle
Drag the slider below. Watch the point \(e^{i\theta}\) move around the circle, and see how its real part (\(\cos\theta\)) and imaginary part (\(\sin\theta\)) change.
Unit circle — e^(iθ) interactive
The key geometric insight: as \(\theta\) increases from \(0\) to \(2\pi\), the point \(e^{i\theta}\) traces exactly one full circle. The real part \(\cos\theta\) oscillates left and right. The imaginary part \(\sin\theta\) oscillates up and down. Together they trace the circle.
The magnitude \(|e^{i\theta}| = \sqrt{\cos^2\theta + \sin^2\theta} = 1\) always. No matter what \(\theta\) is. This is why the imaginary exponent produces pure rotation — it can never shrink or grow the magnitude because the Pythagorean identity forbids it.
The famous special case: \(e^{i\pi} = -1\)
Set \(\theta = \pi\) (a half-turn, 180°).
Rearrange by adding 1 to both sides:
Five fundamental constants — \(e\), \(i\), \(\pi\), \(1\), \(0\) — in one equation. Each constant arises independently: \(e\) from calculus, \(i\) from algebra, \(\pi\) from geometry, \(1\) from counting, \(0\) from nothing. The fact that they satisfy this single clean relationship is not a trick — it's a deep structural truth about how rotation, growth, and number systems are unified.
Why this matters — the engineering payoff
Euler's formula is not just beautiful. It's a computational superpower. Here are the three most important consequences.
1. Sinusoids become exponentials — algebra becomes trivial
Any sinusoidal signal can be written as a complex exponential:
This matters enormously in AC circuits. Instead of carrying \(\cos\) and \(\sin\) through every calculation (constantly using trig identities), you work with \(e^{i\omega t}\) — which multiplies, divides, and differentiates cleanly. The phase \(\phi\) and amplitude \(A\) bundle into the single complex number \(Ae^{i\phi}\), called a phasor.
2. Differentiation becomes multiplication
If \(f(t) = e^{i\omega t}\), then:
Differentiating with respect to time just multiplies by \(i\omega\). That means solving differential equations for oscillating systems turns into solving algebraic equations — multiply instead of differentiate.
3. The trig identities fall out for free
From \(e^{i\theta} = \cos\theta + i\sin\theta\) alone:
And the angle addition formula becomes a single line — just multiply two exponentials:
Expand the right side and match real and imaginary parts to recover the double-angle identities instantly. No memorization needed — just Euler's formula and algebra.
The deepest takeaway: Euler's formula reveals that oscillation is rotation. A sine wave in time is a shadow of circular motion in the complex plane. The Fourier transform, which decomposes any signal into frequencies, is just decomposing a function into a sum of circular rotations at different speeds — all enabled by Euler's formula.