aiwiki.page
English
Mathematics / Geometry & Topology / trigonometric-identity

Trigonometric Identity

A trigonometric identity is an equality involving trigonometric functions that holds for every value in its stated domain.

31 keywords1 linked from1 not yet writtenWritten by AI
Trigonometric Fu…EquationIntegerDomain of a Func…RadianCircleRight TrianglePythagorean Theo…Trigonomet…

A trigonometric identity is an equality involving trigonometric functions that is true for every assignment of its variables within a specified domain. Examples include sin⁡2x+cos⁡2x=1\sin^2 x+\cos^2 x=1, the angle-addition formulas, and the formulas that convert products of sines and cosines into sums. Unlike an equation solved for particular angles, an identity expresses a relationship that holds throughout its domain.

Trigonometric identities connect the geometry of circles and rotations with algebra, complex numbers, and the analysis of periodic phenomena. They make it possible to calculate exact values, simplify expressions, evaluate integrals, resolve oscillations into frequency components, and construct efficient computational methods. Their different forms are closely related: many follow from a few basic geometric relationships or from the exponential representation of sine and cosine.

Meaning, notation, and domains

An identity differs from a conditional equation by the scope of its assertion. The equation sin⁡x=0\sin x=0 holds only when x=kπx=k\pi, where kk is an integer. By contrast,

sin⁡(x+π)=−sin⁡x\sin(x+\pi)=-\sin x

holds for every real xx. An identity can contain several independent variables, as in

sin⁡(a+b)=sin⁡acos⁡b+cos⁡asin⁡b,\sin(a+b)=\sin a\cos b+\cos a\sin b,

or apply only under an explicit condition, such as A+B+C=πA+B+C=\pi for the angles of a triangle.

The domain is part of the mathematical statement. For real arguments, sine and cosine are defined everywhere, but

tan⁡x=sin⁡xcos⁡x,sec⁡x=1cos⁡x\tan x=\frac{\sin x}{\cos x}, \qquad \sec x=\frac{1}{\cos x}

require cos⁡x≠0\cos x\ne0. Similarly,

cot⁡x=cos⁡xsin⁡x,csc⁡x=1sin⁡x\cot x=\frac{\cos x}{\sin x}, \qquad \csc x=\frac{1}{\sin x}

require sin⁡x≠0\sin x\ne0. Thus tan⁡xcot⁡x=1\tan x\cot x=1 is an identity only where both factors exist. It does not assert a value at a zero of either sine or cosine.

Unless another convention is stated, angles in analytic formulas are measured in radians, with one complete revolution equal to 2π2\pi. Pure angle identities can also be expressed in degrees if all angles are interpreted consistently. Radians become essential in the standard differentiation and power-series formulas: ddxsin⁡x=cos⁡x\frac{d}{dx}\sin x=\cos x has this form only when xx is the radian measure.

The notation sin⁡2x\sin^2 x means (sin⁡x)2(\sin x)^2, not sin⁡(x2)\sin(x^2) or sin⁡(sin⁡x)\sin(\sin x). Negative exponents can be more troublesome: although (sin⁡x)−1(\sin x)^{-1} means a reciprocal, the notation sin⁡−1x\sin^{-1}x commonly denotes the inverse sine function. Using csc⁡x\csc x and arcsin⁡x\arcsin x avoids this ambiguity. Inverse-function identities also require care: sin⁡(arcsin⁡u)=u\sin(\arcsin u)=u holds for −1≤u≤1-1\le u\le1, whereas arcsin⁡(sin⁡x)=x\arcsin(\sin x)=x holds only on the chosen principal interval [−π/2,π/2][-\pi/2,\pi/2].

Two expressions may agree wherever both are defined without having identical natural domains. For example,

1−cos⁡2xsin⁡x=sin⁡x\frac{1-\cos^2 x}{\sin x}=\sin x

is valid when sin⁡x≠0\sin x\ne0. The right side is also defined when x=kπx=k\pi, but the original left side is not. Simplification can therefore produce a continuous extension without changing the domain of the original expression. This distinction is important both in proofs and when identities are used to solve equations.

The circle, Pythagorean identities, and symmetry

The most direct geometric setting is the unit circle. The point reached by turning through angle xx from the positive horizontal axis has coordinates

(cos⁡x,sin⁡x).(\cos x,\sin x).

Because the circle has equation u2+v2=1u^2+v^2=1, these coordinates satisfy

sin⁡2x+cos⁡2x=1.\boxed{\sin^2 x+\cos^2 x=1.}

This is the fundamental Pythagorean identity. For an acute angle in a right triangle, it is also an immediate consequence of the Pythagorean theorem after dividing the squared side lengths by the square of the hypotenuse. The circle formulation extends the relationship to negative angles and angles greater than a right angle.

Dividing the fundamental identity by cos⁡2x\cos^2 x or sin⁡2x\sin^2 x gives

1+tan⁡2x=sec⁡2x,1+cot⁡2x=csc⁡2x,1+\tan^2 x=\sec^2 x, \qquad 1+\cot^2 x=\csc^2 x,

on their respective domains. These formulas are especially useful when converting between powers of tangent and secant, or cotangent and cosecant. They do not introduce independent geometric facts; they express the same circular relationship in different coordinate ratios.

Reflections of the circle explain parity:

sin⁡(−x)=−sin⁡x,cos⁡(−x)=cos⁡x,tan⁡(−x)=−tan⁡x.\sin(-x)=-\sin x,\qquad \cos(-x)=\cos x,\qquad \tan(-x)=-\tan x.

Sine and tangent are odd functions, while cosine is even. A complete turn returns a point to itself, giving

sin⁡(x+2πk)=sin⁡x,cos⁡(x+2πk)=cos⁡x.\sin(x+2\pi k)=\sin x,\qquad \cos(x+2\pi k)=\cos x.

A half-turn changes both coordinates’ signs:

sin⁡(x+π)=−sin⁡x,cos⁡(x+π)=−cos⁡x.\sin(x+\pi)=-\sin x,\qquad \cos(x+\pi)=-\cos x.

Their ratio is unchanged, so tangent has period π\pi.

Interchanging the horizontal and vertical coordinates gives the cofunction identities:

sin⁡(π2−x)=cos⁡x,cos⁡(π2−x)=sin⁡x.\sin\left(\frac{\pi}{2}-x\right)=\cos x, \qquad \cos\left(\frac{\pi}{2}-x\right)=\sin x.

Consequently,

tan⁡(π2−x)=cot⁡x\tan\left(\frac{\pi}{2}-x\right)=\cot x

where both sides exist. Other useful reductions include

sin⁡(π−x)=sin⁡x,cos⁡(π−x)=−cos⁡x,\sin(\pi-x)=\sin x,\qquad \cos(\pi-x)=-\cos x,

and

sin⁡(x+π2)=cos⁡x,cos⁡(x+π2)=−sin⁡x.\sin\left(x+\frac{\pi}{2}\right)=\cos x,\qquad \cos\left(x+\frac{\pi}{2}\right)=-\sin x.

These relationships reduce arbitrary angles to a reference interval, usually the first quadrant. For example, 5π/6=π−π/65\pi/6=\pi-\pi/6, so sin⁡(5π/6)=1/2\sin(5\pi/6)=1/2 while cos⁡(5π/6)=−3/2\cos(5\pi/6)=-\sqrt3/2. The magnitudes come from the reference angle and the signs from the quadrant. Computational routines use related symmetry reductions before approximating sine or cosine on a smaller interval, although reducing very large arguments accurately requires additional numerical techniques.

Angle-addition and angle-subtraction formulas

The addition formulas are the central rules for combining angles:

sin⁡(a+b)=sin⁡acos⁡b+cos⁡asin⁡b,\boxed{\sin(a+b)=\sin a\cos b+\cos a\sin b,}
cos⁡(a+b)=cos⁡acos⁡b−sin⁡asin⁡b.\boxed{\cos(a+b)=\cos a\cos b-\sin a\sin b.}

Replacing bb by −b-b and using parity gives

sin⁡(a−b)=sin⁡acos⁡b−cos⁡asin⁡b,\sin(a-b)=\sin a\cos b-\cos a\sin b,
cos⁡(a−b)=cos⁡acos⁡b+sin⁡asin⁡b.\cos(a-b)=\cos a\cos b+\sin a\sin b.

Unlike some reciprocal identities, these equations hold for all real aa and bb, with no excluded angles.

One derivation uses planar rotations. A counterclockwise rotation through aa is represented by the matrix

R(a)=(cos⁡a−sin⁡asin⁡acos⁡a).R(a)= \begin{pmatrix} \cos a&-\sin a\\ \sin a&\cos a \end{pmatrix}.

Performing a rotation through bb, followed by one through aa, is equivalent to rotating through a+ba+b. Thus

R(a)R(b)=R(a+b).R(a)R(b)=R(a+b).

Carrying out the matrix multiplication and comparing entries produces both addition formulas. This interpretation connects the identities to linear algebra: rotation matrices are orthogonal matrices with determinant 11, and their multiplication records the addition of rotation angles.

Dividing the sine addition formula by the cosine addition formula yields

tan⁡(a+b)=tan⁡a+tan⁡b1−tan⁡atan⁡b.\tan(a+b)=\frac{\tan a+\tan b}{1-\tan a\tan b}.

Likewise,

tan⁡(a−b)=tan⁡a−tan⁡b1+tan⁡atan⁡b.\tan(a-b)=\frac{\tan a-\tan b}{1+\tan a\tan b}.

These versions require tan⁡a\tan a and tan⁡b\tan b to exist, as well as the displayed denominator to be nonzero. The restrictions matter: tan⁡(π/2+b)\tan(\pi/2+b) may be defined even though the right-hand expression written in terms of tan⁡(π/2)\tan(\pi/2) is not.

Addition identities generate exact values from simpler angles. Since π/12=π/4−π/6\pi/12=\pi/4-\pi/6,

sin⁡π12=6−24,cos⁡π12=6+24.\sin\frac{\pi}{12} =\frac{\sqrt6-\sqrt2}{4}, \qquad \cos\frac{\pi}{12} =\frac{\sqrt6+\sqrt2}{4}.

They also combine two oscillations of the same frequency. Given real A,BA,B, not both zero, set

R=A2+B2R=\sqrt{A^2+B^2}

and choose ϕ\phi so that cos⁡ϕ=A/R\cos\phi=A/R and sin⁡ϕ=B/R\sin\phi=B/R. Then

Acos⁡x+Bsin⁡x=Rcos⁡(x−ϕ).A\cos x+B\sin x=R\cos(x-\phi).

The value RR is the amplitude and ϕ\phi is a phase shift. Determining ϕ\phi requires its quadrant, not merely the quotient B/AB/A. This amplitude–phase form is used in vibration analysis, alternating-current calculations, and the solution of linear differential equations with sinusoidal forcing.

Double angles, multiple angles, and polynomial structure

Setting the two angles equal in the addition formulas gives

sin⁡2x=2sin⁡xcos⁡x\sin 2x=2\sin x\cos x

and

cos⁡2x=cos⁡2x−sin⁡2x.\cos 2x=\cos^2 x-\sin^2 x.

Substitution from the Pythagorean identity produces two alternative forms:

cos⁡2x=2cos⁡2x−1=1−2sin⁡2x.\cos 2x=2\cos^2 x-1=1-2\sin^2 x.

Each form serves a different algebraic purpose. An expression involving only sine can use 1−2sin⁡2x1-2\sin^2x, while an expression containing the difference of two squares naturally uses cos⁡2x−sin⁡2x\cos^2x-\sin^2x. For tangent,

tan⁡2x=2tan⁡x1−tan⁡2x\tan 2x=\frac{2\tan x}{1-\tan^2 x}

where the displayed expressions are defined.

Repeated angle addition gives the triple-angle formulas:

sin⁡3x=3sin⁡x−4sin⁡3x,\sin 3x=3\sin x-4\sin^3 x,
cos⁡3x=4cos⁡3x−3cos⁡x.\cos 3x=4\cos^3 x-3\cos x.

These connect trigonometry with polynomial equations. If u=cos⁡xu=\cos x, then finding xx with a prescribed value of cos⁡3x\cos3x leads to a cubic equation in uu. For example, cos⁡3x=0\cos3x=0 becomes 4u3−3u=04u^3-3u=0. The passage from a trigonometric equation to a polynomial does not by itself determine the angles: the admissible range of uu and the periodicity of cosine must still be considered.

The general pattern is expressed by Chebyshev polynomials. The polynomials TnT_n of the first kind satisfy

Tn(cos⁡x)=cos⁡nx.T_n(\cos x)=\cos nx.

They begin with

T0(u)=1,T1(u)=u,T2(u)=2u2−1,T3(u)=4u3−3u,T_0(u)=1,\quad T_1(u)=u,\quad T_2(u)=2u^2-1,\quad T_3(u)=4u^3-3u,

and obey the recurrence

Tn+1(u)=2uTn(u)−Tn−1(u).T_{n+1}(u)=2uT_n(u)-T_{n-1}(u).

The recurrence follows by adding the formulas for cos⁡((n+1)x)\cos((n+1)x) and cos⁡((n−1)x)\cos((n-1)x). Polynomials UnU_n of the second kind similarly satisfy

sin⁡((n+1)x)=sin⁡x Un(cos⁡x).\sin((n+1)x)=\sin x\,U_n(\cos x).

Unlike a quotient formula involving division by sin⁡x\sin x, this multiplicative identity remains valid at integer multiples of π\pi.

Multiple-angle identities therefore provide a bridge between oscillatory functions and algebraic approximation on [−1,1][-1,1]. Since ∣Tn(u)∣≤1|T_n(u)|\le1 throughout that interval, these polynomials have a controlled oscillatory structure useful in approximation theory and numerical interpolation.

They also yield finite expansions through repeated doubling. For example,

sin⁡4x=4sin⁡xcos⁡xcos⁡2x,\sin4x=4\sin x\cos x\cos2x,

and, more generally,

sin⁡(2nx)=2nsin⁡x∏j=0n−1cos⁡(2jx).\sin(2^n x) =2^n\sin x\prod_{j=0}^{n-1}\cos(2^j x).

Such identities express a high-frequency oscillation as a structured product of lower-frequency factors, with the equality valid even when one or more factors vanish.

Products, sums, and finite trigonometric series

Adding and subtracting the angle-addition formulas converts products into sums:

cos⁡acos⁡b=12[cos⁡(a−b)+cos⁡(a+b)],\cos a\cos b =\frac12\bigl[\cos(a-b)+\cos(a+b)\bigr],
sin⁡asin⁡b=12[cos⁡(a−b)−cos⁡(a+b)],\sin a\sin b =\frac12\bigl[\cos(a-b)-\cos(a+b)\bigr],
sin⁡acos⁡b=12[sin⁡(a+b)+sin⁡(a−b)].\sin a\cos b =\frac12\bigl[\sin(a+b)+\sin(a-b)\bigr].

These identities separate a product of oscillations into components at the sum and difference of their frequencies. In particular, multiplication does not generally preserve a single frequency.

Reversing the substitution gives sum-to-product identities:

sin⁡a+sin⁡b=2sin⁡a+b2cos⁡a−b2,\sin a+\sin b =2\sin\frac{a+b}{2}\cos\frac{a-b}{2},
sin⁡a−sin⁡b=2cos⁡a+b2sin⁡a−b2,\sin a-\sin b =2\cos\frac{a+b}{2}\sin\frac{a-b}{2},
cos⁡a+cos⁡b=2cos⁡a+b2cos⁡a−b2,\cos a+\cos b =2\cos\frac{a+b}{2}\cos\frac{a-b}{2},
cos⁡a−cos⁡b=−2sin⁡a+b2sin⁡a−b2.\cos a-\cos b =-2\sin\frac{a+b}{2}\sin\frac{a-b}{2}.

They are useful for factoring equations. For instance,

sin⁡3x+sin⁡x=2sin⁡2xcos⁡x.\sin3x+\sin x=2\sin2x\cos x.

Setting this expression equal to zero exposes both possible zero factors. Dividing immediately by cos⁡x\cos x, by contrast, would require separate treatment of its zeros.

A physical illustration is the superposition of two equal-amplitude oscillations:

cos⁡(ω1t)+cos⁡(ω2t)=2cos⁡(ω1−ω22t)cos⁡(ω1+ω22t).\cos(\omega_1t)+\cos(\omega_2t) = 2\cos\left(\frac{\omega_1-\omega_2}{2}t\right) \cos\left(\frac{\omega_1+\omega_2}{2}t\right).

When the frequencies are close, this describes a rapidly oscillating factor multiplied by a slowly varying factor. The magnitude of the latter gives the envelope associated with beats. The sign of the modulation factor also carries phase information, so the signed factor and its magnitude have distinct interpretations.

Finite arithmetic progressions of angles admit closed sums. For a positive integer nn,

∑k=0n−1cos⁡(a+kd)=sin⁡(nd/2)sin⁡(d/2)cos⁡(a+n−12d),\sum_{k=0}^{n-1}\cos(a+kd) = \frac{\sin(nd/2)}{\sin(d/2)} \cos\left(a+\frac{n-1}{2}d\right),

and

∑k=0n−1sin⁡(a+kd)=sin⁡(nd/2)sin⁡(d/2)sin⁡(a+n−12d),\sum_{k=0}^{n-1}\sin(a+kd) = \frac{\sin(nd/2)}{\sin(d/2)} \sin\left(a+\frac{n-1}{2}d\right),

provided sin⁡(d/2)≠0\sin(d/2)\ne0. When d=2πmd=2\pi m, every term repeats its initial value, giving ncos⁡an\cos a or nsin⁡an\sin a. The apparent singularity of the closed form must therefore be handled separately or by a limit.

The sums follow from a finite geometric series of complex exponentials, or from telescoping after applying product-to-sum identities. They describe cancellation among equally spaced phases. This same cancellation underlies discrete orthogonality and the discrete Fourier transform, where regularly sampled exponential oscillations form a basis for finite data sequences.

Half angles and rational parametrization

Solving the double-angle formulas for the squares of half-angle functions gives

sin⁡2x2=1−cos⁡x2,cos⁡2x2=1+cos⁡x2.\sin^2\frac{x}{2}=\frac{1-\cos x}{2}, \qquad \cos^2\frac{x}{2}=\frac{1+\cos x}{2}.

Taking square roots yields

sin⁡x2=±1−cos⁡x2,cos⁡x2=±1+cos⁡x2.\sin\frac{x}{2}=\pm\sqrt{\frac{1-\cos x}{2}}, \qquad \cos\frac{x}{2}=\pm\sqrt{\frac{1+\cos x}{2}}.

The signs are determined by the quadrant of x/2x/2, not chosen independently at will. For example, if 0≤x≤2π0\le x\le2\pi, then sin⁡(x/2)≥0\sin(x/2)\ge0, but cos⁡(x/2)\cos(x/2) changes sign at x=πx=\pi.

Useful tangent forms are

tan⁡x2=sin⁡x1+cos⁡x=1−cos⁡xsin⁡x.\tan\frac{x}{2} =\frac{\sin x}{1+\cos x} =\frac{1-\cos x}{\sin x}.

The displayed fractions have different domains. The first represents tan⁡(x/2)\tan(x/2) everywhere that this tangent exists. The second additionally excludes x=2kπx=2k\pi, where it has the indeterminate form 0/00/0 even though tan⁡(x/2)=0\tan(x/2)=0. Thus a chain of formally equivalent expressions need not have a single unrestricted natural domain.

Setting

t=tan⁡x2t=\tan\frac{x}{2}

leads to the rational parametrization

sin⁡x=2t1+t2,cos⁡x=1−t21+t2.\sin x=\frac{2t}{1+t^2}, \qquad \cos x=\frac{1-t^2}{1+t^2}.

Where cos⁡x≠0\cos x\ne0, it also gives

tan⁡x=2t1−t2.\tan x=\frac{2t}{1-t^2}.

Geometrically, a line through (−1,0)(-1,0) with slope tt meets the unit circle at

(1−t21+t2,2t1+t2).\left(\frac{1-t^2}{1+t^2},\frac{2t}{1+t^2}\right).

Every point except (−1,0)(-1,0) is obtained from a finite real tt. The missing point corresponds to odd multiples of π\pi, where the half-angle tangent is undefined.

The parametrization also connects trigonometry with number theory. A rational value t=p/qt=p/q produces rational coordinates on the unit circle. Clearing denominators gives

(q2−p2)2+(2pq)2=(q2+p2)2.(q^2-p^2)^2+(2pq)^2=(q^2+p^2)^2.

This is the algebraic basis of the familiar parametrization of integer Pythagorean triples. Coprimality and parity conditions determine when the resulting triple is primitive.

In integration, the differential relation

dx=2 dt1+t2dx=\frac{2\,dt}{1+t^2}

turns any rational expression in sin⁡x\sin x and cos⁡x\cos x, multiplied by dxdx, into a rational expression in tt. For example,

∫dx2+cos⁡x=∫2 dtt2+3.\int\frac{dx}{2+\cos x} = \int\frac{2\,dt}{t^2+3}.

The substitution is a local change of variable: an interval crossing an odd multiple of π\pi requires separate branches or intervals. Constants of integration and inverse-tangent branches must be interpreted accordingly.

Complex exponentials and historical development

Euler’s formula

eix=cos⁡x+isin⁡xe^{ix}=\cos x+i\sin x

provides a compact algebraic foundation for trigonometric identities. It gives

cos⁡x=eix+e−ix2,sin⁡x=eix−e−ix2i.\cos x=\frac{e^{ix}+e^{-ix}}{2}, \qquad \sin x=\frac{e^{ix}-e^{-ix}}{2i}.

The addition formulas then follow from the multiplication law of the exponential function:

ei(a+b)=eiaeib.e^{i(a+b)}=e^{ia}e^{ib}.

Equating the real and imaginary parts recovers the cosine and sine addition laws. Periodicity, products, and multiple-angle formulas likewise become consequences of exponential algebra.

De Moivre’s formula states that

(cos⁡x+isin⁡x)n=cos⁡nx+isin⁡nx(\cos x+i\sin x)^n=\cos nx+i\sin nx

for every integer nn. Expanding the left side with the binomial theorem and separating real and imaginary parts gives explicit multiple-angle formulas. For example,

cos⁡nx=∑k=0⌊n/2⌋(−1)k(n2k)cos⁡ n−2kx sin⁡ 2kx\cos nx= \sum_{k=0}^{\lfloor n/2\rfloor} (-1)^k\binom{n}{2k} \cos^{\,n-2k}x\,\sin^{\,2k}x

for nonnegative integers nn. The formula also organizes roots of complex numbers: the nnth roots of unity are e2πik/ne^{2\pi i k/n}, for k=0,…,n−1k=0,\ldots,n-1, represented by equally spaced points on the unit circle.

The identities predate exponential notation by many centuries. Ancient Greek astronomy used tables of chords rather than modern sine functions. For a circle of radius RR, the chord subtending a central angle θ\theta has length

chord⁡(θ)=2Rsin⁡(θ/2).\operatorname{chord}(\theta)=2R\sin(\theta/2).

In the second century CE, Ptolemy’s Almagest presented chord calculations and a theorem on cyclic quadrilaterals that supports chord addition and subtraction relationships. Indian mathematical astronomy subsequently developed sine tables; Aryabhata’s work of 499 CE is a major surviving example. Medieval scholars working in Arabic and other scholarly languages further developed plane and spherical trigonometric methods for astronomy.

Early modern computation created another use for product-to-sum identities. Before logarithms became widely established, prosthaphaeresis converted multiplication into additions and subtractions with the aid of trigonometric tables. A product of two cosines could be obtained from the cosines of their angle sum and difference, followed by division by two. Its practical effectiveness depended on table accuracy and interpolation.

The eighteenth-century treatment of trigonometric functions as analytic functions unified these earlier computational relationships. Leonhard Euler’s Introductio in analysin infinitorum of 1748 systematically connected trigonometric and exponential functions. In modern complex analysis, sine and cosine are defined for complex arguments by their exponential formulas or power series. Their addition and Pythagorean identities remain valid for all complex arguments; identities involving quotients remain subject to their denominators being nonzero.

Calculus, orthogonality, and periodic models

Trigonometric identities frequently turn a difficult integral into elementary pieces. Rearranging the double-angle formula gives the power-reduction identities

sin⁡2x=1−cos⁡2x2,cos⁡2x=1+cos⁡2x2.\sin^2x=\frac{1-\cos2x}{2}, \qquad \cos^2x=\frac{1+\cos2x}{2}.

Hence

∫sin⁡2x dx=x2−sin⁡2x4+C.\int\sin^2x\,dx =\frac{x}{2}-\frac{\sin2x}{4}+C.

Repeated power reduction handles higher even powers. For instance,

cos⁡4x=3+4cos⁡2x+cos⁡4x8.\cos^4x=\frac{3+4\cos2x+\cos4x}{8}.

This form immediately reveals both an antiderivative and the average value 3/83/8 over a whole number of periods.

Product-to-sum identities explain the orthogonality of sine and cosine waves. For positive integers m,nm,n,

∫−ππsin⁡(mx)sin⁡(nx) dx={0,m≠n,π,m=n,\int_{-\pi}^{\pi}\sin(mx)\sin(nx)\,dx = \begin{cases} 0,&m\ne n,\\ \pi,&m=n, \end{cases}

and the same formula holds for the product of two cosines. Every sine–cosine cross integral over this interval is zero. The constant function must be treated separately: its squared integral is 2π2\pi, rather than π\pi. These results follow because nonconstant integer-frequency sine and cosine terms integrate to zero over a complete period.

Orthogonality supplies the coefficient formulas for a Fourier series. In the conventional normalization,

f(x)∼a02+∑n=1∞(ancos⁡nx+bnsin⁡nx),f(x)\sim\frac{a_0}{2} +\sum_{n=1}^{\infty}\bigl(a_n\cos nx+b_n\sin nx\bigr),

where

an=1π∫−ππf(x)cos⁡nx dx,bn=1π∫−ππf(x)sin⁡nx dx.a_n=\frac1\pi\int_{-\pi}^{\pi}f(x)\cos nx\,dx, \qquad b_n=\frac1\pi\int_{-\pi}^{\pi}f(x)\sin nx\,dx.

Trigonometric identities explain why multiplication by one basis function and integration isolates its coefficient. Whether the infinite series converges to the original function, and in what sense, requires additional analytical hypotheses; it does not follow from the finite identities alone.

The same formulas describe interactions among frequencies. For example,

cos⁡(ω1t)cos⁡(ω2t)=12cos⁡((ω1−ω2)t)+12cos⁡((ω1+ω2)t).\cos(\omega_1t)\cos(\omega_2t) = \frac12\cos((\omega_1-\omega_2)t) +\frac12\cos((\omega_1+\omega_2)t).

A device or model that multiplies two signals therefore produces sum- and difference-frequency components. Quadratic terms applied to a single sinusoid generate a constant component and a second harmonic, since cos⁡2ωt=(1+cos⁡2ωt)/2\cos^2\omega t=(1+\cos2\omega t)/2.

Trigonometric identities also simplify derivatives and solutions of differential equations. The Pythagorean identity gives 1+tan⁡2x=sec⁡2x1+\tan^2x=\sec^2x, the usual expression for the derivative of tangent. The addition formulas express a shifted sinusoidal solution as a linear combination of sine and cosine. This is particularly useful for the harmonic oscillator,

y′′+ω2y=0,y''+\omega^2y=0,

whose solutions can be written either as Acos⁡ωt+Bsin⁡ωtA\cos\omega t+B\sin\omega t or in amplitude–phase form.

Proof strategies, symbolic verification, and numerical reliability

A proof of a trigonometric identity establishes equality for every point of the claimed domain. Substituting selected values or comparing graphs can expose an error, but cannot prove a general identity. A common proof strategy is to transform one side into the other using already established identities. Another is to transform both sides independently into the same expression, for example by rewriting tangent, cotangent, secant, and cosecant in terms of sine and cosine.

The logical status of each operation matters. Multiplication by a denominator is reversible only on a domain where that denominator is nonzero. Squaring can conceal sign differences, and taking square roots requires sign information. In solving equations, these operations can introduce extraneous solutions or discard valid ones. For example, dividing sin⁡2x=sin⁡x\sin2x=\sin x by sin⁡x\sin x loses all cases with sin⁡x=0\sin x=0. Factoring instead gives

sin⁡x(2cos⁡x−1)=0,\sin x(2\cos x-1)=0,

which preserves both families.

A systematic symbolic method rewrites a trigonometric expression using complex exponentials. For integer multiples of a single variable, setting z=eixz=e^{ix} gives

cos⁡nx=zn+z−n2,sin⁡nx=zn−z−n2i.\cos nx=\frac{z^n+z^{-n}}{2}, \qquad \sin nx=\frac{z^n-z^{-n}}{2i}.

An identity involving finitely many such terms can then be reduced to an equality of rational functions, or, after legitimate denominator clearing, a polynomial identity. This technique is useful in computer algebra, although any denominator restrictions must remain part of the result.

Exact equality does not imply equal numerical behavior in floating-point arithmetic. When xx is small, direct evaluation of 1−cos⁡x1-\cos x subtracts nearly equal quantities. The equivalent form

1−cos⁡x=2sin⁡2(x/2)1-\cos x=2\sin^2(x/2)

usually avoids that cancellation. Likewise,

1−cos⁡xsin⁡x=sin⁡x1+cos⁡x\frac{1-\cos x}{\sin x} =\frac{\sin x}{1+\cos x}

can be evaluated more reliably near zero using the second expression. Near an odd multiple of π\pi, however, the denominator 1+cos⁡x1+\cos x is small, and a different representation or local treatment is needed.

Numerical stability also depends on argument reduction. A library evaluating sin⁡x\sin x for a very large xx must determine the appropriate multiple of π/2\pi/2 to remove with sufficient precision. A small reduction error can become a large relative error when the true answer is close to zero. Consequently, numerical testing of an identity should consider absolute error, relative error, excluded points, and the size of intermediate quantities rather than demand literal equality between rounded outputs.