Sine: Difference between revisions

From Citizendium
Jump to navigation Jump to search
imported>Dmitrii Kouznetsov
(Should we copypast here also the "deduction" from the Euclidean geometry?)
 
imported>Meg Taylor
No edit summary
 
(13 intermediate revisions by 5 users not shown)
Line 1: Line 1:
{{copyedit}}
{{subpages}}
In [[trigonometry]], the function '''sine''' and its complementary function  '''cosine''', abbreviated '''sin''' and '''cos''', are central. They are used to express relations between angles and sides of [[triangle]]s. The functions are very common and important  in all branches of science and technology.  It is possible to define sine and cosine  in different ways. Two ways, one based on plane geometry and  one based on differential equations, will be discussed below.


'''sine''' and '''cosine''', or
'''sin''' and '''cos''' are basic [[trigonometirc functions]] (which are, in their turn, basic [[elementary functions]]), used to express relations between angles and sides of [[right-angled triangle]] (or [[right triangle]]). These functions are common in science and technology.
<!-- ==Definitions==  !-->
Various sources define trigonometric funcitons in different ways.
<!--
In this sense, there is no well-established definition of trigonometric functions.
Below, the most important approaches are listed.
!-->
==Geometric approach==
==Geometric approach==
In some countries, schoolchildren begin to study]] mathematics with arithmetics, elementary [[algebra]] and the [[Euclidean geometry]] ([[planimetry]]) <ref name="kiselev">
In most countries, schoolchildren begin to study mathematics with arithmetics, elementary [[algebra]] and 2-dimensional [[Euclidean geometry]] (plane geometry). <ref name="kiselev">
{{cite book
{{cite book
|author=A.P.Kiselev
|author=A. P. Kiselev
|translator=Alexander Givental
|translator=Alexander Givental
|ISBN=0-9779852-0-2
|ISBN=0-9779852-0-2
|isbn=0-9779852-0-2
|isbn=0-9779852-0-2
|year=1892 (first efition in Russia)
|year=1892 (first edition in Russian)
|url=http://math.berkeley.edu/~giventh/
|url=http://math.berkeley.edu/~giventh/
}}</ref>.
}}</ref>.
<!-- http://www.sumizdat.org/ !-->
When they learn about sin and cos, the students already have some idea about addition and measurement of [[angle]]s,  segments and areas; they already have accepted [[Euclid's elements|the axioms of Euclid]], in particular his famous fifth [[Euclid's parallel axiom|parallel axiom]], and they know the [[Pythagorean theorem]]. Therefore, the trigonometric functions are '''defined''' as ratios of the lengths of sides of  [[right-angled triangle]]s, for example, the ratio of the length of a [[leg(geometry)|leg]] to that of the [[hypothenuse]]. We will outline this approach, but point out that an alternative approach is possible which does not rely on Euclid's axioms. This approach, which is  more advanced in that it needs some knowledge of  differential calculus, will be given in the next section.
 
===Definition of sine and cosine in plane geometry===
{{Image|Def sine cosine.png|right|200px|Fig .1. Acute angle &alpha; is &ang;CAB; &ang;CBA = 90°.}}
The most elementary definition of the functions '''sine''' and '''cosine''' follows from Fig. 1. We see a right-angled triangle ABC (black) with an acute angle &alpha;. By definition sin&alpha; is equal to the the length ''a'' of the side opposite angle &alpha; divided by the [[hypotenuse]], which has length ''b''. The cosine of &alpha; is the length ''c'' of the  side adjacent to &alpha; divided by the length of the  hypotenuse ''b'':
:<math>
\sin\alpha\, \stackrel{\mathrm{def}}{=}\, \frac{a}{b}, \qquad \cos\alpha\, \stackrel{\mathrm{def}}{=}\, \frac{c}{b}.
</math>
Applying the same definitions to the complementary angle &ang;DAC (= 90°&minus;&alpha;) we find from the dashed blue triangle in Fig. 1,
:<math>
\sin(90^\circ-\alpha)  = \cos\alpha= \frac{c}{b},\qquad\cos(90^\circ-\alpha) = \sin\alpha= \frac{a}{b}.
</math>
We see that for complementary angles the functions sine and cosine are interchanged, hence the name "cosine" (short for "complementary sine").
 
The [[Pythagorean theorem]] states that ''b''<sup>2</sup> = ''a''<sup>2</sup>+''c''<sup>2</sup>, so that we get the following important relation between the sine and the cosine functions
:<math>
(\sin\alpha)^2 + (\cos\alpha)^2 = \frac{a^2}{b^2}+ \frac{c^2}{b^2}= \frac{a^2+c^2}{b^2} =  \frac{b^2}{b^2}=1.
</math>
 
Since this elementary definition of sine and cosine is in terms of ratios of lengths, which are positive, the definition needs to be generalized in order to account for  negative function values.
===Cartesian definition===
We recall that a vector '''r''' in  a 2-dimensional plane can be represented by two [[Cartesian coordinates]], which  are [[real number]]s: '''r''' = (''x'', ''y''). These numbers are the projections of '''r''' on the ''x''- and the ''y''-axis, respectively. The ''length'' of '''r''' is denoted by |'''r'''| and is defined by
:<math>
|\mathbf{r}| \, \stackrel{\mathrm{def}}{=}\,  \sqrt{x^2+y^2}.
</math>
Assuming that the length of '''r''' is non-zero we can divide the components of the vector by it and obtain a ''unit vector'' '''e'''<sub>''r''</sub>, which by definition has length unity. Unit vectors in the plane have their endpoints on a unit circle (circle with unit radius).
 
Consider Fig. 1 again. Assume now that the length ''b'' equals 1, then the vector '''e'''<sub>AC</sub> pointing from point A to point C is a unit vector. Assume a Cartesian system of axes with the origin in A and let the ''x''-axis be along A&ndash;B and the ''y''-axis be along A&ndash;D.  Clearly cos&alpha; =  ''c'' is the projection of '''e'''<sub>AC</sub> on the ''x''-axis, i.e., cos&alpha; = ''c''  is the ''x''-coordinate of this unit vector. Similarly sin&alpha; = ''a'' is the ''y'' coordinate of the unit vector. This observation leads to the following definition, which is a generalization of the earlier one.
 
Write  '''e'''<sub>''x''</sub> for a unit vector on the ''x''-axis and  '''e'''<sub>''y''</sub> for a unit vector on the ''y''-axis.
 
'''Definition''': If '''e''' is a unit vector obtained from '''e'''<sub>''x''</sub>  by a positive (counter-clockwise) rotation over the angle &alpha;, then cos&alpha; is the ''x'' coordinate of '''e''' and sin&alpha; is the ''y'' coordinate of '''e''':
:<math>
\mathbf{e}\, \stackrel{\mathrm{def}}{=}\,  \cos\alpha\,\mathbf{e}_x + \sin\alpha\,\mathbf{e}_y,  \qquad |\mathbf{e}| \equiv 1 = (\cos\alpha)^2 + (\sin(\alpha)^2.
</math>
 
 
The unit vector '''e''' is depicted (as a red vector with endpoint on the unit circle) in the four different quadrants in Fig. 2.
<br>
{{Image|Sign sincos.png|right|700px|Fig. 2. Unit vector '''e''' (shown in red) in different quadrants. Projection on ''x''-axis is cos &alpha;, projection on ''y''-axis is sin &alpha;.}}
 
<div align="left">
<table width="25%">
<tr><td colspan="4"> <hr> </td></tr>
<tr><td colspan="4" align="center">
<b>Signs of sine and cosine</b>
</tr>
<tr><td colspan="4"> <hr> </td></tr>
<tr><td>Quadrant      </td>
    <td>Range &alpha; </td>
    <td align="left">cos</td>
    <td align="left">sin</td>
</tr>
<tr><td colspan="4"> <hr> </td></tr>
<tr><td>I  <td>&nbsp;&nbsp;  0<sup>0</sup>< &alpha; <  90<sup>0</sup> <td>+<td>+</tr>
<tr><td>II  <td>&nbsp; 90<sup>0</sup>< &alpha; < 180<sup>0</sup> <td>&minus;<td>+</tr>
<tr><td>III <td>180<sup>0</sup>< &alpha; < 270<sup>0</sup> <td>&minus;<td>&minus;</tr>
<tr><td>IV  <td>270<sup>0</sup>< &alpha; < 360<sup>0</sup> <td>+<td>&minus;</tr>
</table>
</div>
 
<br><br><br><br><br>
Clearly, if &alpha;=0° then '''e''' = '''e'''<sub>''x''</sub>. If &alpha;=90° then '''e''' = '''e'''<sub>''y''</sub>. If &alpha;=180° then '''e''' = &minus;'''e'''<sub>''x''</sub>. Thus,
:<math>
\begin{align}
\cos  0^\circ  &=  1, & \sin  0^\circ  &= 0\\
\cos 90^\circ  &=  0, & \sin 90^\circ  &= 1 \\
\cos180^\circ  &= -1, & \sin180^\circ  &= 0
\end{align}
</math>
 
It is also clear from the definition that, when &alpha; is rotated over more than  360°, the unit vector is back at an angle between 0° and 360°. Therefore, for positive integer ''k'',
:<math>
\cos(\alpha + 360^\circ\sdot k) = \cos(\alpha),\qquad\sin(\alpha + 360^\circ\sdot k) = \sin(\alpha).
</math>
 
Negative angles are defined by rotating the unit vector ''clockwise''. By definition reflection in the ''x''-axis changes the components of the unit vector thus,
:<math>
x\mapsto x, \qquad y \mapsto -y
</math>
Further one sees easily  that &alpha; is mapped to &minus;&alpha; under this reflection. That is, upon reflection &alpha; and the ''y''-coordinate change sign and the ''x''-coordinate is invariant (stays the same), hence
:<math>
\sin(-\alpha) = -\sin(\alpha), \qquad \cos(-\alpha) = \cos(\alpha).
</math>
When one rotates the unit vector clockwise over more than 360° one arrives again at a position between 0° and 360°, so that  for positive integer ''k'',
:<math>
\cos(\alpha - 360^\circ\sdot k) = \cos(\alpha),\qquad\sin(\alpha - 360^\circ\sdot k) = \sin(\alpha).
</math>
 
By standard arguments of plane [[Euclidean geometry]] one can compute the numerical values of sine and cosine for a few specific angles. The application of differential calculus (see later in this article) allows the computation of the function values for ''all'' angles. Thus, one obtains the graphs shown in Fig. 3.
{{Image|Sine and cosine.png|center|600px|Fig. 3. The functions sin(''x'') and cos(''x'') for ''x'' running from &minus;360° to 360°}}


Beginning to deal with sin and cos, students already have some idea about "addition" and measurement of [[angle]]s,
===Sum formulas===
segments and areas; they already have accepted the [[axioms of planimetry]], in particular, the [[Euclid's parallel axiom]], and they know the [[Pythagorean theorem]].
The following equations hold:
Therefore, the trigonometric functions are '''defined''' as ratios of the lengths of sides of the [[right-angled triangle]]s, for example, ratio of length of a [[leg(geometry)}|leg]] to that of the [[hypothenuse]].
:<math>
\begin{align}
\cos(\alpha+\beta) &= \cos\alpha\cos\beta - \sin\alpha\sin\beta\\
\sin(\alpha+\beta) &= \sin\alpha\cos\beta+\cos\alpha\sin\beta
\end{align}
</math>


Geometric approach is described in many sources and, in particular, in
'''Proof'''
http://en.wikipedia.org/wiki/Trigonometric_functions|wikipedia.
{{Image|Sum angles.png|center|750px| Fig. 4A: Rotated (green) unit vectors <font color="green">
'''e'''<sub>''x'' &prime;</sub></font> and <font color="green">
'''e'''<sub>''y'' &prime;</sub></font> expressed in original (black) unit vectors '''e'''<sub>''x''</sub> and '''e'''<sub>''y''</sub>. <br>
Fig. 4B: Rotated (red) unit vector <font color="red">
'''e'''<sub>''x'' &quot;</sub></font> expressed with respect to the primed (green) vectors. <br>
Fig. 4C: The vector <font color="red">'''e'''<sub>''x'' &quot;</sub></font> expressed with respect to the unprimed (black) vectors}}


The geometrical approach rather assumes properties of sin and cos than deduce them. The assumptions are masked as the geometric axioms.  However the deduction of properties of trigonometric functions does not require axioms of geometry. Properties of sin and cos do not depend, for example, on the acceptance of rejection of the [[Euclid's parallel axiom]]. Funcitons sin and cos can be defined algebraically, then their properties can be deduced, not postulated.
The proof is easiest in matrix notation. Because we do not expect readers to know this notation we give the proof in long-hand notation and only indicate briefly the equivalent matrix form. The unit vectors and angles that play a role in the proof are depicted in Fig. 4.
Basic elements of such a deduction are suggested below.


==Differential equations==
As shown in Fig. 4A, the singly primed (green) unit vectors are given in terms of the unprimed (black) unit vectors by
The definitions of [[derivative]] and [[integral]] does not refer to trigonometric functions, and their basic properties (in particular the [[Cauchy-Kowalevski theorem]] of existence and uniqueness of solution of the [[Cauchy problem]]) can be deduced without to refer to sin, cos, or any axioms of the [[Euclidean Geometry]]. Therefore, the natural way of definition of functions sin and cos uses the differential equation.
:<math>
\begin{align}
\mathbf{e}_{x'} &= \cos\alpha\,\mathbf{e}_{x} + \sin\alpha\, \mathbf{e}_{y} \\
\mathbf{e}_{y'} &= -\sin\alpha\,\mathbf{e}_{x} + \cos\alpha\, \mathbf{e}_{y}
\end{align}
\quad \Longleftrightarrow\quad
\begin{pmatrix}\mathbf{e}_{x'}\\ \mathbf{e}_{y'}\end{pmatrix}=
\begin{pmatrix}\cos\alpha & \sin\alpha\\ -\sin\alpha& \cos\alpha \end{pmatrix}
\begin{pmatrix}\mathbf{e}_{x}\\ \mathbf{e}_{y}\end{pmatrix}
</math>
In Fig. 4B the doubly primed (red) unit vector '''e'''<sub>''x'' &quot;</sub> is given in terms of the singly primed (green) unit vectors  by
:<math>
\mathbf{e}_{x''} =  \cos\beta\,\mathbf{e}_{x'} + \sin\beta\, \mathbf{e}_{y'}
\quad \Longleftrightarrow\quad\mathbf{e}_{x''} =
(\cos\beta,\,\,\sin\beta) \begin{pmatrix}\mathbf{e}_{x'}\\ \mathbf{e}_{y'}\end{pmatrix}
</math>
In Fig. 4C  the doubly primed (red)  vector '''e'''<sub>''x'' &quot;</sub> is given in terms of the unprimed (black) vectors:
:<math>
\mathbf{e}_{x''} = \cos(\alpha+\beta)\,\mathbf{e}_{x} + \sin(\alpha+\beta)\, \mathbf{e}_{y}
\qquad\qquad\qquad\qquad \qquad\qquad\qquad\qquad\qquad (1)
</math>


'''Definition'''. Functions sin and cos are solution of system of [[equation]]s
Substitution of the expression for the singly primed unit vectors (green) in terms of the unprimed vectors (black) into the expression for '''e'''<sub>''x'' &quot;</sub>  gives :
:(1) <math> \sin'(t)= ~ \cos(t)</math>
:<math>
:(2) <math> \cos'(t)= - \sin(t)</math>
\mathbf{e}_{x''} = \cos\beta( \cos\alpha\,\mathbf{e}_{x}+ \sin\alpha\,\mathbf{e}_{y})
with condition
+\sin\beta(-\sin\alpha\,\mathbf{e}_{x} +\cos\alpha\,\mathbf{e}_{y})
:(3) <math> \sin(0)=0</math>
</math>
:(4) <math> \cos(0)=1</math>
Upon rearranging,
:<math>
\mathbf{e}_{x''} = (\cos\alpha\cos\beta - \sin\alpha\sin\beta)\,\mathbf{e}_{x} + (\sin\alpha\cos\beta+\cos\alpha\sin\beta)\,\mathbf{e}_{y}
\qquad\qquad\qquad (2)
</math>
(In matrix notation the last two expression are equivalent to the multiplication of a row vector with a square matrix). Putting Eqs. (1) and (2) for '''e'''<sub>''x'' &quot;</sub> in terms of the unprimed unit vectors  side by side, we find the equations to be proved. This conclusion is allowed because  '''e'''<sub>''x'' </sub> and '''e'''<sub>''y'' </sub> are linearly independent.


(end of definition).
==Definition by differential equations==
Usually, it is assumed, that independent variable <math>t</math> is real; in this case, values of functions sin and cos are also [[real number]]s. However, the definition above can be used for [[complex number]]s too.
One could say that the geometrical approach assumes the properties of sin and cos rather than deduce them. These assumptions are masked as the geometric axioms. However, as we will now show, the deduction of properties of trigonometric functions does not require axioms of geometry. Properties of sin and cos do not depend, for example, on the acceptance or rejection of the [[Euclid's parallel axiom]]. The functions sin and cos can be defined algebraically, and their properties can be deduced, not postulated. The basic elements of such a deduction are outlined below.


Such definition does not implies additional concepts about [[angle]] and sum of angles; and does not require [[Pythagorean theorem]] or even [[Parallel postulate]] of the Euclidean geometry. Such a definition does not refer to the concept of number [[Pi]] (<math>\pi</math>) and therefore can be used for its definition.
The definitions of [[derivative]] and [[integral]] do not depend on trigonometric functions, and the  basic properties of integrals and derivatives (in particular the [[Cauchy-Kowalevski theorem]] of existence and uniqueness of the solution of the [[Cauchy problem]]) can be deduced without reference to sin, cos, or any axioms of [[Euclidean geometry]]. Therefore, a natural and consistent way of defining the functions sin and cos is via differential equations.  


However, the naively-obvious properties of sin and cos should be deduced from the system (1)-(4).
We will indicate the first derivative of a function with respect to its variable by a prime:
:<math>
f'(t) \equiv \frac{df(t)}{dt}.
</math>
 
===Definition===
The functions sin and cos are solutions of the following system of [[equation]]s
:(1)&nbsp;&nbsp;&nbsp; <math> \sin'(t)= ~ \cos(t)\,</math>
:(2)&nbsp;&nbsp;&nbsp; <math> \cos'(t)= - \sin(t)\,</math>
with conditions
:(3)&nbsp;&nbsp;&nbsp; <math> \sin(0)=0\,</math>
:(4)&nbsp;&nbsp;&nbsp; <math> \cos(0)=1\,</math>
 
Usually, it is assumed that the independent variable ''t'' is real; in that case, the values of the functions sin and cos are also [[real number]]s. However, the definition above can be used for [[complex number]]s too.
 
The present definition does not imply additional concepts about [[angle]]s and sums of angles; and does not require the [[Pythagorean theorem]] or even the [[Euclid's elements|parallel axiom]] of  Euclidean geometry. Furthermore the definition does not refer to the concept of the number [[Pi]] (<math>\,\pi</math>) and therefore can be used for the definition of <math>\,\pi</math>. But,
one pays for this by having to deduce the properties of sin and cos from the system (1)&ndash;(4).


==Sum of squares==
==Sum of squares==
Consider function  
Consider the function  
<math>z(t)=\sin(t)^2+\cos(t)^2</math>. From equations (1) and (2) it follows that  
<math>\,z(t)=\sin(t)^2+\cos(t)^2</math>. From equations (1) and (2) it follows that  
<math>z'(t)=0</math>, i.e.,  
<math>\,z'(t)=0</math>, i.e.,  
<math>z(t)=z(0)=\rm constant </math>. Evaluation of this constant from equations (2) and (3) gives  
<math>\,z(t)=z(0)= </math>constant . Evaluation of this constant from equations (2) and (3) gives  
<math>z(t)=1</math> for all  
<math>\,z(t)=1</math> for all  
<math>t</math>, id est,
<math>\,t</math>, i.e.,
: (5) <math> \sin(t)^2+\cos(t)^2=1</math>
: (5)&nbsp;&nbsp;&nbsp; <math>\, \sin(t)^2+\cos(t)^2=1</math>


In this paper, the superscript as indicator of exponentiation is always written after to specify the argument, avoiding the confusions.
In this article, the superscript as indicator of exponentiation is always written last to specify the argument, avoiding the confusions.
<ref name="footnote">
<ref name="footnote">
Some authors implicitly define also functions with superscript,  
Some authors implicitly define also functions with superscript,  
<math>\sin^n(x)=\sin(x)^n</math> and <math>\sin^n(x)=\sin(x)^n</math>.
<math>\,\sin^n(x)=\sin(x)^n</math> and <math>\cos^n(x)=\cos(x)^n</math>.
However, such a notation leads to confusions as soon as one needs to consider the inverse function  
However, such a notation leads to confusions as soon as one needs to consider the inverse function  
(for a function <math>f</math>, it is common to use notation <math>f^{-1}</math> for inverse function, such that  <math>f\Big(f^{-1}(z)\Big)=z</math>)  
(for a function <math>\,f</math>, it is common to use notation <math>\,f^{-1}</math> for inverse function, such that  <math>\,f\Big(f^{-1}(z)\Big)=z</math>)  
or multiple combination of functions, for example:  
or multiple combination of functions, for example:  
<math>\sin^{-1}(x)={\rm arcsin}(x)</math>;  
<math>\sin^{-1}(x)={\rm arcsin}(x)</math>;  
Line 70: Line 202:
</ref>
</ref>


For the equation (5) it follows, that for all real values of <math>t</math> the functions  
For the equation (5) it follows, that for all real values of ''t'' the functions sin(''t'') and cos(''t'') are bounded:
<math>\sin(t)</math> and  
: (6)&nbsp;&nbsp;&nbsp; <math> -1\le \sin(t) \le 1 </math>
<math>\cos(t)</math> are bounded:
: (7)&nbsp;&nbsp;&nbsp; <math> -1\le \cos(t) \le 1 </math>
: (6) <math> -1\le \sin(t) \le 1 </math>
: (7) <math> -1\le \cos(t) \le 1 </math>
These properties allow efficient [[bracketing of functions]] sin and cos.
These properties allow efficient [[bracketing of functions]] sin and cos.


==Bracketing==
==Bracketing==
Consider <math> sin(x)</math> and <math>\cos(x)</math> in the case <math> x> 0</math>.  
Consider sin(''x'') and cos(''x'') in the case ''x'' > 0. From equation (1) it follows that  
From equation (1) it follows, that  
sin'(''x'') &le; 1 .  
<math> \sin'(x)\le 1 </math>.  
Integration of the last inequality with respect to <math>x </math> from 0 to <math>t</math> gives:
Integration of the last inequality with respect to <math>x </math> from 0 to <math>t</math> gives:
<math> \sin(t)-\sin(0) < t </math>. Using equation (3) we write  
<math> \sin(t)-\sin(0) < t </math>. Using equation (3) we write  
Line 104: Line 233:
::: <math> - x^2/2 < \cos(x)-\cos(0) < -x^2/2 + x^3/6 </math>
::: <math> - x^2/2 < \cos(x)-\cos(0) < -x^2/2 + x^3/6 </math>
: (12) <math> 1- x^2/2 < \cos(x)< 1 -x^2/2 + x^4/24 </math>
: (12) <math> 1- x^2/2 < \cos(x)< 1 -x^2/2 + x^4/24 </math>
Continuing this exercise, one can obtain more and more narrow bounds for values of sin and cos.  
Continuing this exercise, one can obtain more and more narrow bounds for values of sin and cos.
 
However, even equations (9,12) are sufficient to see that both both sin and cos remain positive while, the argument is within (0,1) range.
This property is used below to reveal periodicity of these functions.


==Mathematical induction==
==Mathematical induction==
Line 150: Line 282:


Change of variables  
Change of variables  
<math>t<math> to  
<math>t</math> to  
<math> -t</math> and replacement  
<math> -t</math> and replacement  
<math>\sin</math> to
<math>\sin</math> to
Line 181: Line 313:
The efficient algorithms for evaluation of <math>\pi</math> are mentioned in the spacial article about [[Pi]].
The efficient algorithms for evaluation of <math>\pi</math> are mentioned in the spacial article about [[Pi]].
Usually, the efficient algorithms require more efforts for the deduction.
Usually, the efficient algorithms require more efforts for the deduction.
==Notes==
==Notes==
<references/>
{{reflist}}
==See also==
*[[trigonometric functions]]
*[[Pi]]
*[[evaluation of Pi]]
*[[Linear differential equations]]
 
[[Category:Mathematics]]
[[Category:Elementary functions]]
[[Category:Trigonometry]]
[[Category:Mathematical theorems]]

Latest revision as of 09:53, 10 October 2013

This article is developing and not approved.
Main Article
Discussion
Related Articles  [?]
Bibliography  [?]
External Links  [?]
Citable Version  [?]
 
This editable Main Article is under development and subject to a disclaimer.

In trigonometry, the function sine and its complementary function cosine, abbreviated sin and cos, are central. They are used to express relations between angles and sides of triangles. The functions are very common and important in all branches of science and technology. It is possible to define sine and cosine in different ways. Two ways, one based on plane geometry and one based on differential equations, will be discussed below.

Geometric approach

In most countries, schoolchildren begin to study mathematics with arithmetics, elementary algebra and 2-dimensional Euclidean geometry (plane geometry). [1]. When they learn about sin and cos, the students already have some idea about addition and measurement of angles, segments and areas; they already have accepted the axioms of Euclid, in particular his famous fifth parallel axiom, and they know the Pythagorean theorem. Therefore, the trigonometric functions are defined as ratios of the lengths of sides of right-angled triangles, for example, the ratio of the length of a leg to that of the hypothenuse. We will outline this approach, but point out that an alternative approach is possible which does not rely on Euclid's axioms. This approach, which is more advanced in that it needs some knowledge of differential calculus, will be given in the next section.

Definition of sine and cosine in plane geometry

CC Image

The most elementary definition of the functions sine and cosine follows from Fig. 1. We see a right-angled triangle ABC (black) with an acute angle α. By definition sinα is equal to the the length a of the side opposite angle α divided by the hypotenuse, which has length b. The cosine of α is the length c of the side adjacent to α divided by the length of the hypotenuse b:

Applying the same definitions to the complementary angle ∠DAC (= 90°−α) we find from the dashed blue triangle in Fig. 1,

We see that for complementary angles the functions sine and cosine are interchanged, hence the name "cosine" (short for "complementary sine").

The Pythagorean theorem states that b2 = a2+c2, so that we get the following important relation between the sine and the cosine functions

Since this elementary definition of sine and cosine is in terms of ratios of lengths, which are positive, the definition needs to be generalized in order to account for negative function values.

Cartesian definition

We recall that a vector r in a 2-dimensional plane can be represented by two Cartesian coordinates, which are real numbers: r = (x, y). These numbers are the projections of r on the x- and the y-axis, respectively. The length of r is denoted by |r| and is defined by

Assuming that the length of r is non-zero we can divide the components of the vector by it and obtain a unit vector er, which by definition has length unity. Unit vectors in the plane have their endpoints on a unit circle (circle with unit radius).

Consider Fig. 1 again. Assume now that the length b equals 1, then the vector eAC pointing from point A to point C is a unit vector. Assume a Cartesian system of axes with the origin in A and let the x-axis be along A–B and the y-axis be along A–D. Clearly cosα = c is the projection of eAC on the x-axis, i.e., cosα = c is the x-coordinate of this unit vector. Similarly sinα = a is the y coordinate of the unit vector. This observation leads to the following definition, which is a generalization of the earlier one.

Write ex for a unit vector on the x-axis and ey for a unit vector on the y-axis.

Definition: If e is a unit vector obtained from ex by a positive (counter-clockwise) rotation over the angle α, then cosα is the x coordinate of e and sinα is the y coordinate of e:


The unit vector e is depicted (as a red vector with endpoint on the unit circle) in the four different quadrants in Fig. 2.

CC Image
Fig. 2. Unit vector e (shown in red) in different quadrants. Projection on x-axis is cos α, projection on y-axis is sin α.

Signs of sine and cosine


Quadrant Range α cos sin

I    00< α < 900 ++
II   900< α < 1800 +
III 1800< α < 2700
IV 2700< α < 3600 +






Clearly, if α=0° then e = ex. If α=90° then e = ey. If α=180° then e = −ex. Thus,

It is also clear from the definition that, when α is rotated over more than 360°, the unit vector is back at an angle between 0° and 360°. Therefore, for positive integer k,

Negative angles are defined by rotating the unit vector clockwise. By definition reflection in the x-axis changes the components of the unit vector thus,

Further one sees easily that α is mapped to −α under this reflection. That is, upon reflection α and the y-coordinate change sign and the x-coordinate is invariant (stays the same), hence

When one rotates the unit vector clockwise over more than 360° one arrives again at a position between 0° and 360°, so that for positive integer k,

By standard arguments of plane Euclidean geometry one can compute the numerical values of sine and cosine for a few specific angles. The application of differential calculus (see later in this article) allows the computation of the function values for all angles. Thus, one obtains the graphs shown in Fig. 3.

CC Image
Fig. 3. The functions sin(x) and cos(x) for x running from −360° to 360°

Sum formulas

The following equations hold:

Proof

CC Image

The proof is easiest in matrix notation. Because we do not expect readers to know this notation we give the proof in long-hand notation and only indicate briefly the equivalent matrix form. The unit vectors and angles that play a role in the proof are depicted in Fig. 4.

As shown in Fig. 4A, the singly primed (green) unit vectors are given in terms of the unprimed (black) unit vectors by

In Fig. 4B the doubly primed (red) unit vector ex " is given in terms of the singly primed (green) unit vectors by

In Fig. 4C the doubly primed (red) vector ex " is given in terms of the unprimed (black) vectors:

Substitution of the expression for the singly primed unit vectors (green) in terms of the unprimed vectors (black) into the expression for ex " gives :

Upon rearranging,

(In matrix notation the last two expression are equivalent to the multiplication of a row vector with a square matrix). Putting Eqs. (1) and (2) for ex " in terms of the unprimed unit vectors side by side, we find the equations to be proved. This conclusion is allowed because ex and ey are linearly independent.

Definition by differential equations

One could say that the geometrical approach assumes the properties of sin and cos rather than deduce them. These assumptions are masked as the geometric axioms. However, as we will now show, the deduction of properties of trigonometric functions does not require axioms of geometry. Properties of sin and cos do not depend, for example, on the acceptance or rejection of the Euclid's parallel axiom. The functions sin and cos can be defined algebraically, and their properties can be deduced, not postulated. The basic elements of such a deduction are outlined below.

The definitions of derivative and integral do not depend on trigonometric functions, and the basic properties of integrals and derivatives (in particular the Cauchy-Kowalevski theorem of existence and uniqueness of the solution of the Cauchy problem) can be deduced without reference to sin, cos, or any axioms of Euclidean geometry. Therefore, a natural and consistent way of defining the functions sin and cos is via differential equations.

We will indicate the first derivative of a function with respect to its variable by a prime:

Definition

The functions sin and cos are solutions of the following system of equations

(1)   
(2)   

with conditions

(3)   
(4)   

Usually, it is assumed that the independent variable t is real; in that case, the values of the functions sin and cos are also real numbers. However, the definition above can be used for complex numbers too.

The present definition does not imply additional concepts about angles and sums of angles; and does not require the Pythagorean theorem or even the parallel axiom of Euclidean geometry. Furthermore the definition does not refer to the concept of the number Pi () and therefore can be used for the definition of . But, one pays for this by having to deduce the properties of sin and cos from the system (1)–(4).

Sum of squares

Consider the function . From equations (1) and (2) it follows that , i.e., constant . Evaluation of this constant from equations (2) and (3) gives for all , i.e.,

(5)   

In this article, the superscript as indicator of exponentiation is always written last to specify the argument, avoiding the confusions. [2]

For the equation (5) it follows, that for all real values of t the functions sin(t) and cos(t) are bounded:

(6)   
(7)   

These properties allow efficient bracketing of functions sin and cos.

Bracketing

Consider sin(x) and cos(x) in the case x > 0. From equation (1) it follows that sin'(x) ≤ 1 . Integration of the last inequality with respect to from 0 to gives: . Using equation (3) we write

(8) .

Rewrite (8) as and apply equation(2):

Integrating this inequality with respect to from 0 to gives:

Using equation(4) gives

(9)

The last "<" just follows from equation (7). Using equation (1) gives

(10)

Integrating equation (12) with respect to from 0 to gives

(11)

Rewrite it as

and apply equation (2):

Integrating this equation with respect to from 0 to gives

(12)

Continuing this exercise, one can obtain more and more narrow bounds for values of sin and cos.

However, even equations (9,12) are sufficient to see that both both sin and cos remain positive while, the argument is within (0,1) range. This property is used below to reveal periodicity of these functions.

Mathematical induction

Applying the mathematical induction, it is possible to show that for positive integer

(13)
(14)

and equality takes place only at .

Is cos always larger than sin?

Consider the range such that both and are positive; let be exact upper bound of such interval. Within this interval, sin is monotonously increasing, and cos is monotonously decreasing. Therefore, there exist one and only one solution of the equation

(13)

Let us estimate value of this solution.

Consider the special case, substitute into equation(13) and into equation(11). This gives

In such a way,

(14)

hence,

(15) .

Substitute into equation (11); this gives

(16)
(17)

From equation (5) it follows that

(18)
(19)

Both sin and cos are positive in the range (0,1); and, therefore,

(20)

Therefore,

(21) .

Symmetry and sense of number

Change of variables to and replacement to keeps the system (1)-(4) invariant. This means that

(22)
(23)

Consider replacement to , cos to sin, and sin to cos. Due to equation (13), such a replacement preserves the equations (1-4); therefore,

(24)
(25)

Then, , and

(26)
(27)

It follows, that functions sin and cos are periodic; and

(30)

is period. The deduction above can be used as definition of number . In such a way, appears as half of period of functions which are solutions of equations (1,2).

For definition of , equations (1,2) are sufficient: due to the linearity of equations and the translational invariance, conditions (3,4) do not affect the period; all the solutions with different conditions at zero have the same period.

From estimate (21) it follows that

(31)

The improvement of this estimate using the expansions (13), (14) is possible but computationally non-efficient. The efficient algorithms for evaluation of are mentioned in the spacial article about Pi. Usually, the efficient algorithms require more efforts for the deduction.

Notes

  1. A. P. Kiselev (1892 (first edition in Russian)). {{{title}}}. ISBN 0-9779852-0-2. 
  2. Some authors implicitly define also functions with superscript, and . However, such a notation leads to confusions as soon as one needs to consider the inverse function (for a function , it is common to use notation for inverse function, such that ) or multiple combination of functions, for example: ; .