映画にもなった小説「博士の愛した数式」(小川洋子著新潮文庫)に出てくるオイラーの公式を高校の数学を用いて導いてみよう。
自然対数の底
ネイピア数とも呼ばれる自然対数の底[math]e[/math]は、昔は高校の数Ⅲの教科書の最後に載っていたが、次の(1)式で定義される。
[math]e=\lim _{ n\rightarrow \infty }{ { \left( 1+\frac { 1 }{ n } \right) }^{ n } }[/math] (1)
[math]n[/math]を正の整数として、[math]{ e }_{ n }={ \left( 1+{ 1 }/{ n } \right) }^{ n }[/math]とすると、[math]{e}_{1}[/math]=2, [math]{e}_{2}[/math] =2.25, [math]{e}_{3}[/math]=2.370370370・・・、 [math]{e}_{4}[/math]=2.44140625・・・と続き、その極限値が[math]e[/math]であるが、エクセルを用いると簡単に求めれることができ、[math]e[/math]=2.71828・・・・・・・と小数点以下が無限に続く無理数となる。
複利計算と指数関数
元金[math]A[/math]円を年利率[math]a[/math]の定期預金に自動継続で預けたとき、[math]x[/math]年後の元利合計を[math]y[/math]とすると、年利率は同じでも、利息が元金に1年ごとに繰り入れられる預金よりも、半年ごとに繰り入れられる預金のほうが預金者にとって有利になる。
①自動継続で1年ごとに利息が元金に繰り入れられる定期貯金では1年間の利率が[math]a[/math]であるから、正の整数[math]x[/math]に対して、[math]x[/math]年後の元利合計[math]{ y }_{ 1 }[/math]は次のようになる。
[math]{y}_{1}[/math]=[math]A{ \left( 1+a \right) }^{ x }[/math] (2)
②一方、自動継続で、半年ごとに利息が元金に繰り入れられる定期預金の場合、半年間の利率は[math]a/2[/math]であり、[math]x[/math]年の間に[math]2x[/math]回繰り入れられるので、1/2の倍数[math]x[/math]に対して、その場合の元利合計[math]{ y }_{ 2 }[/math]は次のようになる。
[math]{y}_{2}[/math]=[math]A{ \left( 1+a/2 \right) }^{ 2x }[/math] (3)
(3)式の右辺は[math]{ A\left( 1+a+{ { a }^{ 2 } }/{ 4 } \right) }^{ x }[/math]となるので、[math]{ y }_{2 }[/math]は[math]{ y }_{ 1 }[/math]より大きくなる。
③一般に、自動継続で[math]1/n[/math]年ごとに、利息が元金に繰り入れられる定期預金の場合、その期間の利率は[math]a/n[/math]であり、[math]x[/math]年の間に[math]nx[/math]回、利息が繰り込まれる定期預金での[math]x[/math]年後の元利合計は、1/[math]n[/math]の倍数[math]x[/math]に対して次のようになる。
[math]{y}_{n}[/math]=[math]A{ \left( 1+a/n \right) }^{ nx }[/math] (4)
ここまでは簡単であろう。ここで、改めて、[math]n/a[/math]→[math]n[/math]と置き換えれば、(4)式の右辺は[math]A{ \left( 1+1/n \right) }^{ nax }[/math]となり、さらに[math]n[/math]→∞の極限では(1)式から、[math]A{ e }^{ ax }[/math]となる。すなわち、
[math]y[/math]=[math]A{ e }^{ ax }[/math] (5)
となる。つまり、
[math]{ e }^{ ax }[/math]=[math]\lim _{ n\rightarrow \infty }{ { \left( 1+\frac { a }{ n } \right) }^{ nx } }[/math] (6)
となる。
指数関数の微分
一般に[math]x[/math]の関数[math]f\left( x \right)[/math]の微分は次のように表される。
[math]\frac { df\left( x \right) }{ dx } =\lim _{ h\rightarrow 0 }{ \frac { f\left( x+h \right) -f\left( x \right) }{ h } }[/math] (7)
尚、[math]f\left( x \right)[/math]の微分を[math]{ { \left( f\left( x \right) \right) }^{ \prime } }[/math]または[math]{ f }^{ \prime }\left( x \right)[/math]とも表す。
指数関数[math]{ e }^{ ax }[/math]の微分を求めるのに、[math]h=1/n[/math]とすると、[math]h[/math]→0とするのは[math]n[/math]→∞とするのと同じだから、(6)式を用いると、
[math]{ \left( { e }^{ ax } \right) }^{ \prime }[/math]=[math]\lim _{ n\rightarrow \infty }{ \frac { { \left( 1+a/n \right) }^{ n\left( x+1/n \right) }-{ \left( 1+a/n \right) }^{ nx } }{ 1/n } }[/math]=[math]a{ e }^{ ax }[/math]
となる。すなわち、指数関数[math]{ e }^{ ax }[/math]を[math]x[/math]で微分することは、もとの関数に[math]a[/math]をかけることと同じである。[math]a[/math]=1のとき、つまり、指数関数[math]{ e }^{ x }[/math]は微分しても変化しない。
関数の級数展開
関数 [math]{ f }\left( x \right)[/math]が、つぎのように、[math]x[/math]のべき級数に展開できるとしよう。
[math]f\left( x \right) ={ a }_{ 0 }+{ a }_{ 1 }x+{ a }_{ 2 }{ x }^{ 2 }+\cdots +{ a }_{ n }{ x }^{ n }+\cdots[/math] (8)
次に、(8)式右辺の各項の係数を求める。(8)式において、[math]x[/math]=0と置くことにより、[math]{a}_{0}[/math]=[math]{ f }\left( 0 \right)[/math]であり、(8)式の両辺を1回微分したのち、[math]x[/math]=0と置くことにより、[math]{a}_{1}[/math]=[math]{ f }^{ \prime }\left( 0 \right)[/math]となり、8)式の両辺を2回微分したのち、[math]x[/math]=0と置くことにより、[math]{a}_{2}[/math]=[math]{ f }^{ \prime\prime }\left( 0 \right)[/math]/2!となる。一般に(8)式を[math]n[/math]回微分したのち、[math]x[/math]=0と置くことにより、[math]{a}_{n}[/math]=[math]\frac { f_{ }^{ \left( n \right) }{ \left( 0 \right) } }{ n! }[/math]となる。よつて(8)式は、
[math]f\left( x \right) =f\left( 0 \right) +f^{ \prime }\left( 0 \right) x+\frac { f^{ \prime \prime }\left( 0 \right) }{ 2! } { x }^{ 2 }+\cdots +\frac { f^{ \left( n \right) }\left( 0 \right) }{ n! } { x }^{ n }+\cdots [/math] (9)
① [math]f\left( x \right) ={ e }^{ x }[/math]の場合:[math]f\left( 0 \right) =f^{ \prime }\left( 0 \right) =f^{ \prime \prime }\left( 0 \right) =\cdots =f^{ \left( n \right) }\left( 0 \right) =\cdots =1[/math]である。これを(9)式に代入すると、次のように級数展開が得られる。
[math]{ e }^{ x }=1+x+\frac { { x }^{ 2 } }{ 2! } +\frac { { x }^{ 3 } }{ 3! } +\frac { { x }^{ 4 } }{ 4! } +\frac { { x }^{ 5 } }{ 5! } +\cdots[/math] (10)
② 三角関数の級数展開: まず、[math]f\left( x \right) =\sin { x }[/math]の場合、[math]f\left( 0 \right) =0[/math]、[math]f^{ \prime }\left( 0 \right) =1[/math]、[math]f^{ \prime \prime }\left( 0 \right) =0[/math]、[math]f^{ \prime \prime\prime }\left( 0 \right) [/math]=-1であり、後は0、1、0、-1の繰り返しだから、奇数次の項のみが残り、
[math] \sin { x=x-\frac { { x }^{ 3 } }{ 3! } +\frac { { x }^{ 5 } }{ 5! } } -\frac { { x }^{ 7 } }{ 7! } +-\cdots[/math] (11)
となる。同様に[math]f\left( x \right) =\cos { x }[/math]の場合、偶数次の項のみが残り次のように展開できる。
[math] \cos { x=1-\frac { { x }^{ 2 } }{ 2! } +\frac { { x }^{4 } }{ 4! } } -\frac { { x }^{ 6 } }{ 6! } +-\cdots[/math] (12)
当然(11)を微分すると(12)になる。
複素空間への拡張
指数関数[math]{ e }^{ x }[/math]は実数[math]x[/math]に対して定義されているが、これを一般に複素数[math]x[/math]に拡張することができる。(10)式に、θを実数、[math]i[/math]を虚数単位として、[math]x=i\theta[/math]とすると、[math]{ i }^{ 2 }=-1[/math]、[math]{ i }^{ 3 }=-i[/math]、[math]{ i }^{ 4 }=1[/math]、[math]{ i }^{ 5 }=i[/math]であるから、
[math]{ e }^{ i\theta }=1-\frac { { \theta }^{ 2 } }{ 2! } +\frac { { \theta }^{ 4 } }{ 4! } -+\cdots +i\left( \theta -\frac { { \theta }^{ 3 } }{ 3! } +\frac { { \theta }^{ 5 } }{ 5! } -+\cdots \right)[/math] (13)
となる。右辺実数部は(12)式から、[math]\cos { \theta }[/math]であり、虚数部は(11)式から[math]\sin { \theta }[/math]である。よつて、次のオイラーの公式が導かれる。
[math]{ e }^{ i\theta }=\cos { \theta +i\sin { \theta } }[/math] (14)
θの値が円周率πに等しいとき、cosπ=-1、sinπ=0であるから、オイラーの公式から次の関係が成り立つ。
[math]{ e }^{ i\pi }+1=0[/math] (15)
(15)式には、自然対数の底[math]e[/math]、円周率π、虚数単位[math]i[/math]、そして実数1と、性格の全く異なる数が一つの式のなかに仲良く同居している。さらに(14)式の両辺を[math]n[/math]乗すると、次のド・モアブルの定理が導かれる。
[math]{ \left( \cos { \theta +i\sin { \theta } } \right) }^{ n }=\cos { n\theta +i\sin { n\theta } }[/math] (16)
(16)式において、[math]n[/math]=2として、両辺の実数部どうし、および虚数部どうしを等しいとおくと、高校の三角関数で習う倍角の公式となる。また、[math]n[/math]=3から3倍角の公式が導かれる。
オイラーの公式(14)から導かれる(15)式は、「博士が愛した数式」では、世界一美しい数式として紹介されているが、チコちゃんが世界一美しいと思う数式はどんな式かな? エッ、結婚式! 博士の愛した数式も実数と虚数の結婚式のようなものだね。

コメント