自然对数(英语:Natural logarithm)为以数学常数e为底数的对数函数,标记作
ln
x
{\displaystyle \ln x}
或
log
e
x
{\displaystyle \log _{e}x}
,其反函数为指数函数
e
x
{\displaystyle e^{x}}
。[注 1]
自然对数
ln
(
x
)
{\displaystyle \ln(x)}
的函数图像
自然对数
ln
(
x
)
{\displaystyle \ln(x)}
的积分定义
自然对数积分定义为对任何正实数
x
{\displaystyle x}
,由
1
{\displaystyle 1}
到
x
{\displaystyle x}
所围成,
x
y
=
1
{\displaystyle xy=1}
曲线下的面积。如果
x
{\displaystyle x}
小于1,则计算面积为负数。
ln
x
=
∫
1
x
d
t
t
{\displaystyle \ln x=\int _{1}^{x}{\frac {dt}{t}}\,}
e
{\displaystyle e}
则定义为唯一的实数
x
{\displaystyle x}
使得
ln
x
=
1
{\displaystyle \ln x=1}
。
自然对数一般表示为
ln
x
{\displaystyle \ln x\!}
,数学中亦有以
log
x
{\displaystyle \log x\!}
表示自然对数。[1][注 2]
目录
1 历史
1.1 十七世纪
1.2 十八世纪
2 形式定义
3 性质
4 导数
5 幂级数
6 积分
6.1 例子
7 与双曲函数的关系
8 连分数
9 复数对数
9.1 主值定义
10 科学应用
11 注释
12 参考资料
13 延伸阅读
历史
编辑
十七世纪
编辑
双曲线扇形是笛卡尔平面
{
(
x
,
y
)
}
{\displaystyle \{(x,y)\}}
上的一个区域,由从原点到
(
a
,
1
a
)
{\textstyle (a,{\frac {1}{a}})}
和
(
b
,
1
b
)
{\textstyle (b,{\frac {1}{b}})}
的射线,以及双曲线
x
y
=
1
{\displaystyle xy=1}
围成。在标准位置的双曲线扇形有
a
=
1
{\displaystyle a=1}
且
b
>
1
{\displaystyle b>1}
,它的面积为
ln
(
b
)
{\displaystyle \ln(b)}
[2],此时双曲线扇形对应正双曲角。
当直角双曲线下的两段面积相等时,
x
{\displaystyle x}
的值呈等比数列,
x
2
x
1
=
x
1
x
0
=
k
{\textstyle {\frac {x_{2}}{x_{1}}}={\frac {x_{1}}{x_{0}}}=k}
,
y
{\displaystyle y}
的值也呈等比数列,
x
2
x
1
=
x
1
x
0
=
1
k
{\textstyle {\frac {x_{2}}{x_{1}}}={\frac {x_{1}}{x_{0}}}={\frac {1}{k}}}
。
约翰·纳皮尔在1614年[3]以及约斯特·比尔吉在6年后[4],分别发表了独立编制的对数表,当时通过对接近1的底数的大量乘幂运算,来找到指定范围和精度的对数和所对应的真数。当时还没出现有理数幂的概念,按后世的观点,约翰·纳皮尔的底数0.999999910000000相当接近
1
e
{\textstyle {\frac {1}{e}}}
[5],而约斯特·比尔吉的底数1.000110000相当接近自然对数的底数
e
{\displaystyle e}
。实际上不需要做开高次方这种艰难运算,约翰·纳皮尔用了20年时间进行相当于数百万次乘法的计算,亨利·布里格斯(英语:Henry Briggs (mathematician))建议纳皮尔改用10为底数未果,他用自己的方法[6]于1624年部分完成了常用对数表的编制。
形如
f
(
x
)
=
x
p
{\displaystyle f(x)=x^{p}}
的曲线都有一个代数反导数,除了特殊情况
p
=
−
1
{\displaystyle p=-1}
对应于双曲线的弓形面积(英语:Quadrature (mathematics)),即双曲线扇形;其他情况都由1635年发表的卡瓦列里弓形面积公式(英语:Cavalieri's quadrature formula)给出[7],其中抛物线的弓形面积由公元前3世纪的阿基米德完成(抛物线的弓形面积),双曲线的弓形面积需要发明一个新函数。1647年圣文森特的格列高利(英语:Grégoire de Saint-Vincent)将对数联系于双曲线
x
y
=
1
{\displaystyle xy=1}
的弓形面积,他发现x轴上
[
a
,
b
]
{\displaystyle [a,b]}
两点对应的双曲线线段与原点围成的双曲线扇形同
[
c
,
d
]
{\displaystyle [c,d]}
对应的扇形,在
a
b
=
c
d
{\textstyle {\frac {a}{b}}={\frac {c}{d}}}
时面积相同,这指出了双曲线从
x
=
1
{\displaystyle x=1}
到
x
=
t
{\displaystyle x=t}
的积分
f
(
t
)
{\displaystyle f(t)}
满足[8]:
f
(
t
u
)
=
f
(
t
)
+
f
(
u
)
{\displaystyle f(tu)=f(t)+f(u)\,}
1649年,萨拉萨的阿尔丰斯·安东尼奥(英语:Alphonse Antonio de Sarasa)将双曲线下的面积解释为对数。大约1665年,伊萨克·牛顿推广了二项式定理,他将
1
1
+
x
{\textstyle {\frac {1}{1+x}}}
展开并逐项积分,得到了自然对数的无穷级数。“自然对数”最早描述见于尼古拉斯·麦卡托在1668年出版的著作《Logarithmotechnia》中[9],他也独立发现了同样的级数,即自然对数的麦卡托级数。
十八世纪
编辑
大约1730年,欧拉定义互为逆函数的指数函数和自然对数为[10][11]:
e
x
=
lim
n
→
∞
(
1
+
x
n
)
n
,
{\displaystyle e^{x}=\lim _{n\rightarrow \infty }\left(1+{\frac {x}{n}}\right)^{n},}
ln
(
x
)
=
lim
n
→
∞
n
(
x
1
n
−
1
)
{\displaystyle \ln(x)=\lim _{n\rightarrow \infty }n\left(x^{\frac {1}{n}}-1\right)}
1742年威廉·琼斯发表了现在的幂指数概念[12]。
形式定义
编辑
欧拉定义自然对数为序列的极限:
ln
(
x
)
=
lim
n
→
∞
n
(
x
1
n
−
1
)
.
{\displaystyle \ln(x)=\lim _{n\rightarrow \infty }n\left(x^{\frac {1}{n}}-1\right).}
ln
(
a
)
{\displaystyle \ln(a)}
正式定义为积分,
ln
(
a
)
=
∫
1
a
1
x
d
x
.
{\displaystyle \ln(a)=\int _{1}^{a}{\frac {1}{x}}\,dx.}
这个函数为对数是因满足对数的基本性质:
ln
(
a
b
)
=
ln
(
a
)
+
ln
(
b
)
.
{\displaystyle \ln(ab)=\ln(a)+\ln(b).\,\!}
这可以通过将定义了
ln
(
a
b
)
{\displaystyle \ln(ab)}
的积分拆分为两部分,并在第二部分中进行换元
x
=
t
a
{\displaystyle x=ta}
来证实:
ln
(
a
b
)
=
∫
1
a
b
1
x
d
x
=
∫
1
a
1
x
d
x
+
∫
a
a
b
1
x
d
x
=
∫
1
a
1
x
d
x
+
∫
1
b
1
a
t
d
(
a
t
)
{\displaystyle \ln(ab)=\int _{1}^{ab}{\frac {1}{x}}\;dx=\int _{1}^{a}{\frac {1}{x}}\;dx\;+\int _{a}^{ab}{\frac {1}{x}}\;dx=\int _{1}^{a}{\frac {1}{x}}\;dx\;+\int _{1}^{b}{\frac {1}{at}}\;d(at)}
=
∫
1
a
1
x
d
x
+
∫
1
b
1
t
d
t
=
ln
(
a
)
+
ln
(
b
)
.
{\displaystyle =\int _{1}^{a}{\frac {1}{x}}\;dx\;+\int _{1}^{b}{\frac {1}{t}}\;dt=\ln(a)+\ln(b).}
幂公式
ln
(
t
r
)
=
r
ln
(
t
)
{\displaystyle \ln(t^{r})=r\ln(t)}
可如下推出:
ln
(
t
r
)
=
∫
1
t
r
1
x
d
x
=
∫
1
t
1
u
r
d
(
u
r
)
=
∫
1
t
1
u
r
(
r
u
r
−
1
d
u
)
=
r
∫
1
t
1
u
d
u
=
r
ln
(
t
)
.
{\displaystyle \ln(t^{r})=\int _{1}^{t^{r}}{\frac {1}{x}}dx=\int _{1}^{t}{\frac {1}{u^{r}}}d\left(u^{r}\right)=\int _{1}^{t}{\frac {1}{u^{r}}}\left(ru^{r-1}\,du\right)=r\int _{1}^{t}{\frac {1}{u}}\,du=r\ln(t).}
第二个等式使用了换元
u
=
x
1
r
{\displaystyle u=x^{\frac {1}{r}}}
。
自然对数还有在某些情况下更有用的另一个积分表示:
ln
(
x
)
=
−
lim
ϵ
→
0
∫
ϵ
∞
d
t
t
(
e
−
x
t
−
e
−
t
)
.
{\displaystyle \ln(x)=-\lim _{\epsilon \to 0}\int _{\epsilon }^{\infty }{\frac {dt}{t}}\left(e^{-xt}-e^{-t}\right).}
性质
编辑
ln
(
1
)
=
∫
1
1
1
t
d
t
=
0
{\displaystyle \ln(1)=\int _{1}^{1}{\frac {1}{t}}\,dt=0\,}
ln
(
−
1
)
=
i
π
{\displaystyle \operatorname {ln} (-1)=i\pi \,}
(参见复数对数)
ln
(
x
)
<
ln
(
y
)
f
o
r
0
<
x
<
y
{\displaystyle \ln(x)<\ln(y)\quad {\rm {for}}\quad 0 lim x → 0 ln ( 1 + x ) x = 1 {\displaystyle \lim _{x\to 0}{\frac {\ln(1+x)}{x}}=1\,} ln ( x y ) = y ln ( x ) {\displaystyle \ln(x^{y})=y\,\ln(x)\,} x − 1 x ≤ ln ( x ) ≤ x − 1 f o r x > 0 {\displaystyle {\frac {x-1}{x}}\leq \ln(x)\leq x-1\quad {\rm {for}}\quad x>0\,} ln ( 1 + x α ) ≤ α x f o r x ≥ 0 , α ≥ 1 {\displaystyle \ln {(1+x^{\alpha })}\leq \alpha x\quad {\rm {for}}\quad x\geq 0,\alpha \geq 1\,} 证明 lim h → 0 ln ( 1 + h ) h = lim h → 0 ln ( 1 + h ) − ln 1 h = d d x ln x | x = 1 = 1 {\displaystyle \lim _{h\to 0}{\frac {\ln(1+h)}{h}}=\lim _{h\to 0}{\frac {\ln(1+h)-\ln 1}{h}}={\frac {d}{dx}}\ln x{\Bigg |}_{x=1}=1} 导数 编辑 自然对数的图像和它在 x = 1.5 {\displaystyle x=1.5} 处的切线。 ln ( 1 + x ) {\displaystyle \ln(1+x)} 的泰勒多项式只在 − 1 < x ≤ 1 {\displaystyle -1 范围内有逐步精确的近似。 自然对数的导数为 d d x ln ( x ) = 1 x . {\displaystyle {\frac {d}{dx}}\ln(x)={\frac {1}{x}}.\,} 证明一(微积分第一基本定理): d d x ln ( x ) = d d x ∫ 1 x 1 t d t = 1 x {\displaystyle {\frac {d}{dx}}\ln(x)={\frac {d}{dx}}\int _{1}^{x}{\frac {1}{t}}\,dt={\frac {1}{x}}} 证明二:按此影片(页面存档备份,存于互联网档案馆) d d x ln ( x ) = lim h → 0 ln ( x + h ) − ln ( x ) h {\displaystyle {\frac {d}{dx}}\ln(x)=\lim _{h\to 0}{\frac {\ln(x+h)-\ln(x)}{h}}} = lim h → 0 ln ( x + h x ) h {\displaystyle =\lim _{h\to 0}{\frac {\ln({\frac {x+h}{x}})}{h}}} = lim h → 0 [ 1 h ln ( 1 + h x ) ] {\displaystyle =\lim _{h\to 0}\left[{\frac {1}{h}}\ln \left(1+{\frac {h}{x}}\right)\right]\quad } = lim h → 0 ln ( 1 + h x ) 1 h {\displaystyle =\lim _{h\to 0}\ln \left(1+{\frac {h}{x}}\right)^{\frac {1}{h}}} 设 u = h x ⇒ u x = h {\displaystyle u={\frac {h}{x}}\Rightarrow ux=h} 1 h = 1 u x {\displaystyle {\frac {1}{h}}={\frac {1}{ux}}} d d x ln ( x ) = lim u → 0 ln ( 1 + u ) 1 u x {\displaystyle {\frac {d}{dx}}\ln(x)=\lim _{u\to 0}\ln(1+u)^{\frac {1}{ux}}} = lim u → 0 ln [ ( 1 + u ) 1 u ] 1 x {\displaystyle =\lim _{u\to 0}\ln \left[(1+u)^{\frac {1}{u}}\right]^{\frac {1}{x}}} = 1 x lim u → 0 ln ( 1 + u ) 1 u {\displaystyle ={\frac {1}{x}}\lim _{u\to 0}\ln(1+u)^{\frac {1}{u}}} 设 n = 1 u ⇒ u = 1 n {\displaystyle n={\frac {1}{u}}\Rightarrow u={\frac {1}{n}}} d d x ln ( x ) = 1 x lim n → ∞ ln ( 1 + 1 n ) n {\displaystyle {\frac {d}{dx}}\ln(x)={\frac {1}{x}}\lim _{n\to \infty }\ln \left(1+{\frac {1}{n}}\right)^{n}} = 1 x ln [ lim n → ∞ ( 1 + 1 n ) n ] {\displaystyle ={\frac {1}{x}}\ln \left[\lim _{n\to \infty }\left(1+{\frac {1}{n}}\right)^{n}\right]} = 1 x ln e {\displaystyle ={\frac {1}{x}}\ln e} = 1 x {\displaystyle ={\frac {1}{x}}} 用自然对数定义的更一般的对数函数, log b ( x ) = ln ( x ) ln ( b ) {\displaystyle \log _{b}(x)={\frac {\ln(x)}{\ln(b)}}} ,根据其逆函数即一般指数函数的性质,它的导数为[13][14]: d d x log b ( x ) = 1 x ln ( b ) . {\displaystyle {\frac {d}{dx}}\log _{b}(x)={\frac {1}{x\ln(b)}}.} 根据链式法则,以 f ( x ) {\displaystyle f(x)} 为参数的自然对数的导数为 d d x ln [ f ( x ) ] = f ′ ( x ) f ( x ) . {\displaystyle {\frac {d}{dx}}\ln[f(x)]={\frac {f'(x)}{f(x)}}.} 右手端的商叫做 f {\displaystyle f} 的对数导数(英语:logarithmic derivative),通过 ln ( f ( x ) ) {\displaystyle \ln(f(x))} 的导数的方法计算 f ′ ( x ) {\displaystyle f'(x)} 叫做对数微分[15]。 幂级数 编辑 自然对数的导数性质导致了 ln ( 1 + x ) {\displaystyle \ln(1+x)} 在0处的泰勒级数,也叫做麦卡托级数: ln ( 1 + x ) = ∑ n = 1 ∞ ( − 1 ) n + 1 n x n = x − x 2 2 + x 3 3 − ⋯ {\displaystyle \ln(1+x)=\sum _{n=1}^{\infty }{\frac {(-1)^{n+1}}{n}}x^{n}=x-{\frac {x^{2}}{2}}+{\frac {x^{3}}{3}}-\cdots } 对于所有 | x | ≤ 1 , {\displaystyle \left|x\right|\leq 1,} 但不包括 x = − 1. {\displaystyle x=-1.} 把 x − 1 {\displaystyle x-1} 代入 x {\displaystyle x} 中,可得到 ln ( x ) {\displaystyle \ln(x)} 自身的级数。通过在麦卡托级数上使用欧拉变换,可以得到对绝对值大于1的任何 x {\displaystyle x} 有效的如下级数: ln x x − 1 = ∑ n = 1 ∞ 1 n x n = 1 x + 1 2 x 2 + 1 3 x 3 + ⋯ . {\displaystyle \ln {x \over {x-1}}=\sum _{n=1}^{\infty }{1 \over {nx^{n}}}={1 \over x}+{1 \over {2x^{2}}}+{1 \over {3x^{3}}}+\cdots \,.} 这个级数类似于贝利-波尔温-普劳夫公式。 还要注意到 x x − 1 {\displaystyle x \over {x-1}} 是自身的逆函数,所以要生成特定数 y {\displaystyle y} 的自然对数,简单把 x x − 1 {\displaystyle x \over {x-1}} 代入 x {\displaystyle x} 中。 ln x = ∑ n = 1 ∞ 1 n ( x − 1 x ) n = ( x − 1 x ) + 1 2 ( x − 1 x ) 2 + 1 3 ( x − 1 x ) 3 + ⋯ {\displaystyle \ln {x}=\sum _{n=1}^{\infty }{1 \over {n}}\left({x-1 \over x}\right)^{n}=\left({x-1 \over x}\right)+{1 \over 2}\left({x-1 \over x}\right)^{2}+{1 \over 3}\left({x-1 \over x}\right)^{3}+\cdots \,} 对于 Re ( x ) ≥ 1 2 . {\displaystyle \operatorname {Re} (x)\geq {\frac {1}{2}}\,.} 自然数的倒数的总和 1 + 1 2 + 1 3 + ⋯ + 1 n = ∑ k = 1 n 1 k , {\displaystyle 1+{\frac {1}{2}}+{\frac {1}{3}}+\cdots +{\frac {1}{n}}=\sum _{k=1}^{n}{\frac {1}{k}},} 叫做调和级数。它与自然对数有密切联系:当 n {\displaystyle n} 趋于无穷的时候,差 ∑ k = 1 n 1 k − ln ( n ) , {\displaystyle \sum _{k=1}^{n}{\frac {1}{k}}-\ln(n),} 收敛于欧拉-马歇罗尼常数。这个关系有助于分析算法比如快速排序的性能。[16] 积分 编辑 自然对数通过分部积分法积分: ∫ ln ( x ) d x = x ln ( x ) − x + C . {\displaystyle \int \ln(x)\,dx=x\ln(x)-x+C.} 假设: u = ln ( x ) ⇒ d u = d x x {\displaystyle u=\ln(x)\Rightarrow du={\frac {dx}{x}}} d v = d x ⇒ v = x {\displaystyle dv=dx\Rightarrow v=x\,} 所以: ∫ ln ( x ) d x = x ln ( x ) − ∫ x x d x = x ln ( x ) − ∫ 1 d x = x ln ( x ) − x + C {\displaystyle {\begin{aligned}\int \ln(x)\,dx&=x\ln(x)-\int {\frac {x}{x}}\,dx\\&=x\ln(x)-\int 1\,dx\\&=x\ln(x)-x+C\end{aligned}}} 自然对数可以简化形如 g ( x ) = f ′ ( x ) f ( x ) {\displaystyle g(x)={\frac {f'(x)}{f(x)}}} 的函数的积分: g ( x ) {\displaystyle g(x)} 的一个原函数给出为 ln ( | f ( x ) | ) {\displaystyle \ln(\left\vert f(x)\right\vert )} 。这是基于链式法则和如下事实: d d x ln | x | = 1 x . {\displaystyle \ {d \over dx}\ln \left|x\right|={1 \over x}.} 换句话说, ∫ 1 x d x = ln | x | + C {\displaystyle \int {1 \over x}dx=\ln |x|+C} 且 ∫ f ′ ( x ) f ( x ) d x = ln | f ( x ) | + C . {\displaystyle \int {{\frac {f'(x)}{f(x)}}\,dx}=\ln |f(x)|+C.} 例子 编辑 下面是 g ( x ) = tan x {\displaystyle g(x)=\tan x} 的例子: ∫ tan x d x = ∫ sin x cos x d x = ∫ − d d x cos x cos x d x . {\displaystyle {\begin{aligned}\int \tan x\,dx&=\int {\sin x \over \cos x}\,dx\\&=\int {-{d \over dx}\cos x \over {\cos x}}\,dx.\\\end{aligned}}} 设 f ( x ) = cos x {\displaystyle f(x)=\cos x} 且 f ′ ( x ) = − sin x {\displaystyle f'(x)=-\sin x} : ∫ tan x d x = − ln | cos x | + C = ln | sec x | + C {\displaystyle {\begin{aligned}\int \tan x\,dx&=-\ln {\left|\cos x\right|}+C\\&=\ln {\left|\sec x\right|}+C\\\end{aligned}}} 与双曲函数的关系 编辑 在直角双曲线(方程 y = 1 x {\displaystyle y={\frac {1}{x}}} )下,双曲线三角形(黄色),和对应于双曲角 u {\displaystyle u} 的双曲线扇形(红色)。这个三角形的边分别是双曲函数中 cosh {\displaystyle \cosh } 和 sinh {\displaystyle \sinh } 的 2 {\displaystyle {\sqrt {2}}} 倍。 射线出原点交单位双曲线 x 2 − y 2 = 1 {\displaystyle \scriptstyle x^{2}\ -\ y^{2}\ =\ 1} 于点 ( cosh a , sinh a ) {\displaystyle \scriptstyle (\cosh \,a,\,\sinh \,a)} ,这里的 a {\displaystyle \scriptstyle a} 是射线、双曲线和 x {\displaystyle \scriptstyle x} 轴围成的面积的二倍。对于双曲线上位于x轴下方的点,这个面积被认为是负值。 在18世纪,约翰·海因里希·兰伯特介入双曲函数[17],并计算双曲几何中双曲三角形的面积[18]。对数函数是在直角双曲线 x y = 1 {\displaystyle xy=1} 下定义的,可构造双曲线直角三角形,底边在线 y = x {\displaystyle y=x} 上,一个顶点是原点,另一个顶点在双曲线。这里以自然对数即双曲角作为参数的函数,是自然对数的逆函数指数函数,即要形成指定双曲角 u {\displaystyle u} ,在渐近线即x或y轴上需要有的 x {\displaystyle x} 或 y {\displaystyle y} 的值。显见这里的底边是 ( e u + e − u ) 2 2 {\displaystyle \left(e^{u}+e^{-u}\right){\frac {\sqrt {2}}{2}}} ,垂线是 ( e u − e − u ) 2 2 {\displaystyle \left(e^{u}-e^{-u}\right){\frac {\sqrt {2}}{2}}} 。 通过旋转和缩小线性变换,得到单位双曲线下的情况,有: cosh x = e x + e − x 2 {\displaystyle \cosh x={\frac {e^{x}+e^{-x}}{2}}} sinh x = e x − e − x 2 {\displaystyle \sinh x={\frac {e^{x}-e^{-x}}{2}}} 单位双曲线中双曲线扇形的面积是对应直角双曲线 x y = 1 {\displaystyle xy=1} 下双曲角的 1 2 {\displaystyle {\frac {1}{2}}} 。 连分数 编辑 尽管自然对数没有简单的连分数,但有一些广义连分数如: ln ( 1 + x ) = x 1 1 − x 2 2 + x 3 3 − x 4 4 + x 5 5 − ⋯ = x 1 − 0 x + 1 2 x 2 − 1 x + 2 2 x 3 − 2 x + 3 2 x 4 − 3 x + 4 2 x 5 − 4 x + ⋱ {\displaystyle {\begin{aligned}\ln(1+x)&={\frac {x^{1}}{1}}-{\frac {x^{2}}{2}}+{\frac {x^{3}}{3}}-{\frac {x^{4}}{4}}+{\frac {x^{5}}{5}}-\cdots \\&={\cfrac {x}{1-0x+{\cfrac {1^{2}x}{2-1x+{\cfrac {2^{2}x}{3-2x+{\cfrac {3^{2}x}{4-3x+{\cfrac {4^{2}x}{5-4x+\ddots }}}}}}}}}}\\\end{aligned}}} ln ( 1 + x y ) = x y + 1 x 2 + 1 x 3 y + 2 x 2 + 2 x 5 y + 3 x 2 + ⋱ = 2 x 2 y + x − ( 1 x ) 2 3 ( 2 y + x ) − ( 2 x ) 2 5 ( 2 y + x ) − ( 3 x ) 2 7 ( 2 y + x ) − ⋱ {\displaystyle {\begin{aligned}\ln \left(1+{\frac {x}{y}}\right)&={\cfrac {x}{y+{\cfrac {1x}{2+{\cfrac {1x}{3y+{\cfrac {2x}{2+{\cfrac {2x}{5y+{\cfrac {3x}{2+\ddots }}}}}}}}}}}}\\&={\cfrac {2x}{2y+x-{\cfrac {(1x)^{2}}{3(2y+x)-{\cfrac {(2x)^{2}}{5(2y+x)-{\cfrac {(3x)^{2}}{7(2y+x)-\ddots }}}}}}}}\\\end{aligned}}} 这些连分数特别是最后一个对接近1的值快速收敛。但是,更大的数的自然对数,可以轻易的用这些更小的数的自然对数的加法来计算,带有类似的快速收敛。 例如,因为 2 = 1.25 3 × 1.024 {\displaystyle 2=1.25^{3}\times 1.024} ,2的自然对数可以计算为: ln 2 = 3 ln ( 1 + 1 4 ) + ln ( 1 + 3 125 ) = 6 9 − 1 2 27 − 2 2 45 − 3 2 63 − ⋱ + 6 253 − 3 2 759 − 6 2 1265 − 9 2 1771 − ⋱ . {\displaystyle {\begin{aligned}\ln 2&=3\ln \left(1+{\frac {1}{4}}\right)+\ln \left(1+{\frac {3}{125}}\right)\\&={\cfrac {6}{9-{\cfrac {1^{2}}{27-{\cfrac {2^{2}}{45-{\cfrac {3^{2}}{63-\ddots }}}}}}}}+{\cfrac {6}{253-{\cfrac {3^{2}}{759-{\cfrac {6^{2}}{1265-{\cfrac {9^{2}}{1771-\ddots }}}}}}}}.\\\end{aligned}}} 进而,因为 10 = 1.25 10 × 1.024 3 {\displaystyle 10=1.25^{10}\times 1.024^{3}} ,10的自然对数可以计算为: ln 10 = 10 ln ( 1 + 1 4 ) + 3 ln ( 1 + 3 125 ) = 20 9 − 1 2 27 − 2 2 45 − 3 2 63 − ⋱ + 18 253 − 3 2 759 − 6 2 1265 − 9 2 1771 − ⋱ . {\displaystyle {\begin{aligned}\ln 10&=10\ln \left(1+{\frac {1}{4}}\right)+3\ln \left(1+{\frac {3}{125}}\right)\\&={\cfrac {20}{9-{\cfrac {1^{2}}{27-{\cfrac {2^{2}}{45-{\cfrac {3^{2}}{63-\ddots }}}}}}}}+{\cfrac {18}{253-{\cfrac {3^{2}}{759-{\cfrac {6^{2}}{1265-{\cfrac {9^{2}}{1771-\ddots }}}}}}}}.\\\end{aligned}}} 复数对数 编辑 主条目:复数对数 指数函数可以扩展为对任何复数 x {\displaystyle x} 得出复数值为 e x {\displaystyle e^{x}} 的函数,只需要简单使用 x {\displaystyle x} 为复数的无穷级数;这个指数函数的逆函数形成复数对数,并带有正常的对数的多数性质。但是它涉及到了两个困难:不存在 x {\displaystyle x} 使得 e x = 0 {\displaystyle e^{x}=0} ;并且有着 e 2 π i = 1 = e 0 {\displaystyle e^{2\pi i}=1=e^{0}} 。因为乘法性质仍适用于复数指数函数, e z = e z + 2 n π i {\displaystyle e^{z}=e^{z+2n\pi i}} ,对于所有复数 z {\displaystyle z} 和整数 n {\displaystyle n} 。 所以对数不能定义在整个复平面上,并且它是多值函数,就是说任何复数对数都可以增加 2 π i {\displaystyle 2\pi i} 的任何整数倍而成为等价的对数。复数对数只能在切割平面上是单值函数。例如, log i = 1 2 π i {\displaystyle \log i={\frac {1}{2}}\pi i} 或 5 2 π i {\displaystyle {\frac {5}{2}}\pi i} 或 − 3 2 π i {\displaystyle -{\frac {3}{2}}\pi i} 等等;尽管 i 4 = 1 {\displaystyle i^{4}=1} , 4 log = i {\displaystyle 4\log =i} 不能定义为 2 π i {\displaystyle 2\pi i} 或 10 π i {\displaystyle 10\pi i} 或 − 6 π i {\displaystyle -6\pi i} ,以此类推。 自然对数函数在复平面(主分支)上的绘图 z=Re(ln(x+iy)) 前三图的叠加 主值定义 编辑 对于每个非0复数 z = x + y i {\displaystyle z=x+yi} ,主值 log z {\displaystyle \log z} 是虚部位于区间 ( − π , π ] {\displaystyle (-\pi ,\pi ]} 内的对数。表达式 log 0 {\displaystyle \log 0} 不做定义,因为没有复数 w {\displaystyle w} 满足 e w = 0 {\displaystyle e^{w}=0} 。 要对 log z {\displaystyle \log z} 给出一个公式,可以先将 z {\displaystyle z} 表达为极坐标形式, z = r e i θ {\displaystyle z=re^{i\theta }} 。给定 z {\displaystyle z} ,极坐标形式不是确切唯一的,因为有可能向 θ {\displaystyle \theta } 增加 2 π {\displaystyle 2\pi } 的整数倍,所以为了保证唯一性而要求 θ {\displaystyle \theta } 位于区间 ( − π , π ] {\displaystyle (-\pi ,\pi ]} 内;这个 θ {\displaystyle \theta } 叫做幅角的主值,有时写为 arg z {\displaystyle \operatorname {arg} z} 或 atan 2 ( y , x ) {\displaystyle \operatorname {atan} 2(y,x)} 。则对数的主值可以定义为[19] : Log z := ln r + i θ = ln | z | + i Arg z = ln x 2 + y 2 + i atan2 ( y , x ) . {\displaystyle \operatorname {Log} z:={\text{ln }}r+i\theta =\ln |z|+i\operatorname {Arg} z=\operatorname {ln} {\sqrt {x^{2}+y^{2}}}+i\operatorname {atan2} (y,x).} 例如, Log ( − 3 i ) = ln 3 − π i 2 {\displaystyle \operatorname {Log} (-3i)=\ln 3-{\frac {\pi i}{2}}} 。 科学应用 编辑 自然指数有应用于表达放射衰变(放射性)之类关于衰减的过程,如放射性原子数目的微分方程 N {\displaystyle N} 随时间变化率 d N d t = − p N {\displaystyle {\frac {dN}{dt}}=-pN} ,常数 p {\displaystyle p} 为原子衰变概率,积分得 N ( t ) = N ( 0 ) exp ( − p t ) {\displaystyle N(t)=N(0)\exp(-pt)} 。 注释 编辑 ^ 根据微积分学,某函数之定义域为其反函数之值域,反之其值域为其反函数之定义域。因 e x {\displaystyle e^{x}} 的值域为 ( 0 , ∞ ) {\displaystyle (0,\infty )} ,且其为 ln x {\displaystyle \ln x} 之反函数,故可知 ln x {\displaystyle \ln x} 之定义域为 ( 0 , ∞ ) {\displaystyle (0,\infty )} ,即 ln x {\displaystyle \ln x} 在非正实数系无法定义。 ^ 若要避免与底为10的常用对数 log x {\displaystyle \log x\!} 混淆,可用“全写” log e x {\displaystyle \log _{\boldsymbol {e}}x\!} 。 参考资料 编辑 ^ 例如哈代和赖特所著的《数论入门》"Introduction to the theory of numbers" (1.7, Sixth edition, Oxford 2008)的注解 "log x is, of course the 'Napierian' logarithm of x, to base e. 'Common' logarithms have no mathematical interest."(log x 当然是以e为基,x的“纳皮尔”对数。“常用”对数在数学上毫无重要。) ^ 证明:从1到b积分1/x,增加三角形{(0, 0), (1, 0), (1, 1)},并减去三角形{(0, 0), (b, 0), (b, 1/b)}。 ^ Ernest William Hobson, John Napier and the invention of logarithms, 1614, Cambridge: The University Press, 1914 ^ Boyer, Carl B., 14,Section "Jobst Bürgi", A History of Mathematics, New York: John Wiley & Sons, 1991, ISBN 978-0-471-54397-8 ^ 选取接近e的底数b,对数表涉及的bx为单调增函数,定义域为0到1而值域为1到b;选取接近1/e的底数b,对数表涉及的bx为单调减函数,定义域为0到∞而值域为1到0。 ^ 以 10 1 2 54 {\displaystyle 10^{\frac {1}{2^{54}}}} 这个接近1的数为基础。 ^ 博纳文图拉·卡瓦列里在1635年的《Geometria indivisibilibus continuorum nova quadam ratione promota》中给出定积分: ∫ 0 a x n d x = 1 n + 1 a n + 1 n ≥ 0 , {\displaystyle \int _{0}^{a}x^{n}\,dx={\tfrac {1}{n+1}}\,a^{n+1}\qquad n\geq 0,} 其不定积分形式为: ∫ x n d x = 1 n + 1 x n + 1 + C n ≠ − 1. {\displaystyle \int x^{n}\,dx={\tfrac {1}{n+1}}\,x^{n+1}+C\qquad n\neq -1.} 独立发现者还有:皮埃尔·德·费马、罗贝瓦尔的吉尔(英语:Gilles de Roberval)和埃万杰利斯塔·托里拆利。 ^ 设a=1,x轴上[a,b]两点对应的双曲线线段与原点围成的双曲线扇形面积为f(b),[c,d]对应的扇形面积为f(d)-f(c),d=bc,即为f(bc)-f(c),当且仅当f(bc)=f(b)+f(c)时,两双曲线扇形面积相等。 ^ J. J. O'Connor; E. F. Robertson, The number e, The MacTutor History of Mathematics archive, September 2001 [2009-02-02], (原始内容存档于2012-02-19) ^ 卡瓦列里弓形面积公式,对于负数值的n(x的负数幂),由于在x = 0处有个奇点,因此定积分的下限为1,而不是0,即为: ∫ 1 a x n d x = 1 n + 1 ( a n + 1 − 1 ) n ≠ − 1. {\displaystyle \int _{1}^{a}x^{n}\,dx={\tfrac {1}{n+1}}(a^{n+1}-1)\qquad n\neq -1.} 欧拉的自然对数定义: ln ( x ) = lim n → ∞ n ( x 1 / n − 1 ) = lim n → − 1 1 n + 1 ( x n + 1 − 1 ) {\displaystyle {\begin{aligned}\ln(x)&=\lim _{n\rightarrow \infty }n(x^{1/n}-1)\\&=\lim _{n\rightarrow -1}{\tfrac {1}{n+1}}(x^{n+1}-1)\\\end{aligned}}} ^ Maor, Eli, e: The Story of a Number, Princeton University Press, 2009, ISBN 978-0-691-14134-3 ,sections 1, 1.Eves, Howard Whitley, An introduction to the history of mathematics, The Saunders series 6th, Philadelphia: Saunders, 1992, ISBN 978-0-03-029558-4 , section 9-3Boyer, Carl B., A History of Mathematics, New York: John Wiley & Sons, 1991, ISBN 978-0-471-54397-8 , p. 484, 489 ^ ( 1 + 1 n ) x = ( ( 1 + 1 n ) n ) x n {\displaystyle \left(1+{\frac {1}{n}}\right)^{x}=\left(\left(1+{\frac {1}{n}}\right)^{n}\right)^{\frac {x}{n}}} 在最初的概念下,底数是接近1的数,而对数是整数;经过简单变换后,底数变大了,成为接近数学常量e的数,而对数变小了,成为 x/n。 ^ Lang 1997, section IV.2 ^ Wolfram, Stephen. "Calculation of d/dx(Log(b,x))". from Wolfram Alpha: Computational Knowledge Engine, Wolfram Research. (原始内容存档于2011-07-18) (英语). ^ Kline, Morris, Calculus: an intuitive and physical approach, Dover books on mathematics, New York: Dover Publications, 1998, ISBN 978-0-486-40453-0 , p. 386 ^ Havil, Julian, Gamma: Exploring Euler's Constant, Princeton University Press, 2003, ISBN 978-0-691-09983-5 , sections 11.5 and 13.8 ^ Eves, Howard, Foundations and Fundamental Concepts of Mathematics, Courier Dover Publications: 59, 2012, ISBN 9780486132204, We also owe to Lambert the first systematic development of the theory of hyperbolic functions and, indeed, our present notation for these functions. ^ Ratcliffe, John, Foundations of Hyperbolic Manifolds, Graduate Texts in Mathematics 149, Springer: 99, 2006 [2014-03-28], ISBN 9780387331973, (原始内容存档于2014-01-12), That the area of a hyperbolic triangle is proportional to its angle defect first appeared in Lambert's monograph Theorie der Parallellinien, which was published posthumously in 1786. ^ Sarason, Section IV.9. 延伸阅读 编辑 John B. Conway, Functions of one complex variable, 2nd edition, Springer, 1978. Serge Lang, Complex analysis, 3rd edition, Springer-Verlag, 1993. Gino Moretti, Functions of a Complex Variable, Prentice-Hall, Inc., 1964. Donald Sarason, Complex function theory (页面存档备份,存于互联网档案馆), 2nd edition, American Mathematical Society, 2007. E. T. Whittaker(英语:E. T. Whittaker) and G. N. Watson, A Course in Modern Analysis, fourth edition, Cambridge University Press, 1927.