Home Up Back Forward Leave Message


We know that the exponential function $e^x$ and the natural logarithm $log(x)$ are inverses of each other, thus $e^{log(x)}=x$ and $log(e^x) = x$. We can use Taylor Series to determine a series for each of these functions.

Taylor Series of exponential function

We know the following:
$$ \begin{aligned} f^{(1)}(x) & = {d \over dx}\left( e^x \right) = e^x\\ f^{(n)}(x) & = {d^{(n)} \over dx}\left( e^x \right) = e^x\\ e^0 & = 1\\ f^{(n)}(0) & = {d^{(n)} \over dx}\left( e^0 \right) = e^0 = 1\\ \end{aligned} $$
Therefore the Taylor Series of the exponent function is represented by:
$$ \begin{aligned} e^x & = \sum_{j=0}^\infty f^{(j)}(0) \cdot {{x^j} \over j!}\\ & = \sum_{j=0}^\infty 1 \cdot {{x^j} \over j!}\\ e^x & = \sum_{j=0}^\infty {{x^j} \over j!}\\ \end{aligned} $$
If we apply this to $x=1$ then we get the following:
$$ \begin{aligned} e^1 & = \sum_{j=0}^\infty {{1^j} \over j!}\\ e^1 & = \sum_{j=0}^\infty {1 \over j!}\\ e & = 1 + {1 \over 1!} + {1 \over 2!} + {1 \over 3!} + {1 \over 4!} + ... + {1 \over n!} + .....\\ & = 1 + 1 + {1 \over 2} + {1 \over 6} + {1 \over 24} + {1 \over 120} + ......\\ & = 2.718281828459045...\\ \end{aligned} $$
We can in fact use this to determine $e$ to any number of digits. The first 100 digits determined by this method are:
$$ \begin{aligned} e = 2.71828182845904523536028747135266249775724709369995957496696762772407663035354759457138217852516642742....\\ \end{aligned} $$

Taylor Series for natural logarithm

When we want to work the natural logarithm we have a problem if we try to determine $log(0)$ seeing that it is undefined. We can overcome this by setting $a=1$ in the following:
$$ \begin{aligned} f(x) & = \sum_{j=0}^\infty f^{(j)}(a) \cdot {{(x-a)^j} \over j!}\\ f(x) & = \sum_{j=0}^\infty f^{(j)}(1) \cdot {{(x-1)^j} \over j!}\\ log(1) & = 0\\ log^{(1)}(a) & = {1 \over a}\\ log^{(1)}(1) & = {1 \over 1} = 1\\ log^{(2)}(a) & = -{1 \over {a^2}}\\ log^{(2)}(1) & = -{1 \over {1^2}} = -1\\ log^{(3)}(a) & = {2 \over {a^3}}\\ log^{(3)}(1) & = {2 \over {1^3}} = 2\\ log^{(n)}(a) & = (-1)^{n+1} \cdot {{(n-1)!} \over {a^n}}\\ log^{(n)}(1) & = (-1)^{n+1} \cdot {{(n-1)!} \over {1^n}} = (-1)^{n+1} \cdot (n-1)!\\ \therefore\, log(x) & = \sum_{j=1}^\infty (-1)^{j+1} \cdot (j-1)! \cdot {{(x-1)^j} \over j!}\\ & = \sum_{j=1}^\infty (-1)^{j+1} \cdot {{(x-1)^j} \over j}\\ \therefore\, log(x+1) & = \sum_{j=1}^\infty (-1)^{j+1} \cdot {{x^j} \over j}\\ & = x - {{x^2} \over 2} + {{x^3} \over 3} - {{x^4} \over 4} + ..... (-1)^{n+1} {{x^n} \over n} + ..... \end{aligned} $$
We can however see a problem will arise if $x > 1$. Let's take $x = 1+y$, then $x^n = {(1+y)}^n = 1 + n \cdot y + ....$. If we let $n$ goes towards positive infinity $1 + n \cdot y$ will go toward infinity and seeing that the difference between $x^n$ and $x^{n-1}$ does not converge the sum does not as well.

We can see if we make $x=0$ then
$$ \begin{aligned} log(1+0) & = 0 - {{0^2} \over 2} + {{0^3} \over 3} - {{0^4} \over 4} + ..... (-1)^{n+1} {{0^n} \over n} + .....\\ log(1) & = 0\\ \end{aligned} $$