Skip to main content

Section 1.2 Important Theorems from Calculus

We now review a number of theorems that are usually presented in Calculus. Here we both state a number of important theorems, but show some examples and then prove most of these. For additional information on these, check any standard Calculus book. A nice open-source one can be found at APEX Calculus.
This is a important theorem, however, there is much background needed to prove this. A good open-source book with a proof is in Introduction to Real Analysis by William F. Trench
 1 
The background and proof of this is in section 2.2
.

Proof.

From the Intermediate Value Theorem, \(f\) takes on a minimum value, \(m\) and maximum value, \(M\) on the interval \([a,b]\text{.}\)
  1. If \(m=M\text{,}\) then the function is constant and all values of \(c\) in the interval \((a,b)\) satisfy \(f'(c)=0\text{.}\)
  2. Otherwise, either the minimum or maximum or both of \(f\) occurs in \((a,b)\text{.}\) Let \(c\) be this value. Since the function is differentiable at \(c\) and is a minimum or a maximum, then \(f'(c)=0\text{.}\)

Example 1.2.3.

Show that \(f(x)=x^{2} +6x\) on the interval \([-6,0]\) satisfies all the conditions of Rolle’s Theorem. Find all of the values of \(c\) such that \(f'(c)=0\text{.}\)
Solution.
Since \(f\) is a quadratic, it is continuous and differentiable everywhere, so the first two conditions are satisfied. Also, \(f(-6)=(-6)^{2}+6(-6)=0\) and \(f(0)=0^{2}+6(0)=0\text{,}\) so the function is equal on the endpoints of the interval and the 3rd condition is satisfied.
To find the values of \(c\) such that \(f'(c)=0\text{,}\) we solve \(f'(c)=2c+6=0\) or \(c=-3\text{.}\)
A graphical view of this is as follows:
Figure 1.2.4.
and one can see that the function is 0 at both \(x=-6\) and \(x=0\) and that there is a horizontal tangent line at \(c=-3\text{.}\)
We can extend Rolle’s theorem to functions that have more than 2 equal points. We will show an example of the theorem, but in general, this is very important for proving other theorems.

Proof.

First, for some notation, assume that \(a_{k}\) is the \(k\)th zero of \(f\) for \(k=1,2,\ldots,n+1\) and the \(a_{k}\) are ordered increasingly. Therefore there are \(n\) subintervals of \([a,b]\text{,}\) denoted \([a_{k},a_{k+1}]\) such that \(f(a_{k})=f(a_{k+1})=0\text{.}\) Using Rolle’s theorem, for each subinterval, there exists a number \(c_{k}\) in \((a_{k},a_{k+1})\) such that \(f'(c_{k})=0\text{.}\)
Therefore there are \(n\) numbers, \(c_{k}\) such that the derivative is 0 or subintervals \([c_{k},c_{k+1}]\) such that \(f'(c_{k})=f'(c_{k+1})=0\text{.}\) Using Rolle’s theorem again, for each \(k\text{,}\) there exists a number \(d_{k} \in (c_{k},c_{k+1})\) such that \(f''(d_{k})=0\text{.}\)
Repeat this process \(n-2\) times and the theorem holds.
The following is a simple example of Generalized Rolle’s theorem. We will see this theorem used in a proof later in the text.

Example 1.2.6.

Let \(p(x)=(x+2)(x+1)x(x-1)(x-2)\text{.}\) Show that this satisfies Generalized Rolle’s theorem on \([-2,2]\) for \(n=4\) and find the value of \(c\) such that \(f^{(4)}(c)=0\text{.}\)
First, there are zeros of \(p\) at \(x=-2,-1,0,1,2\) and since \(p\) is a polynomial, it is infinitely differentiable, so it satisfies the conditions of the theorem.
Also, \(p\) can be written:
\begin{equation*} p(x) = x(x^{2}-1)(x^{2}-4) = x^{5}-5x^{3}+4x \end{equation*}
and \(p^{(4)}(x) = 120x\) and when \(c=0\text{,}\) \(p^{(4)}(0)=0\text{.}\)
Also, consider a plot of the above situation.
Figure 1.2.7.
and as can be seen that the function is 0 at the 5 points \(x=-2,-1,0,1,2\) and any function with 5 zeros like this either wiggle a lot like the above function, or is constant.

Proof.

Let
\begin{equation*} g(x) = f(x) +(x-a) \frac{f(b)-f(a)}{b-a}-f(a). \end{equation*}
Then
\begin{equation*} g'(x) = f'(x) -\frac{f(b)-f(a)}{b-a} \end{equation*}
and evaluated \(g\) on the endpoints of the interval shows that
\begin{equation*} \begin{aligned} g(a) \amp = f(a) + (a-a) \frac{f(b)-f(a)}{b-a}- f(a) = 0, \\ g(b)\amp= f(b) + (b-a) \frac{f(b)-f(a)}{b-a}- f(a) = 0. \end{aligned} \end{equation*}
Since \(f\) is continuous on \([a,b]\) and differentiable of \((a,b)\text{,}\) so is \(g\) and therefore \(g\) satisfies Rolle’s Theorem. Then there exists a number \(c\) in the interval \((a,b)\) such that \(g'(c)=0\) or
\begin{align*} g'(c) \amp = f'(c) - \frac{f(b)-f(a)}{b-a}= 0\\ \text{or} \qquad \amp \\ f'(c) \amp = \frac{f(b)-f(a)}{b-a} \end{align*}
so the theorem holds.
The following example shows a nice consequence of the mean value theorem.

Example 1.2.9.

Use the Mean Value Theorem to proof that \(e^{x} \geq x+1\) for all \(x\geq 0\text{.}\)
Solution.
Let \(f(x) = e^{x}\text{,}\) \(b=x\text{,}\) \(a=0\) then the MVT applied to this situation is
\begin{equation*} e^{\xi} = \frac{e^{x} - 1}{x-0} \end{equation*}
since \(e^{\xi}\geq 1\text{,}\) for \(\xi\geq 0\text{,}\) then
\begin{equation*} \begin{aligned} \frac{e^{x}-1}{x}\amp \geq 1 \\ e^{x} \amp \geq x+1 \end{aligned} \end{equation*}