Skip to content

Cool Proofs That You Need to Know or You'll Fail the Exam

$\gdef \N{\mathbb{N}} \gdef \Z{\mathbb{Z}} \gdef \Q{\mathbb{Q}} \gdef \R{\mathbb{R}} \gdef \C{\mathbb{C}} \gdef \setcomp#1{\overline{#1}} \gdef \sseq{\subseteq} \gdef \pset#1{\mathcal{P}(#1)} \gdef \covariant{\operatorname{Cov}} \gdef \of{\circ} \gdef \p{^{\prime}} \gdef \pp{^{\prime\prime}} \gdef \ppp{^{\prime\prime\prime}} \gdef \pn#1{^{\prime\times{#1}}} $

Linearity of Limits of a Sequence

If \(a_n\) and \(b_n\) are infinite sequences, and \(\lim_{n\to \infty}{a_n}=A\) and \(\lim_{n\to \infty}{b_n}=B\), then \(\lim_{n\to \infty}{(a_n+b_n)}=A+B\)

By laws of limits, \(\forall\varepsilon>0,\;\exists N_a,N_b\) such that \(\forall n> \text{max}(N_a,N_b)\), we have:
\(|a_n-A|<\varepsilon/2\) and \(|b_n-B|<\varepsilon/2\).
By the Triangle inequality we have:
\(|a_n+b_n-(A+B)|\le|a_n-A|+|b_n-B|<\varepsilon\)
Therefore, \(\lim_{n\to \infty}{(a_n+b_n)}=A+B\).

\[Q.E.D.\]

Product of Vanishing and Bounded Sequences

If \(a_n\) is a vanishing sequence, and \(b_n\) is a bounded sequence, then \(\lim_{n\to \infty}{(a_nb_n)}=0\)

Let \(\varepsilon>0\). We seek \(N\) such that if \(n>N\), \(|a_nb_n-0|<\varepsilon\).
Because \(b_n\) is bounded, there exists \(M>0\) such that \(|b_n|<M\mid\forall n\in\N\)
Thus, we have \(0\le|a_nb_n|\le M|a_n|\).
\(\lim_{n\to\infty}{a_n}=0\), so \(\exists N\) such that for \(n>N, |a_n-0|<\frac\varepsilon M\)
\(|a_nb_n|<|a_n|\cdot|b_n|<\frac\varepsilon M<\varepsilon\)
So, \(|a_nb_n|<\varepsilon\).

\[Q.E.D.\]

Sum of Derivatives

The derivative of sums is the sum of the derivatives: \({(f+g)}^\prime (x)=f^\prime (x)+g^\prime (x)\)

\[\begin{align} {(f+g)}^\prime (x)&=\lim_{h\to0}{\frac{(f+g)(x+h)-(f+g)(x)}h}\\ &=\lim_{h\to0}{\frac{[f(x+h)+g(x+h)]-[f(x)+g(x)]}h}\\ &=\lim_{h\to0}{\frac{f(x+h)-f(x)+g(x+h)-g(x)}h}\\ &=\lim_{h\to0}{\frac{f(x+h)-f(x)}{h}}+\lim_{h\to0}{\frac{g(x+h)-g(x)}h}\\ &=f^\prime (x)+g^\prime (x)\\ \end{align}\]
\[Q.E.D\]

Product of Derivatives

It’s the product rule! \({(f\cdot g)}^\prime (x)=(fg^\prime )(x)+(f^\prime g)(x)\)

\[\begin{align} {(fg)}^\prime (x)&=\lim_{h\to0}{\frac{(f\cdot g)(x+h)-(f\cdot g)(x)}h}\\ &=\lim_{h\to0}{\frac{[f(x+h)\cdot g(x+h)]-[f(x)\cdot g(x)]}h}\\ &=\lim_{h\to0}{\frac{[f(x+h)\cdot g(x+h)]-f(x+h)\cdot g(x)+f(x+h)\cdot g(x) -[f(x)\cdot g(x)]}h}\\ &=\lim_{h\to0}{\frac{f(x+h)\cdot g(x+h)-f(x+h)\cdot g(x)}h}+\lim_{h\to0}{\frac{f(x+h)\cdot g(x) -f(x)\cdot g(x)}h}\\ &=\lim_{h\to0}{f(x+h)\cdot\frac{g(x+h)-g(x)}{h}}+\lim_{h\to0}{g(x)\cdot\frac{g(x+h)-g(x)}h}\\ &=f(x)g^\prime (x)+f^\prime (x)g(x)\\ &=(fg^\prime )(x)+(f^\prime g)(x) \end{align}\]
\[Q.E.D.\]

Fermat’s Theorem

If \(f\) is a function that is differentiable at \(x_0\), and \(x_0\) is an extreme point on \(f\), then \(f^\prime (x_0)=0\).

W.L.O.G. maximum/minimum, assuming maximum.
In order for \(f\) to be differentiable at \(x_0\), \(f^\prime _-(x_0)=f^\prime _+(x_0)\), and \(f^\prime (x_0)\) must exist.
\(f^\prime _-(x_0)=\lim_{h\to0^-}{\frac{f(x_0+h)-f(x_0)}{h}}\). Since \(f(x_0+h)-f(x_0)<0\), and \(h<0\),
\(f^\prime _-(x_0)\) is the limit of positive numbers, which we know is \(\ge0\).
\(f^\prime _+(x_0)=\lim_{h\to0^+}{\frac{f(x_0+h)-f(x_0)}{h}}\). Since \(f(x_0+h)-f(x_0)<0\), and \(h>0\),
\(f^\prime _-(x_0)\) is the limit of negative numbers, which we know is \(\le0\).
Since \(f^\prime _-(x_0)=f^\prime _+(x_0)\), \(f^\prime (x_0)\) must be zero.
\(\(Q.E.D.\)\)

Rolle’s Theorem

If \(f\) is a function that is continuous in \([a,b]\), and differentiable in \((a,b)\), and \(f(a)=f(b)\), then \(\exists c\in(a,b)\) such that \(f^\prime(c)=0\).

There are two possibilities:

  1. \(f\) is constant in \((a,b)\). Therefore, \(f^\prime (x)=0\mid\forall x\in(a,b)\).
  2. \(f\) is not constant in \((a,b)\). By the Weirstrass Theorem, we know that \(\exists c\in(a,b)\) such that \(c\) is an extreme point on \(f\). By Fermat’s Theorem, \(f^\prime (c)=0\).
\[Q.E.D.\]

Mean Value Theorem (AKA Lagrange’s Theorem)

If \(f\) is a function that is continuous in \([a,b]\), and differentiable in \((a,b)\), then \(\exists c\in(a,b)\) such that \(f^\prime(c)=\frac{f(b)-f(a)}{b-a}\).

Let \(g(x)=f(x)-\frac{(f(b)-f(a))\cdot(x-a)}{b-a}\). Note that \(g(a)=g(b)=f(a)\). Furthermore, \(g(x)\) has the same continuity and differentiability properties as \(f(x)\), since \(g^\prime(x)=f^\prime(x)-\frac{f(b)-f(a)}{b-a}\).
Therefore, we can apply Rolle’s Theorem that \(\exists c\in(a,b)\mid g^\prime(c)=0\).
Thus, we can immediately conclude by the definition of \(g^\prime\) that \(f^\prime(c)=\frac{f(b)-f(a)}{b-a}\).

\[Q.E.D.\]
MVT Corollary I (Constant Corollary)

If the conditions for the MVT are met, and \(f^\prime(x)=0\mid\forall x\in(a,b)\), then \(f\) is constant in \((a,b)\).

Choose \(x_1,x_2\in(a,b)\) so that \(x_1<x_2\). By the MVT, \(\exists c\in(x_1,x_2)\) such that \(f^\prime(c)=\frac{f(x_2)-f(x_1)}{x_2-x_1}\).
By the claim, \(f^\prime(c)=\frac{f(x_2)-f(x_1)}{x_2-x_1}=0\). This can only happen if \(f(x_2)-f(x_1)=0\mid\forall x_1,x_2\in(a,b)\).

\[Q.E.D.\]
MVT Corollary II (Increasing Corollary)

If the conditions for the MVT are met, and \(f^\prime(x)>0\mid\forall x\in(a,b)\), then \(f\) is increasing in \((a,b)\).

This is easily modified for \(f\) decreasing in \((a,b)\), and therefore omitted.

Suppose that \(f\) is not increasing in \((a,b)\). Choose \(x_1,x_2\in(a,b)\) so that \(x_1<x_2\). By the MVT, \(\exists c\in(x_1,x_2)\) such that \(f^\prime(c)=\frac{f(x_2)-f(x_1)}{x_2-x_1}\).
Since \(f\) is not increasing, and \(x_1<x_2\), \(f(x_2)-f(x_1)\le0\), and \(x_2-x_1>0\). Therefore, \(f^\prime(c)\le0\), which is a contradiction of the claim. By contradiction, \(f\) must be increasing in \((a,b)\).

\[Q.E.D.\]
MVT Corollary III (Other Constant Corollary)

If the conditions for the MVT are met for functions \(f\) and \(g\), and \(f^\prime(x)=g^\prime(x)\mid\forall x\in(a,b)\), then \(f\) and \(g\) differ by a constant—that is, \(f(x)=g(x)+C\mid\forall x\in(a,b),C\in\R\).

Define \(h(x)=f(x)-g(x)\). Then, \(h^\prime(x)=f^\prime(x)-g^\prime(x)=0\). By Corollary I, we know that if the derivative of a function is zero, that function is a constant. Therefore, \(h(x)=f(x)-g(x)=C\), which is the same thing as saying \(f(x)=g(x)+C\mid\forall x\in(a,b),C\in\R\)

\[Q.E.D.\]

Fundamental Theorem of Calculus

If \(f\) is continuous on \([a,b]\), and \(F(x)=\int_a^xf(t)dt\) for \(a\le x\le b\), then \(F\) is continuous on \([a,b]\) and differentiable on \((a,b)\), and \(f^\prime (x)=f(x)\).

\[\begin{align} f^\prime (x)&=\lim_{h\to0}{\frac{F(x+h)-F(x)}h}\\ &=\lim_{h\to0}{\frac1h\left[\int_a^{a+h}f(t)dt-\int_a^xf(t)dt\right]}\\ &=\lim_{h\to0}{\frac1{(x+h)-x}\int_x^{x+h}f(t)dt}\\ \text{Via MVT for Integrals: }&=f(c)\mid c\in(x,x+h) \end{align}\]

Since \(x\le c\le x+h\), and \(h\to0\), \(c=x\).
Therefore, \(f^\prime (x)=f(x)\).

\[Q.E.D.\]