Added proof techniques draft
This commit is contained in:
parent
2de7e7536b
commit
59f8cc3657
31
Advanced/Proof Techniques/main.tex
Executable file
31
Advanced/Proof Techniques/main.tex
Executable file
@ -0,0 +1,31 @@
|
||||
% use [nosolutions] flag to hide solutions.
|
||||
% use [solutions] flag to show solutions.
|
||||
\documentclass[
|
||||
solutions,
|
||||
singlenumbering,
|
||||
unfinished
|
||||
]{../../resources/ormc_handout}
|
||||
\usepackage{../../resources/macros}
|
||||
|
||||
\usepackage{units}
|
||||
|
||||
\uptitlel{Math Circle Basics}
|
||||
\uptitler{}
|
||||
\title{Proof Techniques}
|
||||
\subtitle{Prepared by \githref{Mark} on \today{}}
|
||||
|
||||
% Default \implies is ugly
|
||||
\let\implies\Rightarrow
|
||||
\let\rimplies\Leftarrow
|
||||
\let\iff\Leftrightarrow
|
||||
\let\notimplies\nRightarrow
|
||||
|
||||
\begin{document}
|
||||
|
||||
\maketitle
|
||||
|
||||
\input{parts/0 intro}
|
||||
\input{parts/1 contradiction}
|
||||
\input{parts/2 induction}
|
||||
|
||||
\end{document}
|
203
Advanced/Proof Techniques/parts/0 intro.tex
Normal file
203
Advanced/Proof Techniques/parts/0 intro.tex
Normal file
@ -0,0 +1,203 @@
|
||||
\section{Introduction}
|
||||
|
||||
|
||||
\definition{}
|
||||
A \textit{proof} is a mathematical argument that irrefutably
|
||||
demonstrates the truth of a given proposition.
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
Every proof involves some sort of \textit{implication}, denoted $\implies$. \par
|
||||
The statement \say{$A$ implies $B$} (written $A \implies B$), means that $B$ is true whenever $A$ is true.
|
||||
|
||||
\problem{}<trueimplies>
|
||||
Which of the following are true? \par
|
||||
\note{You don't need to provide a proof.}
|
||||
|
||||
\begin{itemize}
|
||||
\item $x$ is prime $\implies$ $x$ is odd.
|
||||
\item $x$ is real $\implies$ $x$ is rational.
|
||||
\item $x$ is odd $\implies$ $x$ is prime.
|
||||
\end{itemize}
|
||||
|
||||
\vfill
|
||||
|
||||
|
||||
|
||||
\problem{}
|
||||
As you saw above, $A \implies B$ does not guarantee that $B \implies A$. \par
|
||||
Find two new statements $A$ and $B$ so that $A \implies B$ but $B \notimplies A$. \par
|
||||
\hint{\say{new} as in \say{not from \ref{trueimplies}}}
|
||||
|
||||
\begin{solution}
|
||||
A fairly trite example is below. \par
|
||||
Note that \say{$X$ is a square} is a subset of the statement \say{$X$ is a rectangle.}
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
$X$ is a square $\implies$ $X$ is a rectangle, but $X$ is a rectangle $\notimplies$ $X$ is a square.
|
||||
\end{solution}
|
||||
|
||||
|
||||
|
||||
\vfill
|
||||
\pagebreak
|
||||
|
||||
|
||||
\definition{}<iffdef>
|
||||
As we just saw, implication is one-directional. \par
|
||||
The statements $A \implies B$ and $B \implies A$ are independent of one another. \par
|
||||
|
||||
\vspace{1mm}
|
||||
|
||||
If both are true, we write $A \iff B$. This can be read as \say{$A$ if and only if $B$.} \par
|
||||
In text, \say{if and only if} is often abbreviated as iff. \par
|
||||
|
||||
\vspace{1mm}
|
||||
|
||||
Bidirectional implication is the strongest relationship we can have between two statements: \par
|
||||
If $A \iff B$, $A$ and $B$ are \textit{equivalent.} They are always either both true or both false.
|
||||
|
||||
\definition{}
|
||||
The \textit{floor} of $x$ is the largest integer $a$ so that $a \leq x$. This is denoted $\lfloor x \rfloor$. \par
|
||||
The \textit{ceiling} of $x$ is the largest integer $a$ so that $a \geq x$. This is denoted $\lceil x \rceil$.
|
||||
|
||||
|
||||
\generic{Property:}
|
||||
If $b_1 \leq a \leq b_2$ and $b_1 = b_2$, we must have that $b_1 = a = b_2$. \par
|
||||
|
||||
\vspace{1mm}
|
||||
|
||||
Also, if $a \leq b$ and $a \geq b$, we must have that $a = b$. \par
|
||||
This is a trick we often use when showing that two quantities are equal.
|
||||
|
||||
|
||||
|
||||
|
||||
\problem{}
|
||||
Although $A \iff B$ looks like a single statement, we often need to prove each direction seperately. \par
|
||||
Show that $x \in \mathbb{Z}$ iff $\lfloor x \rfloor = \lceil x \rceil$
|
||||
|
||||
\begin{solution}
|
||||
\textbf{Forwards:} $x \in \mathbb{Z} ~\implies~ \lfloor x \rfloor = \lceil x \rceil$ \par
|
||||
If $x \in \mathbb{Z}$, by definition we have that $\lfloor x \rfloor = x$ and $\lceil x \rceil = x$ \par
|
||||
So, $\lfloor x \rfloor = \lceil x \rceil$
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
\textbf{Backwards:} $x \in \mathbb{Z} ~\rimplies~ \lfloor x \rfloor = \lceil x \rceil$ \par
|
||||
Assume that $\lfloor x \rfloor = \lceil x \rceil$, and show that $x$ is an integer. \par
|
||||
Note that $\lfloor x \rfloor \leq x \leq \lceil x \rceil$ (by definition) \par
|
||||
Since $\lfloor x \rfloor = \lceil x \rceil$, we must have that $\lfloor x \rfloor = x = \lceil x \rceil$. \par
|
||||
$\lfloor x \rfloor$ is an integer, so $x$ must be an integer.
|
||||
\end{solution}
|
||||
|
||||
|
||||
\vfill
|
||||
\pagebreak
|
||||
|
||||
\problem{}
|
||||
We don't always need to prove each direction of an iff statement seperately. \par
|
||||
|
||||
\begin{itemize}[itemsep = 1mm]
|
||||
\item Convince yourself that we can \say{chain} iffs together: \par
|
||||
If we show that $A \iff B \iff C \iff D$, do we know that $A \iff D$?
|
||||
|
||||
\item Does this still work if $A \iff B \implies C \iff D$?
|
||||
|
||||
\item Show that $x^2 - 6x - 6 = 3 \iff x = 3$ by building a chain of iffs. \par
|
||||
\hint{You remember how to factor quadratics, right?}
|
||||
\end{itemize}
|
||||
|
||||
\begin{solution}
|
||||
Does this still work if $A \iff B \implies C \iff D$? \par
|
||||
Of course not. $D \notimplies A$ since $C \notimplies B$.
|
||||
We can only conclude that $A \implies D$.
|
||||
|
||||
\linehack{}
|
||||
|
||||
$x^2 - 6x - 6 = 3$ \par
|
||||
$\iff x^2 - 6x - 9 = 0$ \par
|
||||
$\iff (x-3)^2 = 0$ \par
|
||||
$\iff x-3 = 0$ \par
|
||||
$\iff x = 0$
|
||||
|
||||
Note that this is a well-defined argument. Every step is an iff statement we can rigorously justify.
|
||||
We're not hand-wavily \say{rearranging} one equation into another,
|
||||
we're building a chain of implications that eventually bring us to our result.
|
||||
This is the logic behind most algebraic proofs.
|
||||
\end{solution}
|
||||
|
||||
|
||||
\vfill
|
||||
|
||||
\problem{}
|
||||
Another trick you may find useful is the \say{implication cycle.} \par
|
||||
Convince yourself that if $A \implies B \implies C \implies D \implies A$, \par
|
||||
we can conclude that $A$, $B$, $C$, and $D$ are equivalent. \par
|
||||
\note{$A \iff B$ means that $A$ and $B$ are equivalent. See \ref{iffdef}.}
|
||||
|
||||
\vfill
|
||||
|
||||
\problem{Bonus}
|
||||
Use an implication cycle to show that the following definitions of a \textit{squarefree integer} are equivalent.
|
||||
%\hint{Show that $A \implies B \implies C \implies D \implies A$}
|
||||
\begin{enumerate}
|
||||
\item $n^2$ does not divide $q$ for any $n \in \mathbb{Z}^+$, $n \neq 1$
|
||||
\item $p^2$ does not divide $q$ for any prime $p$
|
||||
\item $q$ is a product of distinct primes
|
||||
\item $q ~|~ n^k \implies q ~|~ n$ for all $n, k \in \mathbb{Z}^+$
|
||||
\end{enumerate}
|
||||
|
||||
\begin{solution}
|
||||
Assume $q$ has a square factor, so that $q = an^2$ for some $a, n \in \mathbb{Z}^+$ \par
|
||||
By D, we know that $q ~|~ (an)^2 \implies q ~|~ an$ \par
|
||||
But $q ~|~ an \implies an^2 ~|~ an$ \par
|
||||
$\implies n = 1$
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
So, $q$ cannot have a square factor that isn`t 1.
|
||||
\end{solution}
|
||||
|
||||
|
||||
\vfill
|
||||
\pagebreak
|
||||
|
||||
Often enough, proving a statement is simply a matter of \say{definition chasing,}
|
||||
where we expand the symbols used in the statement we're proving, and then do a bit of
|
||||
rearranging to arrive at the result we want.
|
||||
|
||||
|
||||
\definition{}
|
||||
Let $n, x \in \mathbb{Z}$. \par
|
||||
We say \say{$n$ divides $x$} if $x = kn$ for some $k \in \mathbb{Z}$
|
||||
|
||||
\definition{}
|
||||
Let $a, b \in \mathbb{Z}$ and $n \in \mathbb{Z}^+$. \par
|
||||
We say \say{$a$ is congruent to $b$ modulo $n$} (and write $a \equiv_{n} b$) if $n$ divides $a - b$. \par
|
||||
|
||||
\definition{}
|
||||
Let $a, b \in \mathbb{Z}$. We define $a ~\%~ b$ as the remainder of $a \div b$.
|
||||
|
||||
\problem{}
|
||||
Let $a, b, n$ be positive integers. \par
|
||||
Show that $a + b ~\equiv_n (a ~\%~ n) + (b ~\%~ n)$
|
||||
|
||||
\begin{solution}
|
||||
$a + b ~\equiv_n (a ~\%~ n) + (b ~\%~ n)$ \par
|
||||
|
||||
\vspace{2mm}
|
||||
...can be rewritten as... \par
|
||||
$n$ divides $a + b - (a ~\%~ n) - (b ~\%~ n)$ \par
|
||||
|
||||
\vspace{2mm}
|
||||
...which can be rearranged to... \par
|
||||
$n$ divides $(a - (a ~\%~ n)) + (b - (b ~\%~ n))$
|
||||
|
||||
\vspace{2mm}
|
||||
...which is clearly true, if you think about the meaning of \say{$n$ divides $x$} and \say{$a ~\%~ b$}.
|
||||
\end{solution}
|
||||
|
||||
\vfill
|
||||
\pagebreak
|
63
Advanced/Proof Techniques/parts/1 contradiction.tex
Normal file
63
Advanced/Proof Techniques/parts/1 contradiction.tex
Normal file
@ -0,0 +1,63 @@
|
||||
\section{Proofs by Contradiction}
|
||||
|
||||
\definition{}
|
||||
A very common (and somewhat contraversial) proof technique is
|
||||
\textit{proof by contradiction}. It works as follows:
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
Say we want to prove a statement $P$. Assume that $P$ is false, and show that this implies a false statement.
|
||||
In other words, we show that $P$ can't \textit{not} be true. \par
|
||||
If it's false, we either get a known impossibility ($1 = 2$, pigs fly, et cetera), \par
|
||||
or we find that (not $P$) $\implies$ (not (not $P$)), which is a contradiction in itself.
|
||||
|
||||
|
||||
\problem{}
|
||||
Show that the set of integers has no maximum using a proof by contradiction.
|
||||
|
||||
\begin{solution}
|
||||
Assume there is a maximal integer $x$. \par
|
||||
$x + 1$ is also an integer. \par
|
||||
$x + 1$ is larger than $x$, which contradicts our original assumtion!
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
This is a \textit{proof by infinite descent}, a special type of proof by contradiction.\par
|
||||
Such proofs have the following structure:
|
||||
\begin{itemize}
|
||||
\item Assume there is a smallest (or largest) object with property $X$.
|
||||
\item Show that we have an even smaller object that has the same property $X$.
|
||||
\end{itemize}
|
||||
\end{solution}
|
||||
|
||||
\vfill
|
||||
|
||||
\definition{}
|
||||
We say a number $x \in \mathbb{R}$ is \textit{rational} if we can write $x$ as $\nicefrac{p}{q}$, \par
|
||||
where $p, q$ are integers with no common factors.
|
||||
|
||||
\problem{}
|
||||
Show that $\sqrt{2}$ is irrational. \par
|
||||
\hint{Start by chasing definitions. \say{Not irrational} $=$ \say{rational.}}
|
||||
|
||||
\begin{solution}
|
||||
Suppose $\sqrt{2}$ is rational. Then, there exist $p, q$ so that $\sqrt{2} = \frac{p}{q}$.
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
Multiply by $q$ and square to find that $2q^2 = p^2$, which implies that $p^2$ is even. \par
|
||||
This then implies that $p$ is even, \par
|
||||
which implies that $p^2$ is divisible by 4, \par
|
||||
which implies that $q^2$ is divisible by 2, \par
|
||||
and thus we see that $q$ is also even.
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
$p$ and $q$ are both even, so they cannot be coprime. \par
|
||||
Therefore, we cannot write $\sqrt{2}$ as $\nicefrac{p}{q}$ for comprime $p, q$, \par
|
||||
and $\sqrt{2}$ is therefore irrational.
|
||||
\end{solution}
|
||||
|
||||
|
||||
\vfill
|
||||
\pagebreak
|
41
Advanced/Proof Techniques/parts/2 induction.tex
Normal file
41
Advanced/Proof Techniques/parts/2 induction.tex
Normal file
@ -0,0 +1,41 @@
|
||||
\section{Proofs by Induction}
|
||||
|
||||
\definition{}
|
||||
The last proof technique we'll discuss in this handout is \textit{induction.} \par
|
||||
This is particularly useful when we have a \say{countable} variable, usually an integer. \par
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
A proof by induction consists of two parts: a \textit{base case} and a \textit{inductive step}. \par
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
\vfill
|
||||
|
||||
Note that although induction is a powerful proof technique, it usually leads to uninteresting results. \par
|
||||
If we prove a statement using induction, we conclude that it is true---but we get very little insight on
|
||||
\textit{why} that is.
|
||||
|
||||
\vspace{2mm}
|
||||
|
||||
Alternative proofs are take a bit more work than inductive proofs, but they are much more valuable. \par
|
||||
For example, consider the following proof of X:
|
||||
|
||||
\makeatletter
|
||||
\@makeORMCbox{tmpbox}{Alternative Proof}{ogrape!10!white}{ogrape}
|
||||
\makeatother
|
||||
|
||||
\begin{tmpbox}
|
||||
sdfasdf
|
||||
\end{tmpbox}
|
||||
|
||||
|
||||
\pagebreak
|
Loading…
x
Reference in New Issue
Block a user