Discrete Mathematics Guide

Formula Sheet

UseFormula / ruleMeaning
Logic symbols\(\neg p,\ p\land q,\ p\lor q,\ p\oplus q,\ p\to q,\ p\leftrightarrow q\)not; and; inclusive or; exclusive or; if-then; iff
NAND / NOR\(p\uparrow q=\neg(p\land q)\), \(p\downarrow q=\neg(p\lor q)\)NAND means not both; NOR means neither
Implication\(p\to q\equiv \neg p\lor q\)false only when \(p\) is true and \(q\) is false
Contrapositive\(p\to q\equiv \neg q\to \neg p\)same truth value
De Morgan\(\neg(p\land q)\equiv \neg p\lor\neg q\)
\(\neg(p\lor q)\equiv \neg p\land\neg q\)
switch AND/OR when negating
Quantifiers\(\neg\forall xP(x)\equiv\exists x\neg P(x)\)
\(\neg\exists xP(x)\equiv\forall x\neg P(x)\)
flip the quantifier and negate the inside statement
Even / odd\(n=2k\); \(n=2k+1\), \(k\in\mathbb Z\)definitions
Rational\(x=a/b\), \(a,b\in\mathbb Z\), \(b\ne0\)definition
Divisible by \(d\)\(d\mid n\) means \(n=dk\) for some \(k\in\mathbb Z\)definition
Inductionbase case; assume \(P(k)\); prove \(P(k+1)\)prove statements for integers
Factorial\(n!=n(n-1)\cdots2\cdot1\), \(0!=1\)arrange all \(n\) objects
Permutation\(P(n,r)=\dfrac{n!}{(n-r)!}\)choose and arrange \(r\) from \(n\)
Combination\(\binom nr=\dfrac{n!}{r!(n-r)!}\)choose \(r\) from \(n\); order ignored
Pigeonhole\(k(m-1)+1\)force \(m\) objects into one of \(k\) boxes
Probability\(P(E)=\dfrac{|E|}{|S|}\)equally likely outcomes
Conditional probability\(P(A\mid B)=\dfrac{P(A\cap B)}{P(B)}\)probability of \(A\) after \(B\) is known
Bayes\(P(A\mid B)=\dfrac{P(B\mid A)P(A)}{P(B)}\)reverse a conditional probability
Independence\(P(A\cap B)=P(A)P(B)\)also \(P(A\mid B)=P(A)\), if \(P(B)>0\)
Expected value\(E(X)=\sum_{s\in S}X(s)p(s)\)weighted average of the values of \(X\)
Variance\(\operatorname{Var}(X)=E(X^2)-[E(X)]^2\)spread of a random variable

Chapter 1 - Logic and Proofs

Symbols
SymbolRead it asExample
\(\neg p\)not \(p\)not raining
\(p\land q\)\(p\) and \(q\)both true
\(p\lor q\)\(p\) or \(q\), or bothinclusive or
\(p\oplus q\)\(p\) or \(q\), but not bothexclusive or
\(p\to q\)if \(p\), then \(q\)hypothesis \(p\), conclusion \(q\)
\(p\leftrightarrow q\)\(p\) iff \(q\)same truth value
\(p\uparrow q\)not both \(p\) and \(q\)\(\neg(p\land q)\)
\(p\downarrow q\)neither \(p\) nor \(q\)\(\neg(p\lor q)\)
Truth table equivalence

Check whether \(p\to q\) and \(\neg q\to\neg p\) are equivalent.

Answer steps
\(p\)\(q\)\(p\to q\)\(\neg q\to\neg p\)
TTTT
TFFF
FTTT
FFTT

The final columns match, so they are equivalent.

Negating a quantified statement

Negate \(\forall x\exists y\,(x=y^2)\) so that the negation symbol is gone.

Answer steps
\[\neg\forall x\exists y\,(x=y^2)\] \[\equiv \exists x\neg\exists y\,(x=y^2)\] \[\equiv \exists x\forall y\neg(x=y^2)\] \[\equiv \exists x\forall y\,(x\ne y^2).\]
Nested quantifiers

What is the difference between \(\forall x\exists yP(x,y)\) and \(\exists y\forall xP(x,y)\)?

Answer steps

\(\forall x\exists yP(x,y)\): every \(x\) gets at least one \(y\). The \(y\) may change.

\(\exists y\forall xP(x,y)\): one special \(y\) works for every \(x\).

Example: if \(P(x,y)\) means "\(y\) is the parent of \(x\)," then "every person has a parent" is \(\forall x\exists yP(x,y)\). "One person is everyone's parent" is \(\exists y\forall xP(x,y)\).

Exactly one translation

Let \(F(x,y)\) mean "\(x\) can fool \(y\)." Write: there is exactly one person whom everybody can fool.

Answer steps
\[\exists y\left(\forall xF(x,y)\land\forall z(\forall xF(x,z)\to z=y)\right).\]

The first part says everybody can fool \(y\). The second part says any person with that property must be \(y\).

Proof example

Prove: if \(x\) and \(y\) are even integers, then \(3x-4y\) is even.

Answer steps

An integer \(n\) is even if \(n=2k\) for some \(k\in\mathbb Z\).

Since \(x\) and \(y\) are even, let \(x=2a\) and \(y=2b\), where \(a,b\in\mathbb Z\).

\[3x-4y=3(2a)-4(2b)=6a-8b=2(3a-4b).\]

Since \(3a-4b\in\mathbb Z\), the expression is even.

Counterexample

Disprove: the product of two irrational numbers is always irrational.

Answer steps

One counterexample is enough.

\[\sqrt2\cdot\sqrt2=2.\]

Both factors are irrational, but the product is rational.

Chapter 5 - Induction, Recursion, Algorithms

Induction structure
  1. Base case: prove the first case.
  2. Inductive hypothesis: assume \(P(k)\) is true.
  3. Inductive step: use \(P(k)\) to prove \(P(k+1)\).
Recursive formula + induction

Let \(f(0)=0\) and \(f(n)=f(n-1)+2n-1\). Find \(f(1),f(2),f(3),f(4)\), guess a formula, and prove it.

Answer steps
\[f(1)=1,\quad f(2)=4,\quad f(3)=9,\quad f(4)=16.\]

The formula is \(f(n)=n^2\).

Base case: \(f(0)=0=0^2\).

Assume \(f(k)=k^2\). Then

\[f(k+1)=f(k)+2(k+1)-1=k^2+2k+1=(k+1)^2.\]
Recursive algorithm

Write a recursive algorithm that computes \(nx\) using only addition, where \(n\) is a nonnegative integer.

Answer steps
multiply(n, x):
    if n = 0 return 0
    else return x + multiply(n - 1, x)

The recursion uses \(nx=x+(n-1)x\). The base case is \(0x=0\).

Iterative power algorithm

Write an iterative algorithm that computes \(x^n\) using multiplication.

Answer steps
power(n, x):
    answer := 1
    for i := 1 to n
        answer := answer * x
    return answer

After one loop, \(answer=x\). After two loops, \(answer=x^2\). After \(n\) loops, \(answer=x^n\).

Recursive trace

Trace \(f(4)\) for the algorithm \(f(n)=n^2+f(n-1)\), with \(f(0)=0\).

Answer steps
\[f(4)=4^2+f(3)\] \[=16+3^2+f(2)\] \[=16+9+2^2+f(1)\] \[=16+9+4+1^2+f(0)\] \[=16+9+4+1+0=30.\]
Loop final values

For the loop below, determine the possible final values of \(y\).

while y < 0
    y := y + 1
Answer steps

If \(y\) starts negative, the loop adds 1 until \(y=0\). If \(y\) starts nonnegative, the loop does not run.

So the final value is either \(0\), or the original nonnegative value. In all cases, the final value satisfies \(y\ge0\).

Induction proof

Prove: \(3\mid n^3+2n\) for natural numbers \(n\).

Answer steps

Base case: \(1^3+2(1)=3\), so the statement is true for \(n=1\).

Assume \(3\mid k^3+2k\). This means \(k^3+2k=3a\) for some integer \(a\).

\[(k+1)^3+2(k+1)=k^3+3k^2+3k+1+2k+2\] \[=(k^3+2k)+3k^2+3k+3\] \[=(k^3+2k)+3(k^2+k+1).\]

The first part is divisible by 3 by the hypothesis. The second part has a visible factor of 3. So the sum is divisible by 3.

Chapter 6 - Counting

Factorial, permutation, combination
ObjectMeaningExample
\(n!\)arrange all \(n\) objects\(5!=5\cdot4\cdot3\cdot2\cdot1\)
\(P(n,r)\)choose \(r\) objects and arrange thempresident, VP, secretary
\(\binom nr\)choose \(r\) objects; order ignoredcommittee of \(r\) people
Lineups, strings, and officer roles usually care about order. Committees and subsets usually ignore order.
Password counting

How many passwords of length 15 use digits, uppercase letters, and lowercase letters and contain at least one uppercase and at least one lowercase?

Answer steps

Total characters: \(10+26+26=62\), so total passwords are \(62^{15}\).

Subtract passwords with no uppercase: \(36^{15}\). Subtract passwords with no lowercase: \(36^{15}\).

Passwords with only digits got subtracted twice, so add them back: \(10^{15}\).

\[62^{15}-36^{15}-36^{15}+10^{15}.\]
Pigeonhole principle

There are 10 AOCs and 20 teams. How many students guarantee at least 3 students in the same AOC and same team?

Answer steps

The boxes are AOC/team pairs:

\[10\cdot20=200.\]

To avoid 3 in one box, each box can have at most 2 students:

\[200\cdot2=400.\]

One more forces a box with 3 students:

\[401.\]
Gap method

A group has 7 men and 3 women. How many ways can they line up so that no two women are adjacent?

Answer steps

Arrange the 7 men first: \(7!\).

The men create 8 gaps:

_ M _ M _ M _ M _ M _ M _ M _

Choose 3 of those gaps for the women: \(\binom83\). Arrange the women: \(3!\).

\[7!\binom83 3!.\]
Next lexicographic permutations

Find the next three larger permutations after \(4173652\).

Answer steps

For the first next permutation, scan from the right:

\(6>5>2\), but \(3<6\). So \(3\) is the pivot.

To the right of 3, the smallest larger value is 5. Swap 3 and 5:

\[4173652\to4175632.\]

Sort the suffix after 5 in increasing order:

\[4175236.\]

Repeat the same rule:

\[4175236,\quad 4175263,\quad 4175326.\]
Path counting

How many paths go from \((0,0)\) to \((m,n)\) using only right and up moves?

Answer steps

Every path must have exactly \(m\) right moves and \(n\) up moves.

So each path is just an ordering of \(m\) R's and \(n\) U's. There are \(m+n\) total move-slots.

Choose which \(m\) of those slots are R moves. Once those slots are chosen, the other \(n\) slots must be U moves:

\[\binom{m+n}{m}.\]

Choosing the U slots instead gives the same answer:

\[\binom{m+n}{m}=\binom{m+n}{n}.\]

Chapter 7 - Probability

What the notation means
NotationMeaning
\(S\)sample space: all possible outcomes
\(P(A)\)probability that event \(A\) happens
\(P(A\mid B)\)probability that \(A\) happens after \(B\) is known to have happened
\(X\)random variable: a function that assigns a number to each outcome
\(E(X)\)expected value: weighted average of the values of \(X\)
\(\operatorname{Var}(X)\)variance: how spread out \(X\)'s values are
Core probability formulas
UseFormula
Equally likely outcomes\(P(E)=|E|/|S|\)
Conditional probability\(P(A\mid B)=P(A\cap B)/P(B)\)
Bayes\(P(A\mid B)=P(B\mid A)P(A)/P(B)\)
Independence\(P(A\cap B)=P(A)P(B)\), or \(P(A\mid B)=P(A)\)
Expected value\(E(X)=\sum_{s\in S}X(s)p(s)\)
Variance\(E(X^2)-[E(X)]^2\)
Conditional probability + independence

A fair die is rolled twice. Let \(A\) be the event that the sum is 5. Let \(B\) be the event that the first roll is even. Find \(P(A)\), \(P(A\mid B)\), and decide whether \(A\) and \(B\) are independent.

Answer steps

There are 36 total outcomes. Sum 5 happens for:

\[(1,4),(2,3),(3,2),(4,1).\] \[P(A)=4/36=1/9.\]

If \(B\) has happened, the first roll is one of \(2,4,6\). That gives \(3\cdot6=18\) possible outcomes.

Among those, the sum is 5 for:

\[(2,3),(4,1).\] \[P(A\mid B)=2/18=1/9.\]

Since \(P(A\mid B)=P(A)\), the events are independent.

Biased coin

A biased coin has heads twice as likely as tails. It is flipped three times. Find the probability of at least one tail.

Answer steps

Let \(P(T)=x\). Then \(P(H)=2x\). Since probabilities add to 1:

\[x+2x=1,\quad x=1/3.\]

So \(P(T)=1/3\) and \(P(H)=2/3\).

Use the complement. "At least one tail" is the opposite of "all heads."

\[P(\text{at least one tail})=1-P(HHH)=1-\left(\frac23\right)^3=\frac{19}{27}.\]
Bayes example

Suppose 5% of players use steroids. Users test positive 98% of the time. Non-users test positive 12% of the time. Find \(P(\text{user}\mid +)\).

Answer steps

Let \(D\) mean the player uses steroids.

The numerator is the probability of being a user and testing positive:

\[P(+\mid D)P(D)=0.98(0.05)=0.049.\]

The denominator is the total probability of testing positive:

\[P(+)=P(+\mid D)P(D)+P(+\mid\neg D)P(\neg D).\] \[P(+)=0.98(0.05)+0.12(0.95)=0.049+0.114=0.163.\]

Therefore:

\[P(D\mid +)=\frac{0.049}{0.163}\approx0.301.\]
Expected value + variance

A fair coin is flipped three times. Let \(X\) be the number of heads. Find \(E(X)\) and \(\operatorname{Var}(X)\).

Answer steps

The sample space is:

\[S=\{HHH,HHT,HTH,THH,HTT,THT,TTH,TTT\}.\]

Each outcome has probability \(1/8\). The values of \(X\) are:

\[3,2,2,2,1,1,1,0.\] \[E(X)=\frac{3+2+2+2+1+1+1+0}{8}=\frac32.\]

For variance, first compute \(E(X^2)\):

\[E(X^2)=\frac{9+4+4+4+1+1+1+0}{8}=3.\] \[\operatorname{Var}(X)=E(X^2)-[E(X)]^2=3-\left(\frac32\right)^2=\frac34.\]

Practice Problems

Chapter 1 - Logic

Truth table equivalence. Compare the final columns for \(p\to q\) and \(\neg p\lor q\). Are they equivalent?

Quantifier negation. Negate \(\forall x\exists y(P(x,y)\land Q(x,y))\) so that the negation is only on predicates.

Quantifier translation. Let \(F(x,y)\) mean "\(x\) can fool \(y\)." Write: everyone can be fooled by somebody.

Knights and knaves. A says, "B is a knight." B says, "A and I are different types." Determine A and B.

Proofs

Even proof. State the definition of even. Then prove: if \(n\) is even, then \(7n+10\) is even.

Counterexample. Disprove: the sum of two irrational numbers is always irrational.

Contradiction proof. Prove: if \(x\) is rational and \(y\) is irrational, then \(x+y\) is irrational.

Chapter 5 - Induction and Algorithms

Recursive formula. Let \(f(0)=1\) and \(f(n)=f(n-1)+3\). Find \(f(1),f(2),f(3)\), guess a formula, and prove it by induction.

Recursive algorithm. Write a recursive algorithm that computes \(x^n\) using multiplication, where \(n\) is a nonnegative integer.

Recursive trace. Trace \(f(3)\) for \(f(0)=1\) and \(f(n)=n+2f(n-1)\).

Loop final values. Suppose \(y\) is an integer. What are the possible final values after this loop?

while y < 3
    y := y + 1
Chapter 6 - Counting

Password counting. Passwords have length 8 and use digits and lowercase letters. How many contain at least one digit?

Pigeonhole. Students are classified by 4 graduation years and 21 majors. How many students guarantee two share both year and major?

Committee. From 8 men and 5 women, how many committees have exactly 3 men and 2 women?

Gap method. How many ways can 8 men and 5 women stand in a line if no two women stand next to each other?

Next permutation. Find the next larger permutation after \(3651742\).

Chapter 7 - Probability

Conditional probability. Two fair dice are rolled. Let \(A\) be "the sum is 7" and \(B\) be "the first roll is 3." Find \(P(A)\), \(P(A\mid B)\), and decide whether \(A\) and \(B\) are independent.

Biased coin. Heads is three times as likely as tails. The coin is flipped twice. Find the probability of at least one head.

Bayes. In a clinic, 8% of patients are infected. Infected patients test positive 98% of the time. Non-infected patients test positive 3% of the time. Find \(P(\text{infected}\mid +)\).

Expected value and variance. A box has tickets labeled \(1,2,2,3,4\). One ticket is chosen uniformly. Let \(X\) be the number on the ticket. Find \(E(X)\) and \(\operatorname{Var}(X)\).

Practice Problem Answers

Answer steps
Truth table equivalence
\(p\)\(q\)\(p\to q\)\(\neg p\lor q\)
TTTT
TFFF
FTTT
FFTT

The final columns match, so the propositions are equivalent.

Quantifier negation
\[\neg\forall x\exists y(P(x,y)\land Q(x,y))\] \[\equiv \exists x\forall y\neg(P(x,y)\land Q(x,y))\] \[\equiv \exists x\forall y(\neg P(x,y)\lor \neg Q(x,y)).\]

Flip each quantifier as the negation moves inward. Then use De Morgan's law on the AND statement.

Quantifier translation
\[\forall y\exists xF(x,y).\]

Read it as: for every person \(y\), there is some person \(x\) who can fool \(y\).

Knights and knaves

If A were a knight, then A's statement would make B a knight. But then B's statement would say they are different types, which would be false. So A is not a knight.

Therefore A is a knave. A's statement "B is a knight" is false, so B is also a knave. B's statement that they are different types is false, as required.

Answer: A and B are both knaves.

Even proof

An integer \(n\) is even if \(n=2k\) for some integer \(k\).

Assume \(n\) is even. Then \(n=2k\) for some \(k\in\mathbb Z\).

\[7n+10=7(2k)+10=14k+10=2(7k+5).\]

Since \(7k+5\) is an integer, \(7n+10\) is even.

Counterexample

Use one example where the hypothesis is true but the conclusion is false.

\[\sqrt2+(-\sqrt2)=0.\]

Both \(\sqrt2\) and \(-\sqrt2\) are irrational, but their sum is rational. So the statement is false.

Contradiction proof

Assume \(x\) is rational and \(y\) is irrational. Suppose, for contradiction, that \(x+y\) is rational.

Then \(y=(x+y)-x\). A rational number minus a rational number is rational, so this would make \(y\) rational.

That contradicts the assumption that \(y\) is irrational. Therefore \(x+y\) is irrational.

Recursive formula
\[f(1)=4,\qquad f(2)=7,\qquad f(3)=10.\]

The pattern adds 3 each time, starting at 1, so the formula is:

\[f(n)=3n+1.\]

Base case: \(f(0)=1=3(0)+1\).

Inductive step: assume \(f(k)=3k+1\). Then

\[f(k+1)=f(k)+3=(3k+1)+3=3(k+1)+1.\]

So the formula holds for all nonnegative integers \(n\).

Recursive algorithm
power(n, x):
    if n = 0 return 1
    else return x * power(n - 1, x)

The base case is \(x^0=1\). The recursive step uses \(x^n=x\cdot x^{n-1}\).

Recursive trace
\[f(3)=3+2f(2)\] \[=3+2(2+2f(1))\] \[=3+2(2+2(1+2f(0)))\] \[=3+2(2+2(1+2\cdot1))\] \[=3+2(2+6)=19.\]

So \(f(3)=19\).

Loop final values

If \(y<3\), the loop keeps adding 1 until \(y=3\). If \(y\ge3\), the loop does not run.

So the final value is either \(3\), or the original value if it was already at least \(3\). In every case, the final value satisfies \(y\ge3\).

Password counting

There are 36 possible characters total: 10 digits and 26 lowercase letters. So total passwords:

\[36^8.\]

Passwords with no digit use only lowercase letters:

\[26^8.\]

Subtract the no-digit cases:

\[36^8-26^8.\]
Pigeonhole

The boxes are year-major pairs:

\[4\cdot21=84.\]

With 84 students, it is possible to put one in each box. One more forces a repeated box.

\[84+1=85.\]
Committee

Choose the men and women separately. Order does not matter in a committee.

\[\binom83\binom52.\]
Gap method

Arrange the 8 men first:

\[8!.\]

They create 9 gaps around them:

_ M _ M _ M _ M _ M _ M _ M _ M _

Choose 5 of those gaps for the women, then arrange the 5 women:

\[8!\binom95 5!.\]
Next permutation

Start with \(3651742\). Scan from the right. The first place where a number is smaller than the number after it is \(1<7\), so \(1\) is the pivot.

To the right of 1, the smallest number bigger than 1 is 2. Swap 1 and 2:

\[3652741.\]

Then put the suffix after 2 in increasing order:

\[3652147.\]

So the next permutation is \(3652147\).

Conditional probability

There are 36 total dice outcomes. The sum is 7 for 6 outcomes:

\[(1,6),(2,5),(3,4),(4,3),(5,2),(6,1).\] \[P(A)=6/36=1/6.\]

Given \(B\), the first roll is already 3, so only the second die varies. To get sum 7, the second die must be 4.

\[P(A\mid B)=1/6.\]

Since \(P(A\mid B)=P(A)\), the events are independent.

Biased coin

Let \(P(T)=x\). Since heads is three times as likely as tails, \(P(H)=3x\).

\[x+3x=1,\qquad x=1/4.\]

So \(P(T)=1/4\) and \(P(H)=3/4\).

Use the complement: at least one head is the opposite of both flips being tails.

\[P(\text{at least one head})=1-P(TT)=1-\left(\frac14\right)^2=\frac{15}{16}.\]
Bayes

Let \(D\) mean infected. The numerator is infected and positive:

\[P(+\mid D)P(D)=0.98(0.08)=0.0784.\]

The denominator is total positive tests:

\[P(+)=0.98(0.08)+0.03(0.92)=0.0784+0.0276=0.106.\]

Therefore:

\[P(D\mid +)=\frac{0.0784}{0.106}\approx0.740.\]
Expected value and variance

The five equally likely ticket values are \(1,2,2,3,4\).

\[E(X)=\frac{1+2+2+3+4}{5}=\frac{12}{5}.\]

For variance, first compute \(E(X^2)\):

\[E(X^2)=\frac{1^2+2^2+2^2+3^2+4^2}{5}=\frac{34}{5}.\] \[\operatorname{Var}(X)=E(X^2)-[E(X)]^2=\frac{34}{5}-\left(\frac{12}{5}\right)^2=\frac{26}{25}.\]