MATH1061 — Week 5 Notes

Lecture 13 — Sequences and Mathematical Induction

Pre-work: Video 016 (Sequences) and Video 017 (Mathematical Induction). See sequences-and-summations for the notation and sum manipulations, and mathematical-induction for the proof template, the worked examples and the inequality rules. The worked questions are below.

Learning goals

  1. Gain fluency in the use of standard notation for sequences and summations.
  2. Understand how to apply the principle of mathematical induction to prove properties of summations and inequalities.

The first goal is about four notations: \((a_n)_{n \geq 0}\), \(\{a_n\}_{n=1}^{\infty}\), \(a_1, a_2, a_3, \dots\) and \(\sum_{n=1}^{m} a_n\).

Two questions about induction proofs

Do we have to write/know basis step, inductive hypothesis etc? It is a good idea to list these explicitly in your proofs to help the reader to follow the proof.

How does the LHS = RHS in the example from video 17? \(P(k)\) is the statement \(1 + 2 + \cdots + k = \frac{k(k+1)}{2}\), so \(P(k+1)\) is the statement \(1 + 2 + \cdots + k + (k+1) = \frac{(k+1)(k+2)}{2}\). Working from the left:

\[L.H.S = \underbrace{1 + 2 + \cdots + k}_{\text{arithmetic series}} + (k+1) = \frac{k(k+1)}{2} + (k+1) = \frac{k(k+1)}{2} + \frac{2(k+1)}{2} = \frac{(k+1)(k+2)}{2} = R.H.S\]

The step that does the work is replacing \(1 + 2 + \cdots + k\) — an arithmetic series — by the closed form the inductive hypothesis supplies.

Defining a sequence

Three ways: by an explicit formula, recursively from one or more preceding terms, or in words. Full definitions and examples in sequences-and-summations.

Sequences cannot be defined by listing their first few terms (unless you can list all of them). Asked for the next term after \(1, 4, 9, 16\): under \(a_n = n^2\) it is 25; under \(a_n = n^2 \bmod 20\) it is 5. Searching the Online Encyclopedia of Integer Sequences (https://oeis.org) for \(1, 4, 9, 16\) returns 397 possible sequences.

Working with sums

\[\sum_{i=m}^{n} a_i \pm \sum_{i=m}^{n} b_i = \sum_{i=m}^{n}(a_i \pm b_i) \qquad\qquad \sum_{i=m}^{n} c\,a_i = c\sum_{i=m}^{n} a_i\]

Worked: change of variable

(a) What expression results from applying the change of variable \(j = i+1\) to \(\displaystyle\sum_{i=0}^{n-1}(2i+5)\)?

With \(i = j - 1\):

  • lower limit: \(j = 0 + 1 = 1\)
  • upper limit: \(j = (n-1) + 1 = n\)
  • general term: \(2i + 5 = 2(j-1) + 5 = 2j + 3\)

\[\sum_{i=0}^{n-1}(2i+5) = \sum_{j=1}^{n}(2j+3)\]

(b) Use the change of variable from part (a) to simplify \(\displaystyle 3\sum_{j=1}^{n}(j^2+2j) + \sum_{i=0}^{n-1}(2i+5)\).

Rewriting the second sum over \(j\) puts both sums on the same index range, so they can be combined termwise:

\[\sum_{j=1}^{n} 3(j^2+2j) + \sum_{j=1}^{n}(2j+3) = \sum_{j=1}^{n}\Big(3(j^2+2j) + (2j+3)\Big) = \sum_{j=1}^{n}\big(3j^2 + 8j + 3\big)\]

Q4 from the pre-class multiple choice questions

The point of the question is splitting off the last term of a sum, which is done often when proving a summation property by mathematical induction:

\[\sum_{i=1}^{n+1}(3i+2) = \underbrace{(3\cdot1+2)}_{i=1} + \underbrace{(3\cdot2+2)}_{i=2} + \cdots + \underbrace{(3n+2)}_{i=n} + \underbrace{(3(n+1)+2)}_{i=n+1} = \sum_{i=1}^{n}(3i+2) + \big(3(n+1)+2\big)\]

Worked: three rewrites of one sum

Let \(k \in \mathbb{Z}\), \(k \geq 2\). Which of the following expressions is equal to \(\displaystyle\sum_{i=1}^{k+1}(2i-1)\)?

The restriction \(k \geq 2\) matters for option C: \(\sum_{i=1}^{k-1}(2i-1) + 4k\) only has something to sum once \(k - 1 \geq 1\).

All three of A, B and C. Each is the same sum described differently:

  • A. \(\displaystyle\sum_{j=0}^{k}(2j+1)\) — a change of variable, \(j = i - 1\). The limits shift from \(1 \dots k+1\) to \(0 \dots k\), and \(2i - 1 = 2(j+1) - 1 = 2j + 1\).
  • B. \(\displaystyle\sum_{i=1}^{k}(2i-1) + 2k + 1\) — split off the last term, \(i = k+1\), which is \(2(k+1) - 1 = 2k + 2 - 1 = 2k+1\).
  • C. \(\displaystyle\sum_{i=1}^{k-1}(2i-1) + 4k\) — split off the last two terms, \(i = k\) and \(i = k+1\), which together give \((2k-1) + (2k+1) = 4k\).

Written as sigma notation the three look like different objects. Written out as terms they are visibly one list, bracketed three ways (Figure 1). The split in B is the one worth memorising: it is the move Q4 is about — splitting off the last term — and the move every induction proof over a sum begins with.

Figure 1: One sum, three groupings — why all three options are correct. The boxed row is \(\sum_{i=1}^{k+1}(2i-1)\) written out. A keeps every term and only relabels the index (\(j = i-1\)); B brackets off the last term; C brackets off the last two and adds them, \((2k-1)+(2k+1) = 4k\). Nothing is added or removed in any of the three.

Mathematical induction

Let \(P(n)\) be a predicate in some integer \(n\). Given that (1) \(P(2)\) is true and (2) \(\forall k \in \mathbb{Z}\) with \(k \geq 2\), if \(P(k)\) then \(P(k+1)\) — then \(P(5)\) follows by applying (2) three times, and in fact \(P(n)\) is true for each integer \(n \geq 2\). The chain of modus ponens steps, and why the basis step is not optional, are in mathematical-induction.

The worked summation example — \(\displaystyle\sum_{i=1}^{n}(3i+2) = \frac{3n^2+7n}{2}\) for each integer \(n \geq 1\) — is written out there in full, with the four steps every induction proof takes: 1 define the predicate, 2 base case, 3 assume the inductive hypothesis \(P(k)\), 4 prove \(P(k+1)\).

Worked: a summation proof

Prove that for each integer \(n \geq 1\), \(\displaystyle\sum_{i=1}^{n}(2i-1) = n^2\).

Let \(P(n)\) be the predicate \(\displaystyle\sum_{i=1}^{n}(2i-1) = n^2\).

Basis step (\(n = 1\)). LHS of \(P(1)\): \(\displaystyle\sum_{i=1}^{1}(2i-1) = 2 - 1 = 1\). RHS of \(P(1)\): \(1^2 = 1\). Therefore \(P(1)\) is true.

Inductive step. Prove \(P(k) \to P(k+1)\) for all \(k \geq 1\) where \(k \in \mathbb{Z}\).

Suppose \(P(k)\) holds. Then \(\displaystyle\sum_{i=1}^{k}(2i-1) = k^2\).

LHS of \(P(k+1)\):

\[\sum_{i=1}^{k+1}(2i-1) = \sum_{i=1}^{k}(2i-1) + \big(2(k+1)-1\big) = k^2 + 2k + 1 \qquad \text{(by the inductive hypothesis } P(k)\text{)}\] \[= (k+1)^2\]

which is the RHS of \(P(k+1)\). Therefore \(P(k) \to P(k+1)\). //

Conclusion. Therefore, by the principle of mathematical induction, \(\displaystyle\sum_{i=1}^{n}(2i-1) = n^2\) for each integer \(n \geq 1\).

The // marks the end of the inductive step. The concluding sentence finishes the proof.

Mathematical induction to prove an inequality

You must be careful when manipulating inequalities — the full table of what can and cannot be concluded is in mathematical-induction. The two traps: multiplying by a negative flips the direction, and \(a < b\) together with \(x > y\) lets you conclude nothing about \(a + x\).

The example from the second video, \(n^2 < 2^n\) for each integer \(n \geq 5\), is worked through there as well.

Worked: an inequality proof

Prove that for each integer \(n \geq 2\), \(5^n + 9 < 6^n\).

Let \(P(n)\) be the predicate \(5^n + 9 < 6^n\).

Basis step (\(n = 2\)). LHS of \(P(2) = 5^2 + 9 = 25 + 9 = 34\); RHS of \(P(2) = 6^2 = 36\). Since \(34 < 36\), \(P(2)\) is true.

Inductive step. Prove \(P(k) \to P(k+1)\) for all \(k \geq 2\), \(k \in \mathbb{Z}\).

Assume \(P(k)\): \(5^k + 9 < 6^k\). The goal is \(P(k+1)\): \(5^{k+1} + 9 < 6^{k+1}\).

\[\text{LHS of } P(k+1) = 5^{k+1} + 9 = 5(5^k + 9) - 4 \times 9\] \[< 5(6^k) - 4 \times 9 \qquad \text{(by the inductive hypothesis } P(k)\text{)}\] \[< 6 \cdot 6^k - 4 \times 9 \qquad (5 < 6)\] \[< 6^{k+1} \qquad (-4 \times 9 < 0)\]

which is the RHS of \(P(k+1)\). Therefore \(P(k) \to P(k+1)\). //

Conclusion. Therefore, by the principle of mathematical induction, \(5^n + 9 < 6^n\) for each integer \(n \geq 2\).

The opening rewrite is the trick worth remembering: \(5(5^k + 9) - 4 \times 9 = 5^{k+1} + 45 - 36 = 5^{k+1} + 9\), which manufactures the exact expression \(5^k + 9\) that the inductive hypothesis bounds.

See also

Lecture 14 — Strong Mathematical Induction and the Well-Ordering Principle

Pre-work: Video 018 (Strong Mathematical Induction and the Well-Ordering Principle). See strong-induction-and-the-well-ordering-principle for the principles themselves, the MI/SMI comparison and the worked examples; the questions worked below apply them. mathematical-induction has the ordinary-induction template both methods share.

Learning goals

  1. Gain fluency in applying the principles of mathematical induction and strong mathematical induction.
  2. Further your understanding of how the basis step relates to the inductive step in mathematical induction and strong mathematical induction.
  3. Appreciate the equivalence between mathematical induction, strong mathematical induction and the well-ordering principle for the integers.

Base cases, least elements and equivalence

Why does pre-class Q1 need 2 values for the basis step, not 3? Because the recurrence in that question, \(b_n = b_{n-2} + b_{n-1}\), reaches back exactly two terms — so the inductive step consumes \(P(k-1)\) and \(P(k)\), and the basis step has to cover \(P(1)\) and \(P(2)\) for the chain to have somewhere to start. Worked in full below. Contrast the sequence \(a_n = a_{n-1} - a_{n-3}\) further down, whose recurrence reaches back three terms and therefore needs three base cases.

In the well-ordering principle — non-empty set, fixed integer, a least element — is the least element an integer in the set or one outside it? In the set — a least element of \(S\) is a member of \(S\) that is \(\leq\) every member of \(S\), not a bound sitting outside it. (A lower bound is allowed to sit outside: the “fixed integer” in the statement is any integer below everything in \(S\), and it need not belong to \(S\) — \(-1\) plays that role for the set of squares below. The two ideas are easy to run together, and the difference is the whole content of the principle: bounds are cheap, a least element is what the WOP promises.) The terminology is unpacked in strong-induction-and-the-well-ordering-principle; four sets are tested against the principle below and drawn together in one figure.

Why are mathematical induction, strong mathematical induction and the well-ordering principle for the integers all equivalent? Anything provable by one is provable by the other two — SMI becomes MI by rewording the predicate to accumulate every earlier case (Epp Q27 p. 279), and MI becomes the WOP by contradiction on the set of counterexamples (Epp Q31 p. 279). One example is proved all three ways; see the same note.

Mathematical induction vs strong mathematical induction

The two differ only in what the inductive step may assume, and correspondingly in how much the basis step must establish. Given the basis \(P(2) \wedge P(3) \wedge P(4)\) and the SMI inductive step \(\forall k \geq 4\), \(\big(P(2) \wedge \cdots \wedge P(k)\big) \to P(k+1)\), we reach \(P(6)\) in two applications — and in fact \(P(n)\) for each integer \(n \geq 2\), exactly as ordinary induction does from \(P(2)\) alone. Side-by-side in strong-induction-and-the-well-ordering-principle.

Worked: \(4 \mid b_n\) (pre-class Q1)

Suppose \(b_1, b_2, b_3, \dots\) is defined by \(b_1 = 4\), \(b_2 = 12\), and \(b_n = b_{n-2} + b_{n-1}\) for each integer \(n \geq 3\). Prove that \(4 \mid b_n\) for each integer \(n \geq 1\).

The recurrence reaches back two terms, so this needs two base cases. The proof is by PSMI. The first few terms:

\[b_3 = b_1 + b_2 = 4 + 12 = 16 \qquad b_4 = b_2 + b_3 = 12 + 16 = 28\]

The full proof by strong induction is in strong-induction-and-the-well-ordering-principle. The point of the example: the inductive step uses both \(P(k)\) and \(P(k-1)\), which is exactly what ordinary induction cannot supply.

Worked: a recurrence that reaches back three terms

Suppose \(a_0, a_1, a_2, \dots\) is defined by \(a_0 = 6\), \(a_1 = 9\), \(a_2 = 12\) and \(a_n = a_{n-1} - a_{n-3}\) for each integer \(n \geq 3\).

(a) Determine the value of \(a_4\).

\(a_4\) cannot be reached in one step, because the recurrence at \(n = 4\) asks for \(a_3\), which has not been computed yet. Two steps:

\[a_3 = a_2 - a_0 = 12 - 6 = 6 \qquad a_4 = a_3 - a_1 = 6 - 9 = -3\]

The sequence is not monotone and not positive — \(6, 9, 12, 6, -3, \dots\) — so nothing about “divisible by 3” can be read off a pattern of growth. It has to be proved.

(b) In a proof by strong mathematical induction that \(3 \mid a_n\) for each integer \(n \geq 0\), what would you need to prove in the basis step?

With \(P(n)\) the predicate \(3 \mid a_n\), the recurrence reaches back three terms, so the basis step needs three cases:

\[P(0):\; 3 \mid a_0 \qquad P(1):\; 3 \mid a_1 \qquad P(2):\; 3 \mid a_2\]

The count is read straight off the recurrence: the term \(a_{n-3}\) means 3 cases. It is the deepest back-reference that sets the number, not the number of terms on the right-hand side — strong-induction-and-the-well-ordering-principle has a figure contrasting this recurrence with the one for \(b_n\) above.

This is the sequence used below to compare three separate proofs of \(3 \mid a_n\) — by SMI, by MI, and by the WOP.

Every integer \(n > 1\) is a product of primes

Proved by strong induction, splitting on whether \(k+1\) is prime or composite; written out in strong-induction-and-the-well-ordering-principle. Strong induction is required rather than ordinary induction because \(r\) and \(s\) in the factorisation \(k+1 = rs\) can be any integers from 2 to \(k\) — the proof has no control over which, so it needs the inductive hypothesis at all of those values.

The prose proof leaves the induction machinery implicit. It is worth writing each part out in terms of a predicate:

Part of the proof In terms of \(P\)
the theorem \(P(n)\): \(n\) can be written as a product of primes, for \(n \geq 2\) (the same as \(n > 1\))
the basis step \(P(2)\) ✓
the inductive step \(\big(P(2) \wedge P(3) \wedge \cdots \wedge P(k)\big) \to P(k+1)\)
“\(1 < r < k+1\) and \(1 < s < k+1\)” \(P(r) \wedge P(s)\), both taken from the conjunction above
“In either case…” \(P(2) \wedge \cdots \wedge P(k) \to P(k+1)\) again, closing the step

Naming the predicate first, then writing out the exact implication the inductive step owes you, is what turns the prose proof into the course’s induction skeleton.

The well-ordering principle for the integers

Let \(S\) be a non-empty set of integers, each of which is greater than some fixed integer. Then \(S\) has a least element.

Three conditions. The four sets below test them — one per condition, plus one case where they all hold. The one that is easiest to forget is “of integers”: a set can be non-empty and bounded below and still have no least element, as long as it is not a set of integers. The set of positive rationals below is exactly that case.

Worked: the integers congruent to \(1 \bmod 4\) (pre-class Q2)

True or False? The set of all integers congruent to \(1 \bmod 4\) has a least element.

False. The set is \(\{\dots, -7, -3, 1, 5, 9, \dots\}\).

Suppose this set did have a least element, call it \(x\). Then, since \(x \equiv 1 \pmod 4\), we have \(x - 4 \equiv 1 \pmod 4\) — so \(x-4\) is also in the set. But \(x-4\) is smaller than \(x\), so we have a contradiction.

The middle clause is the step the contradiction turns on: \(x-4\) being smaller only matters once you know it’s in the set.

The set is non-empty, but not bounded below — so the WOP never applied to it in the first place.

Worked: which of these sets has a least element?

For each statement, decide if it is true or false, and how you would demonstrate that it is true or false.

(a) The set of positive rational numbers has a least element.

False — proof by contradiction.

Suppose \(\frac{p}{q}\) is the smallest possible positive rational, where \(p, q \in \mathbb{Z}^+\). Then

\[0 < \frac{1}{2}\cdot\frac{p}{q} < \frac{p}{q}\]

So \(\frac{p}{2q}\) is a smaller positive rational — contradiction.

Halving is the useful move here: it produces a strictly smaller element that is still in the set, which is exactly what “no least element” needs. The trick is worth remembering on its own, because “multiply by a factor strictly between 0 and 1” is the standard way to defeat a claimed least element in a dense set.

\(\mathbb{Q}^+\), the set of positive rationals, is non-empty and is bounded below — by 0. It satisfies two of the WOP’s three conditions and still has no least element. What it fails is “set of integers”: between \(\tfrac{p}{q}\) and 0 there is always more room, whereas between an integer \(x\) and a fixed lower bound there is not. So (a) is not the “not bounded below” failure that (b) and the mod-4 set are; it is a different failure entirely, and it is the reason the principle is called the well-ordering principle for the integers.

(b) Every non-empty subset of the integers has a least element.

False — take \(S = \mathbb{Z}^-\); the proof is by contradiction. (A contradiction can be marked inline with a crossed X, a lightning stroke or \(\bot\); the three are alternatives — see math1061-diagram-notation.) \(\mathbb{Z}^-\) is non-empty and made of integers, but it has no least element. This is precisely the “greater than some fixed integer” hypothesis of the WOP failing: \(\mathbb{Z}^-\) is not bounded below.

(c) Every non-empty subset of the set of squares \(x^2\) where \(x \in \mathbb{Z}\) has a least element.

True.

Let \(S \subseteq \mathbb{Z}\) with \(S \neq \emptyset\) be a set of such squares. For all \(x \in \mathbb{Z}\) we have \(0 \leq x^2 \in \mathbb{Z}\), so \(-1 < x^2\) for every \(x^2 \in S\). So by the WOP, \(S\) has a least element.

The fixed integer that bounds \(S\) below is \(-1\) (or 0) — squares cannot be negative, which is what makes the WOP applicable here and not in (b). The proof does not have to say which element is least, or produce it. The WOP is an existence statement, and the whole argument is the three-line check that its hypotheses hold.

Figure 2: Four sets tested against the well-ordering principle, and which of its conditions each one fails. A set has no least element exactly when some move always produces a smaller member still inside it — subtract 4, halve, subtract 1 — and each disproof is the exhibition of that move. The rows are drawn to scale: the mod-4 and negative-integer sets march off to the left forever, while the positive rationals stay bounded below by 0 and fail for a different reason, having no integrality to stop them accumulating. Open circles mark numbers not in the set. Each hop arrow is labelled by its operation.

MI, SMI and the WOP on one example

The claim \(3 \mid a_n\), for the sequence \(a_n = a_{n-1} - a_{n-3}\) above, can be proved three times over — by SMI, by MI on the reworded predicate \(Q(n) \equiv P(0) \wedge P(1) \wedge \cdots \wedge P(n)\), and by the WOP taking the least counterexample. All three proofs share the same algebraic core, \(a_{k+1} = a_k - a_{k-2} = 3x - 3y = 3(x-y)\), and in the SMI and MI proofs that line is identical. They differ only in how the earlier cases are made available. All three are written out in strong-induction-and-the-well-ordering-principle.

Three details of those proofs, all recorded in that note:

  • The MI proof’s basis step is \(n = 2\), not \(n = 0\), because \(Q(2)\) is the statement that packs all three of \(3 \mid a_0\), \(3 \mid a_1\), \(3 \mid a_2\) into one predicate.
  • The MI proof’s last step concludes \(Q(k+1)\), not \(P(k+1)\). An induction run on \(Q\) must conclude \(Q(k+1)\).
  • The WOP proof’s “least element \(k\)” is the least counterexample, which is the same information the strong inductive hypothesis carries, obtained a different way.

Worked: where the prime-factorisation proof uses the WOP

How does the following proof use the well-ordering principle for the integers?

Theorem: every integer \(n > 1\) can be written as a product of primes.

Proof: suppose the theorem is false. Then there exists an integer \(n > 1\) that is not a product of primes. Choose the smallest such number \(n\). Since \(n > 1\), then either \(n\) is prime or \(n\) is composite.

If \(n\) is prime, then \(n\) is a product of primes.

If \(n\) is composite, then \(n = rs\) for some positive integers \(r \neq 1\) and \(s \neq 1\). This implies that \(1 < r < n\) and \(1 < s < n\). Because we chose \(n\) to be the smallest integer that is not a product of primes, both \(r\) and \(s\) (which are smaller) are products of primes. Let \(r = p_1p_2\dots p_a\) and \(s = q_1q_2\dots q_b\) where the \(p_i\) and \(q_j\) are primes, and so \(n = rs = p_1p_2\dots p_a q_1q_2\dots q_b\) is a product of primes also.

In either case, \(n\) is a product of primes, contradicting our choice of \(n\). Therefore every integer \(n > 1\) can be written as a product of primes.

At the step “choose the smallest such number \(n\)” — that is the well-ordering principle, and nothing else in the proof invokes it. Written in full, the step reads “By the WOP we can choose the smallest such number \(n\)”. The proof has just established that the set of counterexamples is non-empty, and every member is an integer greater than 1, so the set is bounded below; the WOP is what licenses picking a least element out of it. Everything downstream (“because we chose \(n\) to be the smallest…”) depends on that element existing.

Figure 3: The well-ordering proof that every integer n > 1 is a product of primes, drawn on the number line: the well-ordering principle is used once, and everything after it lives off that one step. The green stretch is not proved — it holds only because n was chosen smallest, and the whole proof consists of showing that the two factors of a composite n land inside it. Compare the same shape in the well-ordering proof of 3 | a_n: least counterexample, a stretch below it that is known to be fine, and a descent that lands inside that stretch. The r and s ticks are placed arbitrarily inside the range, since the argument fixes only that 1 < r, s < n.

Set this next to the strong-induction proof of the same theorem above and the two are the same argument turned inside out. Strong induction builds up: assume \(P(2), \dots, P(k)\), prove \(P(k+1)\). The WOP version descends: assume some \(n\) fails, take the least such, and find that its factors — which are smaller, so they do not fail — hand you the very factorisation \(n\) was supposed to lack. The green stretch in the figure is doing the job that “\(P(2) \wedge \cdots \wedge P(k)\)” does in the induction.

Mathematical induction to prove a divisibility property

Recall that for integers \(n\) and \(d\) with \(d \neq 0\), \(d \mid n\) if and only if there exists some integer \(a\) such that \(n = da\) (see divisibility-and-factorisation).

Worked: \(5 \mid (7^n - 2^n)\)

Prove that for each integer \(n \geq 0\), \(5 \mid (7^n - 2^n)\).

Let \(P(n)\) be the predicate \(5 \mid (7^n - 2^n)\).

Basis step (\(P(0)\)). When \(n = 0\), \(7^n - 2^n = 7^0 - 2^0 = 1 - 1 = 0 = 5 \times 0\). Therefore \(5 \mid 0\), and so \(P(0)\) is true.

Inductive step (\(P(k) \to P(k+1)\)). Suppose \(P(k)\) is true for some \(k \geq 0\). Then \(5 \mid (7^k - 2^k)\), so there exists \(b \in \mathbb{Z}\) such that \(7^k - 2^k = 5b\). Thus \(7^k = 5b + 2^k\).

Consider \(P(k+1)\). Then

\[\text{LHS of } P(k+1) = 7^{k+1} - 2^{k+1} = 7\big(7^k\big) - 2\big(2^k\big)\] \[= 7\big(5b + 2^k\big) - 2\big(2^k\big) \qquad \text{(by the inductive hypothesis } P(k)\text{)}\] \[= 5(7b) + (7-2)2^k = 5\big(7b + 2^k\big)\]

Therefore \(5 \mid (7^{k+1} - 2^{k+1})\), so \(P(k) \to P(k+1)\). \(/\!/\)

Conclusion. By the principle of mathematical induction, \(5 \mid (7^n - 2^n)\) for each integer \(n \geq 0\).

The idea: separate \(7^k\) and \(2^k\) out of \(7^{k+1} - 2^{k+1}\), then use the inductive hypothesis to get rid of the \(7^k\). Replacing it with \(5b + 2^k\) is what makes a factor of 5 appear; the two \(2^k\) terms then collect into \((7-2)2^k\), which is another multiple of 5.

Two things about the shape of this one:

  • Ordinary induction, not strong. Nothing here is a recurrence reaching back several terms — \(P(k+1)\) is built out of \(P(k)\) alone — so one base case is all it needs. It is the contrast to everything above it.
  • The end mark and the conclusion. The double slash \(/\!/\) closes the inductive step at “\(P(k) \to P(k+1)\)”. The Conclusion line that follows, “by the principle of mathematical induction”, completes the course’s induction skeleton. (\(/\!/\) is an end-of-proof mark; \(\square\) and no mark at all are also used — see math1061-diagram-notation, which records four options as sanctioned.)

See also

Reference material

Divisibility & Unique Factorisation

Definition

\[d \mid n \iff d \neq 0 \text{ and } n = dk \text{ for some integer } k\]

Both \(n\) and \(d\) must be integers, and \(d \neq 0\) — this side condition is part of the definition, not an afterthought. Equivalently, \(d \mid n \iff \frac{n}{d} \in \mathbb{Z}\).

All of these say the same thing:

  • \(d\) divides \(n\)
  • \(d\) is a divisor of \(n\)
  • \(d\) is a factor of \(n\)
  • \(n\) is a multiple of \(d\)

\(d \nmid n\) means \(d\) does not divide \(n\). Note that \(d \mid n\) is a statement, not a number: \(2 \mid 6\) is true, it is not “\(2\) divided by \(6\)” and it is not \(3\). Likewise \(10 \nmid 5\) is true, since \(\frac{5}{10} \notin \mathbb{Z}\). Being a statement is what makes \(2 \mid x\) a predicate — it acquires a truth value once \(x\) is fixed — so it is exactly the kind of thing a quantifier can be put in front of.

Divisibility inside a quantified statement

Two quantified divisibility statements, both false, both settled by a single counterexample — one counterexample is enough to make a \(\forall\) statement false, though no single example could ever make one true (see quantified-statements on that asymmetry):

Statement Verdict Counterexample
\(\forall x \in \mathbb{Z}\), if \(2 \mid x\) then \(4 \mid x\) False \(x = 2\): \(2 \mid 2\) is true but \(4 \nmid 2\)
\(\forall x \in \mathbb{Z}\), if \(6\) is divisible by \(x\), then \(x = 2\) False \(x = 3\): \(3 \in \mathbb{Z}\) and \(3 \mid 6\), but \(3 \neq 2\)

A counterexample to a universal conditional has to make the hypothesis true and the conclusion false — that is what \(\sim(p \to q) \equiv p \land \sim q\) says. An \(x\) that fails the hypothesis proves nothing.

The second statement becomes true once the conclusion lists every divisor:

\[\forall x \in \mathbb{Z}, \text{ if } 6 \text{ is divisible by } x, \text{ then } x = \pm 1 \text{ or } \pm 2 \text{ or } \pm 3 \text{ or } \pm 6.\]

The \(\pm\) is not optional. The reason is in the definition: it asks only for \(n = dk\) with \(k \in \mathbb{Z}\), and \(k\) may be negative, so divisors over \(\mathbb{Z}\) always come in \(\pm\) pairs — if \(n = dk\) then \(n = (-d)(-k)\), so \(d \mid n\) exactly when \((-d) \mid n\). A divisor list that quietly restricts itself to \(\mathbb{Z}^+\) is answering a question about a different domain.

Proving and disproving divisibility statements

The pattern for a direct proof: unfold \(d \mid n\) into \(n = dk\), do algebra, then re-fold into the definition by exhibiting the integer witness.

\(\forall a, b, c \in \mathbb{Z}\), if \(c \mid a\) then \(c \mid ab\).

Let \(a, b, c \in \mathbb{Z}\) and suppose \(c \mid a\). Then \(\exists k \in \mathbb{Z}\) such that \(a = ck\). Multiplying both sides by \(b\): \(ab = ckb = c(kb)\). Since \(k, b \in \mathbb{Z}\) we have \(kb \in \mathbb{Z}\). Thus \(c \mid ab\). \(\square\)

\(\forall m \in \mathbb{Z}\), \(6m(2m + 10m^2)\) is divisible by 4.

Let \(m \in \mathbb{Z}\). Then \(6m(2m + 10m^2) = 2(3m) \cdot 2(m + 5m^2) = 4 \big(3m(m + 5m^2)\big)\). Since \(m \in \mathbb{Z}\), \(3m(m + 5m^2) \in \mathbb{Z}\). Thus \(4 \mid 6m(2m + 10m^2)\). \(\square\)

\(\forall a, b, c \in \mathbb{Z}\), if \(a \mid (b + c)\) then \(a \mid b\) or \(a \mid c\) — false. Counterexample \(a = 3, b = 4, c = 2\): \(a \mid (b+c)\) since \(3 \mid 6\), but \(3 \nmid 4\) and \(3 \nmid 2\).

\(\forall a, b, m \in \mathbb{Z}\), if \(m \mid (a+b)\) then \(m \mid (a-b)\) — false. A counterexample needs the hypothesis true and the conclusion false:

\(m \mid (a+b)\) \(m \mid (a-b)\) conditional
\(m=5, a=20, b=10\) \(5 \mid 30\) T \(5 \mid 10\) T T
\(m=5, a=21, b=13\) \(5 \mid 34\) F \(5 \mid 8\) F T
\(m=5, a=27, b=7\) \(5 \mid 34\) F \(5 \mid 20\) T T
\(m=5, a=19, b=-9\) \(5 \mid 10\) T \(5 \mid 28\) F F
\(m=5, a=23, b=12\) \(5 \mid 35\) T \(5 \mid 11\) F F
\(m=5, a=-10, b=-3\) \(5 \mid -13\) F \(5 \mid -7\) F T

Only the two rows with hypothesis T and conclusion F are counterexamples.

\(\forall m, n, q \in \mathbb{Z}\), if \(m \mid n\) and \(n \mid q\) then \(m^2 \mid nq\) — true.

Let \(m, n, q\) be non-zero integers with \(m \mid n\) and \(n \mid q\). By definition there exist \(a, b \in \mathbb{Z}\) with \(n = ma\) and \(q = nb\). Then \[nq = (ma)(nb) = ma \cdot (mab) = m^2(a^2b).\] Since \(a^2 b \in \mathbb{Z}\), we conclude \(m^2 \mid nq\). \(\square\)

If \(k\) is even and \(m\) is odd, then \((k+2)^2 - (m-3)^2\) is divisible by 4.

Let \(k = 2a\) and \(m = 2b+1\) for some \(a, b \in \mathbb{Z}\). Then \((k+2)^2 = (2a+2)^2 = 4(a+1)^2\) and \((m-3)^2 = (2b-2)^2 = 4(b-1)^2\), so \[(k+2)^2 - (m-3)^2 = 4\big((a+1)^2 - (b-1)^2\big),\] and \((a+1)^2 - (b-1)^2 \in \mathbb{Z}\). \(\square\)

\(\forall c, d, e \in \mathbb{Z}\), if \(c \mid d\) and \(c \nmid e\), then \(c \nmid (d+e)\).

Proof by contradiction. Suppose there exist integers \(c, d, e\) with \(c \mid d\), \(c \nmid e\) and \(c \mid (d+e)\). Then \(d = cx\) for some \(x \in \mathbb{Z}\), and \(d + e = cy\) for some \(y \in \mathbb{Z}\). Hence \(cx + e = cy\), so \(e = c(y - x)\) with \(y - x \in \mathbb{Z}\) — so \(c \mid e\), a contradiction. \(\square\)

Every integer \(n > 1\) is a product of primes

Proof by contradiction. Suppose the theorem is false. Then there exists an integer \(n > 1\) that is not a product of primes. Choose the smallest such \(n\) — such a smallest one exists by the well-ordering principle. Choosing the smallest is what lets us assume every integer strictly between \(1\) and \(n\) is a product of primes.

Every integer \(n > 1\) is either prime or composite (not both).

Case 1: \(n\) is prime. Then \(n\) is a product of primes (namely itself) — contradiction.

Case 2: \(n\) is composite. Then \(n = rs\) for positive integers \(r, s\) with \(r \neq 1\) and \(s \neq 1\), which forces \(1 < r < n\) and \(1 < s < n\). Because \(n\) was chosen smallest, both \(r\) and \(s\) are products of primes, say \(r = p_1 p_2 \cdots p_k\) and \(s = q_1 q_2 \cdots q_\ell\). Therefore \(n = rs = p_1 \cdots p_k q_1 \cdots q_\ell\) is a product of primes too.

Either way \(n\) is a product of primes, contradicting our choice of \(n\). \(\square\)

Unique Factorisation Theorem

Provided in the exam. This theorem and the Quotient-Remainder Theorem are both stated on the MATH1061/MATH7861 Examination Formula Page, so they can be cited rather than restated from memory — see also logical-equivalence-laws for the rest of what that page carries.

Given any integer \(n > 1\), the standard factored form of \(n\) is \[n = p_1^{e_1} p_2^{e_2} \cdots p_k^{e_k}\] where \(k \in \mathbb{Z}^+\), the \(p_i\) are primes, the \(e_i\) are positive integers, and \(p_1 < p_2 < \cdots < p_k\).

Both the primes and their exponents are unique — this is what makes the factored form a canonical fingerprint of an integer, and it’s the engine behind divisor counting, gcd and lcm (see gcd-lcm-and-euclidean-algorithm).

Counting divisors from the factorisation

Any positive divisor of \(n = p_1^{e_1} \cdots p_k^{e_k}\) has the form \(p_1^{a_1} \cdots p_k^{a_k}\) with \(0 \leq a_i \leq e_i\), so there are

\[(e_1 + 1)(e_2 + 1) \cdots (e_k + 1)\]

positive divisors.

\(6975 = 3^2 \cdot 5^2 \cdot 31\), found by trial division: \(6975 = 3 \cdot 2325 = 3 \cdot 3 \cdot 775 = 3 \cdot 3 \cdot 5 \cdot 155 = 3^2 \cdot 5^2 \cdot 31\). A divisor has \(0\), \(1\) or \(2\) threes; \(0\), \(1\) or \(2\) fives; and \(0\) or \(1\) thirty-ones — so \(3 \cdot 3 \cdot 2 = 18\) positive divisors.

\(27720 = 2^3 \cdot 3^2 \cdot 5 \cdot 7 \cdot 11\).

\(1400 = 2^3 \cdot 5^2 \cdot 7\), worked through:

  • \(1400 = 2(700) = 2 \cdot 2(350) = 2 \cdot 2(35)(10) = 2^2 (5 \cdot 7)(2 \cdot 5) = 2^3 \cdot 5^2 \cdot 7^1\) — note the primes are written in increasing order.
  • \(1400^2 = (2^3 \cdot 5^2 \cdot 7)^2 = 2^6 \cdot 5^4 \cdot 7^2\) — raising to a power multiplies the exponents.
  • \(1400^3 = 1400 \cdot 1400^2 = 2^9 \cdot 5^6 \cdot 7^3\) — multiplying adds the exponents.
  • Number of positive divisors of \(1400\): \((3+1)(2+1)(1+1) = 24\).
  • Divisors of \(1400\) divisible by \(8\): all factors of the form \(2^3 \cdot 5^* \cdot 7^*\), i.e. \(8, 40, 200, 56, 280, 1400\).
  • Smallest \(n\) with \(1400n\) a perfect square: \(1400n = t^2\) needs every exponent even. \(t^2 = 2^3 \cdot 5^2 \cdot 7^1 \cdot n\), and splitting each power in half (bigger halves on the left) gives \(t = 2^2 \cdot 5^1 \cdot 7^1\) against \(2^1 \cdot 5^1 \cdot 7^0 \cdot n\), so \(n = 2 \cdot 7 = 14\).

Primality testing

Lemma: for every integer \(n > 1\), if no integer \(x\) in the range \(1 < x \leq \sqrt{n}\) divides \(n\), then \(n\) is prime.

This is why trial division only has to go as far as \(\sqrt{n}\) — for \(6975\), only up to \(\sqrt{6975} = 83.51\dots\)

Proof by contraposition. The contrapositive is: \(\forall n \in \mathbb{Z}^{\geq 2}\), if \(n\) is not prime, then there exists an integer \(x\) with \(1 < x \leq \sqrt{n}\) such that \(x \mid n\).

Suppose \(n \in \mathbb{Z}^{\geq 2}\) and \(n\) is not prime. Then \(n\) is composite, so by definition \(\exists r, s \in \mathbb{Z}^+\) with \(n = rs\) and \(1 < r < n\) and \(1 < s < n\).

Case 1: \(r = s\). Then \(r = s = \sqrt{n}\).

Case 2: \(r \neq s\). Then one of \(r, s\) is less than \(\sqrt{n}\) and the other greater — if both were greater we’d have \(rs > n\), and if both were less we’d have \(rs < n\).

In either case there is an \(x \in \{r, s\}\) with \(1 < x \leq \sqrt{n}\) and \(x \mid n\). \(\square\)

There are infinitely many primes

Lemma: for any integer \(a\) and any prime \(p\), if \(p \mid a\) then \(p \nmid (a+1)\).

Proof by contradiction. Let \(a \in \mathbb{Z}\), let \(p\) be prime, suppose \(p \mid a\), and assume for a contradiction that \(p \mid (a+1)\). Since \(p \mid a\) there exists \(k \in \mathbb{Z}\) with \(a = pk\); since \(p \mid (a+1)\) there exists \(\ell \in \mathbb{Z}\) with \(a + 1 = p\ell\). So \[1 = (a+1) - a = p\ell - pk = p(\ell - k),\] and \(\ell - k \in \mathbb{Z}\), so \(p \mid 1\). But \(p \geq 2\) for any prime, whereas \(p \mid 1\) forces \(p = \pm 1\) — a contradiction. \(\square\)

Theorem: there are an infinite number of primes.

Proof by contradiction. Suppose the number of primes is finite. Then some prime \(p\) is the largest, and we can list every prime in ascending order: \(2, 3, 5, 7, 11, \dots, p\).

Let \(N = (2 \cdot 3 \cdot 5 \cdot 7 \cdot 11 \cdots p) + 1\).

Then \(N > 1\), so by the Unique Factorisation Theorem \(N\) is a product of primes and hence divisible by some prime \(q\). Because \(q\) is prime it must be one of \(2, 3, 5, 7, 11, \dots, p\).

So \(q\) divides \(2 \cdot 3 \cdot 5 \cdots p\) — that is, \(q \mid (N - 1)\) — and hence by the lemma \(q \nmid \big((2 \cdot 3 \cdot 5 \cdots p) + 1\big) = N\).

So \(q\) divides \(N\) and \(q\) does not divide \(N\): a contradiction. Therefore the original statement is true. \(\square\)

See also

Diagram and Notation Conventions

The house style for every figure, symbol and proof layout in MATH1061 — read off the course’s own lecture slides rather than assumed. This is the authority any new diagram for this course is checked against. Where the source does not settle something, that is said so explicitly rather than filled in.

What was read

Everything the course ships in the anchor set, in full: all 176 slide pages across the 18 lecture decks (L01–L18), plus the 1-page examination formula sheet. No sampling — the whole corpus was viewed as rendered images so the drawings, ink colours and box shapes were seen, not inferred from extracted text.

Citations below use L07 p2 = PDF page 2 of math1061-lecture-07-…pdf. Two decks have a page-label offset worth knowing: L03 p2–p3 carry the printed header Lecture 2, and L17 p9–p10 carry the printed header Lecture 18. PDF page position is what is cited.

The course ships one written reference — the MATH1061/MATH7861 Examination Formula Page. It is a genuine notation authority (it fixes ∼, ≡, bold t/c, Aᶜ, ∴, A − B, and the numbering of the set identities), but it is not a drawing guide: it says nothing about figures, colour, layout or symbol shape. Everything in the Diagrams section below is inferred from the decks alone.

Only 18 of the course’s 36 lectures exist in this set (L19–L36 — one-to-one/onto functions, composition, cardinality, relations, equivalence and partial orders, groups, fields, counting, probability, pigeonhole, graph theory, walks, matrix representations, trees — are absent). Conventions for those topics are unestablished; see the last section.

Provenance: two lecturers, one template

The course is split between two lecturers (L01 p12, course schedule): Sara Davies takes L01–L09 and L18–L26; Katie Clinch takes L10–L17 and L27–L36. Both use the same slide frame, but their annotation habits and some box conventions differ. Where they diverge it is recorded under Inconsistencies, not resolved.

Textbook lineage

The recommended text is Susanna Epp, Discrete Mathematics with Applications, 4th or 5th (metric) edition (L01 p3, with the cover image on the slide). It is cited repeatedly by page for statements and exercises — L14 p7 (“Q27 on page 279 of Epp”), L14 p8 (“Q31 on page 279”), L16 p12 (“page 310 (in 4th edition) or page 346 (in 5th edition)”), L18 p8 (“pages 419-421 (ed. 5) or pages 378-380 (ed. 4)”).

Important qualification: no slide anywhere carries a caption naming a textbook as the source of a figure. The Epp lineage is real for wording and content — the logical-equivalence names, the valid-argument-form names and the twelve numbered set identities are Epp’s — but it is not documented evidence for any drawing style. Do not cite Epp as a figure source.

Slide frame and page furniture

Element Convention Evidence Holds?
Header band Page 1 of each deck: MATH1061/MATH7861 Discrete Mathematics Semester 2, 2026 at far left, Lecture N – Title at far right. Later pages: right-hand lecture label only. A full-width thin black rule sits beneath. L02 p1 vs L02 p2; identical in every deck L02–L18 Every deck except L01, which has UQ-branded title slides
Pre-work box A purple rounded-rectangle box, top right of page 1, headed Pre-work video: / Pre-work videos: in bold, listing Video NNN Title with three-digit zero-padded numbers L02 p1, L03 p1, L05 p1, L07 p1, L10 p1, L13 p1, L16 p1, L18 p1 All 17 decks L02–L18 — the single most consistent element in the course
Learning goals Bold Learning Goals heading, numbered list 1. 2. 3. L02 p1 … L18 p1 Every deck L02–L18 (L07 p1 uses an unnumbered list — the one exception)
Student Q&A Bold blue heading Student questions and comments: (or Student questions:), questions in blue, answers in red L02 p1, L04 p1, L06 p1, L08 p1, L11 p1, L12 p1, L15 p1, L18 p1 Every deck L02–L18
Activities Bold Activity N: restarting at 1 each lecture; pre-class items as Q1 from the pre-class questions; Clinch decks add Poll question N: L02 p3, L07 p5, L12 p2, L17 p3 Throughout

Figures are never numbered and never labelled “Figure”. When a caption exists it sits centred directly beneath the drawing, either as the mathematical expression it depicts ((A − B), L18 p2) or as a short prose verdict (not a function / a function, L18 p9).

Colour

Printed colour on slides

  • Red = the thing you must not get wrong. Warnings, answers to student questions, “IMPORTANT NOTE”, “Take care with notation!”, “WARNING”. L01 p4 (“hurdles”), L02 p1, L08 p2 (“WARNING”), L10 p1 (“IMPORTANT NOTE”), L11 p4 (“Take care with notation!”), L17 p2 (“Warning!!!!”), L17 p9. Holds in both lecturers’ decks, all semester.
  • Blue = student voice and explanatory gloss. Student questions (L02 p1 onward, universal) and, in Clinch’s decks, a plain-English restatement placed directly beneath a formal condition written in red — L12 p3 (d|a and d|b in red, “d is a divisor of both a and b” in blue below it), L12 p4 (same pattern for lcm). That paired red-above/blue-below idiom is specific and recurs.
  • Colour as correspondence, not meaning. Where two expressions must be matched piece-by-piece, matching pieces are given matching colours. L03 p9 (a three-row table where a ↔︎ b, a if and only if b, b is necessary and sufficient for a are colour-aligned column by column), L06 p6 (quantifier negation, below), L11 p3 (quotient in red above q, remainder in red above r, and the same two colours on the worked examples). This is the one place colour genuinely carries information.
  • Pastel highlight capsules (yellow, green, blue, pink, orange) mark the single expression being tracked through a calculation — it keeps its capsule colour as it moves line to line. Heavily used by Clinch: L11 p3, L12 p6, L13 p6, L14 p3, L16 p5, L17 p5. Rare in Davies’s decks.

Annotation ink (live handwriting)

Both lecturers annotate over the slides in ink. Blue is the predominant ink for the main worked solution — L01 p13, L02 p7, L04 p4, L07 p4, L08 p5, L10 p5, L13 p7, L16 p7, L18 p3. Other colours (green, red, purple, orange, teal) are layered over it for justifications, second solutions, and circling.

Be honest about what ink colour means: mostly, nothing fixed. It separates successive layers of annotation for legibility, and rotates. The two places it is systematic:

  1. Correspondence — e.g. L02 p6, where each intermediate truth-table column is written in its own colour and numbered ①②③ in that colour to show evaluation order, with the final column ringed and labelled “final answer”.
  2. Verdict marks — a green circled T and a red circled F for true/false answers: L05 p7, L05 p8, L06 p4. Green tick / red cross for correct/incorrect options: L06 p3, L09 p4, L16 p2.

Boxes and callouts

Shape Meaning Evidence Holds?
Purple rounded rectangle A named, quotable object: definition, Fact, Lemma, Theorem — and the pre-work video box L07 p1 (the six definitions), L07 p2 (two Facts), L08 p2, L09 p6 (Fact + Lemma), L10 p6, L10 p7 (Unique Factorisation Theorem), L10 p9, L10 p12, L11 p3 (Quotient-Remainder Theorem), L11 p7, L11 p8 Strong: both lecturers, L07 through L12
Plain thin black rectangle A structural panel — a proof skeleton, or a two-way comparison L07 p3, L08 p3, L09 p2 (the four proof-method skeletons), L11 p3 (“2 equivalent ways of defining…”) Consistent where it appears
Red rectangle Davies: a fact that must be memorised or a warning. Clinch: a numbered step caption or a named principle L02 p8 (De Morgan’s laws), L03 p4, L03 p9, L04 p1 (“Recall:”), L01 p4; and L10 p6 (margin commentary), L12 p10 (steps 1/2/3), L14 p6 (Well-ordering Principle) Both uses recur; see Inconsistencies
Margin callout with arrow A red-outlined box in the right margin, with a red arrow to the exact proof line it explains, and a purple ★ at the anchor point L10 p6 (two callouts, two arrows, two stars) Seen once — extrapolating from a single instance
Hand-drawn cloud An aside or a menu of options: “What type of proof will work?” with ✓/✗ against each method L09 p4, L09 p5, L09 p6, L09 p7; L14 p3 Four consecutive slides in L09, once in L14

Logic notation

Fixed by the formula sheet and used unchanged in every deck:

Symbol Reading Note
∼ p not p The course symbol is the tilde ∼, not ¬. L02 p1 answers this directly: “some common symbols for not include ∼ and ¬” — both exist, this course writes ∼. Used in every deck and on the formula sheet
p ∧ q, p ∨ q and, or ∨ is inclusive or; exclusive or is ⊕ (L02 p5)
p → q if p then q L03 p1; defined as ∼ p ∨ q
p ↔︎ q if and only if A connective that combines statements
p ≡ q, p ≢ q logically equivalent / not A relation that compares statements
t, c tautology, contradiction Bold upright, lower case. Formula sheet; L02 p9
∴ therefore Prefixes the conclusion of an argument

↔︎ versus ≡ is stated explicitly and must not be blurred (L03 p7): “The biconditional (↔︎) is a conjunction used to combine statements (similar to +, -, x, ÷ … with numbers). Logical equivalence (≡) is used when comparing two statements (similar to = with numbers).”

Precedence (L02 p5, extended at L03 p7): ∼ first; ∧ and ∨ equal second; → and ↔︎ equal third; within a tier, work left to right — “but parentheses should always be used to avoid confusion.”

Statement variables are lower-case italic p, q, r. When a problem is in English, the letters are chosen as mnemonics of the English phrase: w = wages, b = buying, d = depression (L04 p4); q = queued, s = sent to the buffer (L05 p2).

Truth tables

The course calls this “standard truth table form” and requires it (L02 p4): “You should construct your truth tables in standard form so that you are sure you have covered all combinations … and it is easy for someone to check your work.”

  • Values are the single upright capitals T and F — never 1/0, never True/False. Universal: L02 p4, L03 p1, L04 p5, L06 p2, L07 p3, L09 p2.
  • 2ⁿ rows for n variables, in a fixed order: the first variable is T for the whole top half then F for the whole bottom half; each subsequent variable halves again; the last variable alternates T,F every row. Shown for n = 1, 2, 3, 4 side by side at L02 p4.
  • A narrow solid-filled amber/orange column separates the input block from the compound-form block. L02 p4 (after the variable columns, before “Statement form”); L04 p5 (after the variables and their simple negations, before the premise and conclusion columns). Two instances, both lecturers’ style aside — treat as the course’s own table furniture.
  • Header row typeset in red italic maths (p, q, p → q), body values in black upright. L03 p1, L03 p4, L04 p1, L04 p2, L04 p5, L07 p3, L08 p3, L09 p2. Consistent throughout Davies’s decks; Clinch draws no printed truth tables.
  • Full ruled grid — every cell bordered.
  • When built by hand, columns are added left to right in evaluation order and numbered ①②③ beneath, each in its own ink colour, with the answer column ringed. L02 p6, L03 p2. At L03 p2 the two columns that must match are marked with a ✳ beneath and the verdict written out: “Starred columns are identical so ∼((∼p∧q) ∨ (∼p∧∼q)) ≡ p”.

Arguments and inference proofs

Argument form (formula sheet; L04 p2, L04 p3, L04 p6, L05 p2, L05 p4): premises stacked one per line, left-aligned; the conclusion on the last line prefixed by ∴. No horizontal rule above the conclusion. Premises are numbered 1. 2. 3. and the ∴ line is left unnumbered.

Rules-of-inference proof (L04 p6, L04 p7, L05 p4): numbered lines down the left, the derived statement in the middle, and the justification in its own right-hand column, phrased as from ② by contraposition, from ① & ④ by transitivity, by double negative law, by definition of →. The annotation at L04 p6 is emphatic: “must write explanations!”.

Equivalence-law chains (L02 p10, L03 p8, L18 p5, L18 p7): first line is the starting expression; every subsequent line begins with ≡ (or = for sets), vertically aligned, with the named law in a right-hand column — by de Morgan's law, by the negation law, by identity law. For set identities the numbered law from the formula sheet is cited: by set diff law (12), by de Morgan (9b), dist. law (3b), complement law (5b), identity law (4a) (L18 p7).

Biconditional proofs are marked by direction: ⇒ and ⇐ label the two halves (L12 p8). Two named methods, with a stated preference — Method 1, prove each direction separately, is annotated “recommended method”; Method 2, an iff chain, is annotated “Be careful! Quicker method but easy to make a logical error” (L12 p8, L12 p9).

Quantifiers

  • ∀x ∈ D, P(x) — universal, comma after the domain.
  • ∃x ∈ D such that P(x) — existential, the words “such that”, abbreviated s.t. (L06 p1, L05 p7).
  • Both established at L05 p5 and used unchanged through L08 p1, L17 p3.
  • ∈ reads “is an element of” / “belongs to”; ∉ “is not an element of” (L05 p5).
  • Negation rules, stated as displayed equivalences at L08 p1: ∼(∃x ∈ D such that P(x)) ≡ ∀x ∈ D, ∼P(x) and ∼(∀x ∈ D, P(x)) ≡ ∃x ∈ D such that ∼P(x).

Number sets

Blackboard-bold, established at L05 p1 and used unchanged to the end: ℝ, ℚ, ℤ, ℕ.

Irrational numbers are written ℚ̄ — ℚ with an overbar (L05 p1, L09 p1, L09 p3, L09 p6). Not ℝ∖ℚ; the equality ℝ∖ℚ = ℚ̄ is written by hand at L09 p1 as a gloss, but the barred form is what the printed slides use.

Restrictions are written as superscripts on the set symbol, in italic, not as subscripts and not as prose:

ℤ⁺ ℤ^even ℤ^odd ℤ^nonneg ℤ^prime ℤ^composite ℤ^{>1} ℤ^{≥2} ℤ^{≤1} ℤ⁻

Evidence: L05 p1 (ℤ⁺, ℤ^even, ℤ^nonneg), L07 p2 (ℤ^even, ℤ^odd, ℤ^prime, ℤ^composite, ℤ^{≤1}, ℤ^{>1}), L07 p6 (ℤ^{≥2}), L10 p12 (ℤ^{≥2}), L17 p10 (ℤ⁺, ℤ⁻). Strong and semester-long.

Proof layout

The four methods each have a fixed printed skeleton, given verbatim on a repeated slide (L07 p3, L08 p3, L09 p2 — the same panel, grown one box at a time as methods are introduced). Reproduce these wordings:

  • Direct proof — Let x ∈ D and suppose P(x). … Hence Q(x). □
  • Disproof by counterexample — Let x = ⋯. Then x ∈ D and P(x) is true. However, Q(x) is false. So x is a counterexample and the statement is false.
  • Proof by contradiction — Suppose the negation of ∀x ∈ D, P(x) → Q(x) is true. That is, suppose ∃x ∈ D such that P(x) is true but Q(x) is false. … Hence something that is obviously wrong… This is a contradiction, and we conclude that … is true. □
  • Proof by contraposition — We prove the contrapositive of ∀x ∈ D, P(x) → Q(x), namely that ∀x ∈ D, ∼Q(x) → ∼P(x). Let x ∈ D and suppose ∼Q(x). … Hence, ∼P(x). □

Each box carries a red caption directly beneath it saying what it establishes: Shows the statement is true. / Shows the statement is false. (L07 p3, L08 p3).

Other layout facts:

  • Proof: label, then the hypothesis line, then the working with = signs vertically aligned in a column, then a Hence/Thus/So conclusion. L07 p4, L07 p5, L08 p5, L09 p3, L10 p5.
  • Method comparison is a two-column layout with bold centred headings — Direct Proof left, Proof by contradiction right (L08 p5, L08 p6); Proof SMI / Proof MI / Proof WOP (L14 p7, L14 p8).
  • Without loss of generality, spelled out then abbreviated WLOG (L08 p8, L09 p5).
  • Case splits are labelled Case 1: / Case 2: (L10 p6, L10 p12, L14 p5).
  • Every proof unfolds definitions. The six definitions in the purple box at L07 p1 / L08 p2 are stated as biconditionals using ↔︎ — e.g. n is an even integer ↔︎ n = 2k for some integer k, d|n ↔︎ d ≠ 0 and n = dk for some integer k — because they are used in both directions.
  • Introduce a new variable for each new existential. Annotated explicitly at L08 p6: “use a new variable!” where m = 2k+1 is followed by n = 2ℓ+1.

End-of-proof mark

Printed proofs end with an open square □: L07 p3, L07 p4, L08 p3, L09 p2, L10 p6, L10 p11, L10 p12.

Handwritten proofs end with either a hand-drawn open square (L07 p5, L08 p8, L09 p4, L09 p5, L09 p6) or a double slash // (L10 p5, L11 p6, L13 p7, L13 p9, L14 p10, L15 p5, L16 p7, L18 p3, L18 p7). At L11 p6 the lecturer writes an explicit legend of the acceptable marks: // or □ ▪ QED. Treat all four as sanctioned.

Contradiction mark

Davies writes (⇒⇐) — two arrows meeting — inline at the point the contradiction lands: L08 p3, L08 p5, L08 p6, L08 p7, L08 p8, L09 p6. Clinch instead uses a small crossed glyph ⨳ inline, e.g. “(proof by ⨳)” for “proof by contradiction” (L14 p6, twice). See Inconsistencies.

Induction

The named-step skeleton is fixed and printed identically in five separate proofs:

Proof:  Let P(n) be the predicate  ⟨statement⟩.
Basis step:        ⟨verify P(a)⟩ … so P(a) is true.
Inductive step:    Prove that ∀k ∈ ℤ with k ≥ a, if P(k) is true, then P(k+1) is true.
Inductive hypothesis:  Suppose that k ≥ a is an integer and that P(k) is true.
        LHS of P(k+1):  ⟨chain of = with the RHS reached⟩
        Therefore P(k+1) is true.
Therefore, by ⟨PMI / SMI⟩, ⟨statement⟩ for each integer n ≥ a.

Evidence: L13 p6, L13 p8, L14 p3, L14 p7, L16 p6, L16 p8, L16 p10. Step labels are printed in blue at L13 p6 and in red at L14 p3/L14 p7.

  • LHS / RHS are the standard abbreviations, and the two sides are named and worked separately: LHS of P(1): … RHS of P(1): … (L13 p6, L13 p7, L16 p7).
  • The step where the hypothesis is used is annotated (by the inductive hypothesis) on the same line (L13 p6, L13 p7, L14 p3).
  • Slides ask for the steps to be named: “It is a good idea to list these explicitly in your proofs to help the reader to follow the proof” (L13 p1).
  • Multiple base cases are listed one per line when strong induction needs them (L14 p4, L15 p8).

Number-theory notation

  • Divides: d | n. Does not divide: d ∤ n (slashed bar). L05 p7, L07 p1, L10 p1, L10 p10, L14 p8.
  • Equivalent phrasings given at L10 p1: d divides n, d is a divisor of n, d is a factor of n, n is a multiple of d.
  • n mod d = r (upright mod as a binary operator, no brackets) versus m ≡ n (mod d) (congruence, (mod d) bracketed after the statement). These are different things and the course flags it twice under a red heading Take care with notation! — L11 p3, L11 p4, recapped L12 p2.
  • gcd(a, b) and lcm(a, b), lower case, upright roman (L12 p3, L12 p4). At L12 p3 the alternative hcf(a,b) is written and then struck through — it is not this course’s name.
  • Floor ⌊x⌋ and ceiling ⌈x⌉ (L11 p2).
  • Standard factored form: n = p₁^{e₁} p₂^{e₂} ⋯ p_k^{e_k} with the primes in increasing order p₁ < p₂ < ⋯ < p_k (L10 p7, annotated “increasing” at L10 p8).
  • Multiplication is a centred dot · in printed slides: 2 · 3 · 5 · 7 · 11 ⋯ p (L10 p11), 5859 = 1232 · 4 + 931 (L12 p6), 2 · 100 + 5 · 10 + 4 · 1 (L11 p8), 3 · 1 + 2 (L13 p6).

Sequences, sums and recurrences

  • Summation Σ with i = m below and n above; index letters i, j, k; change of variable written as j = i + 1 (L13 p3, L13 p4, L13 p6).
  • A recursive definition of a sequence requires two named parts: a recurrence relation and initial conditions (L15 p3, L16 p7, L16 p9).
  • A recursive definition of a set requires three named parts: Base, Recursion, Restriction — always in that order, with the Restriction worded “Nothing is in B other than objects defined by the Base and the Recursion” (L15 p6, L15 p7).
  • Verifying a guessed explicit formula has two named methods: Method 1 (check the initial conditions and the recurrence relation — annotated “quicker”) and Method 2 (prove it by induction — annotated “rigorous”). L16 p4, used at L16 p5–L16 p10.
  • n! with 0! = 1 by definition (L15 p5).

Set-theory notation

Notation Meaning Evidence
x ∈ A is an element of L17 p1
A ⊆ B is a subset of L17 p1, L17 p3
A ⊂ B is a proper subset of (⊆ and there is an element of B not in A) L17 p3 — the course does use ⊂ strictly for proper
A ∩ B, A ∪ B intersection, union; indexed forms ⋃_{i=1}^{3} Aᵢ, ⋂_{i=1}^{3} Aᵢ L17 p1, L17 p6
A − B set difference; A \ B is stated to be equally acceptable L17 p1 — answered directly: “Yes. A − B can be written as A \ B”
Aᶜ complement, superscript c L17 p1, L18 p1, formula sheet
𝒫(A) power set, calligraphic script P L17 p1, L17 p7, L17 p8
A × B Cartesian product; ℝ × ℝ also written ℝ² L17 p1, L17 p9
∅ empty set, also written { } L17 p5
\|A\| cardinality L17 p7, L17 p9
{n ∈ ℤ \| n ≡ 2 (mod 8)} set-builder with a vertical bar L17 p3, L18 p2
[a, b] closed interval {x ∈ ℝ \| a ≤ x ≤ b} L17 p6

Sets take capital italic letters (A, B, C, S, T, D, E, R, X, Y), the universal set is U, elements take lower case (x, y, t). L17 p3, L18 p1, L18 p6.

The ∈ / ⊆ distinction is the course’s flagged trap — a red Warning!!!! at L17 p2 and the first learning goal of both L17 and L18. 2 ≠ {2} (L17 p2); ℤ ∈ ℚ is false, ℤ ⊆ ℚ is true (L17 p1, L17 p2).

Proof of set statements has three named methods (L18 p4): the element method, using the laws of logical equivalence, and using set identities (citing the formula sheet’s numbering, as above).

Functions (L18 p9–L18 p11): f: X → Y; f ⊆ X × Y such that for each x ∈ X there is a unique y ∈ Y with (x, y) ∈ f. Domain X, co-domain Y (hyphenated in print), range f(X) = {f(x) : x ∈ X}. Image of a subset f(A); inverse image or preimage f⁻¹(B) — both words are given as correct.

Diagrams

This is what the decks actually draw. Nine idioms recur; everything else is a one-off and is labelled as such.

1. Nested closed curves for containment (Euler style)

Number systems are drawn as concentric plain closed curves with each label written inside its own region — ℚ: rational numbers outermost, containing ℤ: integers, containing ℕ: natural numbers; ℝ labelled above the frame, ℚ̄: irrational numbers in a separate adjacent panel. Thin blue rule dividing the two panels; curves thin black; no shading; no element dots. L05 p1. Seen once for number systems, but the same nested-oval idiom recurs for subsets at L17 p3 and L18 p3, so the shape convention is safe — the specific number-system figure is not.

2. Rectangle partitions

A plain rectangle with a thin blue/teal outline stands for the ambient set; thin internal straight lines split it into disjoint parts; each part is labelled in red italic with its superscript-qualified name; the ambient set’s own label sits above and outside the top-left corner. Two instances on one page: ℤ split into ℤ^even | ℤ^odd, and ℤ split into ℤ^{≤1} | ℤ^prime / ℤ^composite. L07 p2. Seen on one page only — but it is the course’s only printed partition figure, and it is the natural pairing with the superscript set notation above.

3. Two-circle Venn diagrams

  • Exactly two circles of equal radius, overlapping, drawn side by side in a horizontal row; thin black outline; labels A and B placed inside each circle near its top outer edge.
  • The region being denoted is filled solid grey. The undecorated pair is drawn first, unshaded, as the reference.
  • The expression it depicts is the caption, centred directly beneath — (A − B), (B − A).
  • L18 p2. Three diagrams in a row.
  • Hand-drawn variant at L18 p3: a rectangle for the universe with red diagonal hatching over the complement region, with A (green) inside B (green outline) inside it.

Critical rule, stated by the course itself: at L18 p2 a hand annotation brackets all three Venn diagrams and writes “not a proof!”. A Venn diagram in this course illustrates a set statement; it never discharges one. Proofs go through the element method, the laws of logical equivalence, or the set identities (L18 p4).

4. Small icon strip for set relations

A row of thumbnail two-set diagrams, one per notation, aligned under the list of symbols they gloss: set A drawn as a green ellipse, set B as a pale blue/grey ellipse; an element as a small filled dot; the operation’s region filled or hatched in blue. L17 p1 (the full strip: x ∈ A, A ⊆ B, A ⊂ B, A ∩ B, A ∪ B, A − B, and A = B drawn as coincident ellipses). Repeated as single icons at L17 p3, L17 p4, L17 p6. The green-A / grey-B colour pairing holds across all five appearances.

5. Function arrow diagrams (bipartite)

  • Two vertical ellipses side by side: the domain X on the left, the co-domain Y on the right, each labelled beneath the ellipse in blue.
  • Elements are labelled dots stacked vertically inside each ellipse.
  • Each element of X is joined to its image by a straight blue line segment. Flow is left to right.
  • A subset under discussion is marked by a translucent coloured blob drawn around the relevant dots, labelled just outside in the matching colour with its expression (f({2,3}), f⁻¹({1}), f({1})).
  • L18 p11 — four diagrams on the page, all hand-drawn, all following this arrangement.

6. Cartesian-plane graphs

Axes as thin double-headed arrows crossing at the origin; the y label at the top of the vertical axis and the x label at the right of the horizontal axis, both immediately beside the arrowhead; the curve in blue; marked points as filled dots labelled with their coordinate pair (0,1), (5,11); a prose caption centred beneath (Plot of the function y = 2x + 1., not a function, a function). The vertical line test is shown as a red vertical line crossing the curve twice. L18 p9 — three panels in a row.

7. Number lines

Drawn as a horizontal line with an arrowhead at each end, evenly spaced tick marks, integer labels beneath the ticks, and marked points or brackets above the line. L11 p2 (two of them: the real line from −5 to 4 for floor/ceiling, and a minimal one showing only k and k+1 with k + ½ between); L17 p6 (interval [a,b] drawn above the line with filled dots at the closed endpoints, the axis labelled ℝ at the right end).

8. Colour-coded correspondence diagrams

The single most information-dense idiom in the course. At L06 p6, the negation of a two-quantifier statement is shown as two copies of the same statement schema stacked one above the other and horizontally aligned part-for-part, with each corresponding pair of parts underlined in its own colour and joined by a double-headed arrow of that colour:

  • red: ∀x ∈ X ↔︎ ∃x ∈ X such that
  • green: ∃y ∈ Y such that ↔︎ ∀y ∈ Y,
  • cyan: P(x,y) ↔︎ ∼P(x,y)
  • a purple curved double-headed arrow spanning both rows, labelled negation

Seen once, but it is the canonical figure for the topic. The same alignment-plus-colour technique without arrows appears at L03 p9 and L12 p3.

9. Algorithm and process cascades

  • Euclidean algorithm (L12 p6): the successive equations a = bq + r stacked vertically down the left; each numeral wrapped in a pastel capsule, and each number keeps the same capsule colour as it migrates from b to a and r to b on the next line; red diagonal arrows trace those two migrations; a parallel right-hand column carries the matching gcd(a,b) = gcd(b,r) chain with the same capsule colours. Colour here is pure identity-tracking.
  • Sequence hops (L16 p2, L16 p3, L15 p5): terms written in a horizontal row, short curved arrows above or below each gap, labelled with the operation (+3, ×3, × −1).
  • Recursive-construction tree (L15 p6): root ε at the left, branches as plain straight segments fanning rightwards, node labels at the branch tips, no arrowheads on the branches. The separate derivation chain uses arrows with the rule name written above: ε --1--> 1 --1--> 11 --01--> 1101 (L15 p7).
  • State sequences (L16 p11, Tower of Hanoi): three vertical blue pegs on a base line; discs as short red horizontal bars, widest at the bottom; each state enclosed in a plain black rectangle; states run left to right; green curved arrows between states labelled with the move count (m_{n-1}, 1, m_{n-1}).
  • Partitions of a set as coloured blobs (L17 p10): irregular non-overlapping closed curves in distinct colours, each containing dots, each labelled just outside in its own colour (A₁ green, A₂ red, A₃ yellow, A₄ blue). On the same page a number line is partitioned instead by weaving highlighter stripes — pink for S, green for T — over the tick labels.

Tables

  • Full ruled grid, header row present.
  • Truth tables as described above.
  • Comparison tables put the item varying across the top and the cases down the side, with T/F in the cells (L06 p2: three quantified statements across, three sets S down).
  • Material reproduced from the examination formula sheet is pasted as an image and therefore appears in serif LaTeX type, visually distinct from the sans-serif slide body: L02 p9 (Logical Equivalences), L04 p3 (Valid Argument Forms), L18 p6 (Set Identities). If reproducing one of those tables, reproduce the formula sheet’s own wording and numbering.

Inconsistencies

The course does these things more than one way. Both readings are recorded; no winner is picked. A decision is needed before any new figure fixes one.

  1. End-of-proof mark. Printed proofs use □ (L07 p3, L10 p11). Handwritten proofs use a hand-drawn square (L09 p4) or // (L13 p7, L18 p7). At L11 p6 the lecturer offers all of //, □, ▪ and QED as equivalent. Clinch’s printed induction proofs (L14 p7, L16 p6) end with a concluding sentence and no mark at all. Four options, all sanctioned. Decide.

  2. Contradiction mark. Davies writes (⇒⇐) (L08 p3, L08 p5, L08 p6, L08 p7, L08 p8, L09 p6). Clinch writes a crossed glyph ⨳ and uses it as a noun — “(proof by ⨳)” (L14 p6, two occurrences on one page). Decide.

  3. Name of the induction principle. Within four decks: “by the principle of mathematical induction” (L13 p6), “by PMI” (L13 p8, L16 p12), “by MI” (L16 p10), “by SMI” (L14 p5, L16 p8), “by PSMI” (L14 p3, L14 p7). PMI/MI and SMI/PSMI are used for the same two principles in the same lecturer’s decks. Decide.

  4. Sequence delimiters. Printed: {bₙ}_{n≥0} (L13 p2), {aₙ}_{n≥1} (L15 p3), and bare a₀, a₁, a₂, … (L14 p3, L16 p2). Handwritten on the same slide (L13 p1): (aₙ)_{n≥0} and {aₙ}_{n=1}^{∞}. Four forms. Decide.

  5. Set-builder separator. Vertical bar in {n ∈ ℤ | n ≡ 2 (mod 8)} (L17 p3), {(x,y)|x ∈ ℝ and y ∈ ℝ} (L18 p9); colon in f = {(x,y) ∈ ℝ × ℝ : y = 2x + 1} (L18 p9, same page), f⁻¹(B) = {x ∈ X : f(x) ∈ B} and f(A) = {f(x): x ∈ A} (L18 p11). Both on the same page as each other. Decide.

  6. Multiplication sign. Printed slides use the centred dot: 1232 · 4 (L12 p6), 2 · 3 · 5 · 7 (L10 p11). Handwriting on the very next slide uses ×: 2431 × 1 (L12 p7), 4 × 9 (L13 p9), ×3 (L16 p2). Decide.

  7. Set difference symbol. A − B throughout the printed slides and the formula sheet; A \ B explicitly declared equally correct at L17 p1. Not a conflict of fact, but a live choice.

  8. Which box shape carries a named principle. Purple rounded rectangle for Theorem/Lemma/Definition/Fact (L07 p1, L10 p7, L11 p3) versus a red rectangle for the Well-ordering Principle (L14 p6). Same category of object, two shapes.

  9. Text colour inside the purple definition box. Red at L07 p1, L08 p2, L11 p7; blue at L10 p7 (Unique Factorisation Theorem); black at L07 p2 (the two Facts). Three colours for the same box type.

  10. Capitalisation of the induction step labels. “Inductive hypothesis” (L13 p6, L14 p3) versus “Inductive Hypothesis” (L16 p6, L16 p8, L16 p10). Also “Let P(n) be the predicate” (L13 p7, L16 p6) versus “Let P(n) denote the predicate” (L13 p6).

  11. LHS versus L.H.S. L.H.S / R.H.S with stops at L13 p1; LHS / RHS without at L13 p6, L13 p7, L16 p7, L16 p9.

  12. Does ℕ contain 0? The printed diagram at L05 p1 shows ℕ: natural numbers nested inside ℤ without saying which. The handwritten annotation on the same slide gives both ℕ = {0,1,2,…} and = {1,2,3,…} with the note “take care!” — the course is deliberately flagging the ambiguity rather than settling it. ℤ^nonneg = {0,1,2,3,…} is defined unambiguously on the same slide and is the safe alternative.

  13. Writing two variables in the same domain. Three forms shown side by side at L06 p5: ∃x ∈ D and ∃y ∈ E … (annotated “more clear”), ∃x ∈ D and y ∈ E …, and ∃x, y ∈ D … (annotated “short hand for same domain”). Preference is hinted at but not stated as a rule.

Not established by the source

Be candid: these are where a new diagram would be inventing, not conforming.

  • Arrowheads on function-diagram connectors. Every mapping at L18 p11 is drawn with a plain line segment. Whether an arrowhead is intended, required, or merely omitted in fast handwriting cannot be told from one page. Nothing in the corpus shows an arrowheaded mapping diagram.
  • Three-set Venn diagrams. Only two-circle Venns are ever printed (L18 p2). The set identities involving three sets (associative, distributive — L18 p6) are never illustrated. There is no evidence for the standard three-circle arrangement, region labelling, or which regions get shaded.
  • A universe rectangle in printed Venn diagrams. L18 p2 has no enclosing rectangle at all. The only universe rectangle appears in a hand annotation (L18 p3) with red diagonal hatching. Whether a printed figure should include U is unsettled.
  • Shading colour and fill style. One printed instance uses solid mid-grey (L18 p2); the icon strip uses blue hatching and blue solid interchangeably on the same row (L17 p1). No stated palette.
  • Line weights, stroke colours for plain geometry, and typeface. Nothing is specified anywhere. Slide body text is a Calibri-like sans; reproduced formula-sheet extracts are serif LaTeX. Beyond that, invent nothing.
  • Figure numbering and cross-referencing. No figure in 176 pages carries a number or the word “Figure”. There is no established way to refer to a diagram from body text.
  • Any convention for the second half of the course. L19–L36 have no decks in this set. That means nothing is established for: one-to-one / onto / inverse function diagrams; composition diagrams; cardinality and countability arguments (the Schröder–Bernstein theorem is on the formula sheet but never drawn); relation digraphs; Hasse diagrams for partial orders; equivalence-class pictures; Cayley tables for groups; counting trees and Pascal’s triangle; graph drawings (vertex shape, edge style, labelling, directedness); walks/trails/circuits notation; adjacency-matrix layout; and tree drawings for graph theory. The only graph-theoretic images in the corpus are the Königsberg bridge photograph and a Tower of Hanoi photograph on L01 p13, which establish nothing.
  • Whether hand-annotation colour should be reproduced at all. The ink layers exist because the slides are annotated live. A clean redrawn figure has no equivalent of “second colour for the second pass”, and the corpus gives no printed-only version of most annotated diagrams to copy from.
  • A logic-gate or circuit notation. None appears; MATH1061 is not a circuits course, but L01 p10 motivates logic via computer science, so the absence is worth stating.
  • Whether ⊂ may ever mean plain subset. The course defines it as proper subset (L17 p3) and uses it that way consistently in the pages seen, but never says “we never use ⊂ loosely”, so a reader coming from other texts has no explicit prohibition to point at.

Mathematical Induction

The principle of mathematical induction, the four-step shape every induction proof takes, and the extra care inequalities need. See sequences-and-summations for the summation manipulations these proofs run on, and proof-techniques for the other proof methods in the course.

Why it works

Let \(P(n)\) be a predicate written in terms of some integer \(n\). Suppose you know two statements are true:

  1. \(P(2)\)
  2. \(\forall k \in \mathbb{Z}\) with \(k \geq 2\), if \(P(k)\) then \(P(k+1)\)

Can you conclude \(P(5)\)? Yes — statement 2 is a general implication, so instantiate it repeatedly and apply modus ponens. The conclusion of one step is the premise of the next, which is what turns two facts into a chain (Figure 4).

Figure 4: Why those two statements are enough. The basis step pins one statement true; every hop after it is the same general implication used at one more value of \(k\), so each conclusion becomes the next premise. The panel writes one hop out in full — it is a single modus ponens, nothing more.

Nothing here is special to 5. \(P(5)\) is reached in three hops, but the chain never stops, so \(P(n)\) is true for each integer \(n \geq 2\).

Both statements are needed. Statement 2 alone gives nothing: an implication \(P(k) \to P(k+1)\) is true in three of the four rows of its truth table — including both rows where \(P(k)\) is false. Without the basis step pinning \(P\) true somewhere, the chain has nothing to start from and every link stays vacuously true while every \(P(n)\) stays false.

The shape of an induction proof

Four steps, and it is worth naming each one explicitly in the write-up so the reader can follow the argument:

  1. Define the predicate. Let \(P(n)\) denote (the statement, with \(n\) free).
  2. Basis step. Prove the predicate is true at the smallest value of \(n\) in the claim — evaluate the two sides separately and observe they agree.
  3. Inductive hypothesis. Suppose \(k\) is an integer at or above the basis value, and that the predicate is true for \(n = k\). Write out what \(P(k)\) actually says.
  4. Inductive step. Prove \(P(k+1)\) follows, i.e. \(\forall k \in \mathbb{Z}\) with \(k \geq\) (basis), if \(P(k)\) is true then \(P(k+1)\) is true. Start from the LHS of \(P(k+1)\), use the inductive hypothesis somewhere (say where), and arrive at the RHS of \(P(k+1)\).

Then conclude: therefore, by the principle of mathematical induction, … for each integer \(n \geq\) (basis).

One thing about the order. In the write-up, the inductive step’s goal — prove that \(\forall k \in \mathbb{Z}\) with \(k \geq\) (basis), if \(P(k)\) is true then \(P(k+1)\) is true — goes on the line above the inductive hypothesis, and the assumption is made underneath it: goal first, assumption second. The numbering above counts the four actions; the goal line itself is not numbered. Write the goal down, then assume, then work.

Worked example — a summation

Prove that for each integer \(n \geq 1\), \(\displaystyle\sum_{i=1}^{n} (3i+2) = \frac{3n^2 + 7n}{2}\).

Proof. Let \(P(n)\) denote the predicate \(\displaystyle\sum_{i=1}^{n} (3i+2) = \frac{3n^2+7n}{2}\).

Basis step. Prove the predicate is true for \(n = 1\).

LHS of \(P(1)\): \(\displaystyle\sum_{i=1}^{1}(3i+2) = 3 \cdot 1 + 2 = 5\). RHS of \(P(1)\): \(\dfrac{3(1^2) + 7(1)}{2} = \dfrac{3+7}{2} = 5\). Therefore \(P(1)\) is true.

Inductive step. Prove that \(\forall k \in \mathbb{Z}\) with \(k \geq 1\), if \(P(k)\) is true then \(P(k+1)\) is true.

Inductive hypothesis: suppose \(k \geq 1\) is an integer and the predicate is true for \(n = k\) — that is, \(\displaystyle\sum_{i=1}^{k}(3i+2) = \frac{3k^2+7k}{2}\).

LHS of \(P(k+1)\):

\[\sum_{i=1}^{k+1}(3i+2) = \sum_{i=1}^{k}(3i+2) + \big(3(k+1)+2\big)\] \[= \frac{3k^2+7k}{2} + 3k + 5 \qquad \text{(by the inductive hypothesis } P(k)\text{)}\] \[= \frac{3k^2 + 13k + 10}{2} = \frac{3(k^2+2k+1) + 7k + 7}{2} = \frac{3(k+1)^2 + 7(k+1)}{2}\]

which is the RHS of \(P(k+1)\).

Therefore, by the principle of mathematical induction, \(\displaystyle\sum_{i=1}^{n}(3i+2) = \frac{3n^2+7n}{2}\) for each integer \(n \geq 1\).

The move at the first line is splitting off the last term (see sequences-and-summations) — that is what exposes the \(P(k)\) sum so the hypothesis can be applied to it. Figure 5 tracks that one expression through the chain.

Figure 5: Where the inductive hypothesis actually enters. Splitting off the last term turns the LHS of \(P(k+1)\) into the LHS of \(P(k)\), plus one more term; the hypothesis is then spent exactly once, swapping that sum for its closed form; everything below is algebra. The shading follows that one expression from line to line.

Before starting the chain, write down what the RHS of \(P(k+1)\) actually is — here \(\frac{3(k+1)^2 + 7(k+1)}{2}\). Without it you are simplifying with no idea what counts as finished; with it, the last algebraic step is obvious, because you already know the shape you are steering the numerator into.

Induction with inequalities

Be careful when manipulating inequalities. The rules that hold:

Given Conclusion
\(a < b\) \(a + x < b + x\), regardless of the sign of \(x\)
\(a < b\), \(x\) positive \(ax < bx\)
\(a < b\), \(x\) negative \(ax > bx\) — the direction flips
\(a < b\), \(x\) zero \(ax = bx\)
\(a < b\) and \(x < y\) \(a + x < b + y\)
\(a < b\) and \(x > y\) nothing can be concluded about \(a + x\)

All of these still hold with \(<\) and \(>\) replaced by \(\leq\) and \(\geq\).

Worked example — \(n^2 < 2^n\)

Prove that for each integer \(n \geq 5\), \(n^2 < 2^n\).

Proof. Let \(P(n)\) be the predicate \(n^2 < 2^n\).

Basis step. Consider \(n = 5\). Then \(n^2 = 25\) and \(2^n = 32\). Since \(25 < 32\), \(P(5)\) is true.

Inductive hypothesis. Suppose \(k \geq 5\) is an integer and \(P(k)\) is true. Thus \(k^2 < 2^k\), and hence \(2k^2 < 2 \cdot 2^k\).

Inductive step. Consider \(n = k+1\); the goal is \((k+1)^2 < 2^{k+1} = 2 \cdot 2^k\).

\[(k+1)^2 = k^2 + 2k + 1 < k^2 + 2k + k = k^2 + 3k < k^2 + k^2 = 2k^2 < 2(2^k) = 2^{k+1}\]

The first inequality uses \(1 < k\) (since \(k \geq 5\)); the second uses \(3k < k \cdot k\), which holds because \(5 \leq k\); the last uses the inductive hypothesis. Thus \(P(k+1)\) is true.

Therefore, by PMI, for each integer \(n \geq 5\), \(n^2 < 2^n\).

The pattern for an inequality is a chain: start at the LHS of \(P(k+1)\) and relax it one step at a time — each step justified by a rule from the table or by the inductive hypothesis — until it reaches the RHS of \(P(k+1)\). Every step must be an inequality in the same direction, or the chain says nothing.

The inductive step must hold at every \(k\)

Including the first one. An inductive step that is sound for all large \(k\) but fails at the basis value leaves a gap the chain cannot cross, and the induction proves nothing beyond the basis case — even though every individual line of the argument can look correct.

The classic demonstration is the ‘proof’ that all numbers are equal, whose inductive step is valid for every \(k \geq 2\) and false at \(k = 1\); it is picked apart in 2026-08-31-recursive-definitions. When checking an induction proof, run the inductive step at the smallest \(k\) it claims to cover and see whether it still says anything.

Strong induction

When the inductive step needs more than just the immediately preceding case — typically because a recurrence reaches back several terms, or because the step splits a number into two smaller ones it cannot control — ordinary induction is not enough. See strong-induction-and-the-well-ordering-principle for the stronger form, how many base cases it demands, and the well-ordering principle it is equivalent to.

See also

Practice Problems — the Semester-wide Set

This is the course’s one problem set for the whole semester — a single document rather than a per-week handout. 46 sections, one per pre-work video, each with the relevant textbook pages, a set of practice problems and full worked solutions. Everything is reproduced here in full, so you never need to open the PDF.

Applied Classes start in Week 2 and run to Week 13 — twelve of them, matching the “best 8 of 12” Applied Classes component worth 30% (see math1061). Each week’s class applies the previous week’s lecture content, so the material a class covers runs roughly one week behind the lecture column below. Which sections a given Applied Class draws on is not stated in any of the source material.

These problems belong to the roughly three hours of individual study each week, which are for revising the videos, readings and classes and completing the Practice Problems. They are the out-of-class half of the course, and they are not the same thing as the Assessed Exercises, which are done and submitted inside the class itself. (That class goes by two names, the Active Class and the Applied Class. They are one session.)

How each section is laid out

  1. Relevant textbook pages for Epp 4th edition and 5th (metric) edition
  2. Practice problems
  3. Solutions — full worked solutions, not just answers
  4. Occasionally a non-assessable Application block (see the table below)

Errata

The source document carries an errata page listing corrections made during semester. Last update: 24 August 2026, with two entries:

  • On page 195, the solution to 1(b) was corrected (the note in brackets had previously copied the note from 1(a)) — already applied in §45 below.
  • Question 2 in section 30 has been updated, and the original version moved to section 32 (because it refers to antisymmetry) — applied in §30 and §32 below.

If something looks wrong, post on the Ed Discussion board.

Section index

Every section is below, with its topic and textbook page ranges. Each section goes with one pre-work video (§1 with Video 001, and so on up to Video 046), and the Lecture (week) column places it in the schedule of math1061 by that video number. The Notes column links to the concept note covering that material where one exists; the rest are still ahead in the semester.

§ Topic Lecture (week) 4th ed. 5th ed. Notes
1 Logical Form L2 (wk 1) 023–029 037–043 logical-connectives
2 Logical Equivalence L2 (wk 1) 030–038 043–053 logical-equivalence-laws
3 Conditional Statements L3 (wk 1) 039–050 053–066 conditional-statements
4 Valid and Invalid Arguments L4 (wk 2) 051–063 066–079 valid-argument-forms
5 Methods for Determining Validity L5 (wk 2) 051–063 066–079 determining-argument-validity
6 Quantified Statements L5 (wk 2) 096–108 108–121 quantified-statements
7 Negation of Quantified Statements L6 (wk 2) 108–117 122–131 quantified-statements
8 Statements with Multiple Quantifiers L6 (wk 2) 117–131 131–146 quantified-statements
9 Direct Proofs and Counterexamples L7 (wk 3) 145–163 160–182 proof-techniques
10 Proof by Contradiction L8 (wk 3) 198–207 218–228 proof-techniques
11 Proof by Contraposition L9 (wk 3) 198–207 218–228 proof-techniques
12 Rational Numbers L9 (wk 3) 163–170 183–190 rational-and-irrational-numbers
13 Divisibility L10 (wk 4) 170–179 190–199 divisibility-and-factorisation
14 Modular Arithmetic L11 (wk 4) 482–485 528–531 modular-arithmetic
15 The Euclidean Algorithm L12 (wk 4) 220–224 250–254 gcd-lcm-and-euclidean-algorithm
16 Sequences L13 (wk 5) 227–244 258–275 sequences-and-summations
17 Mathematical Induction L13 (wk 5) 244–268 275–301 mathematical-induction
18 Strong Induction and the Well Ordering Principle L14 (wk 5) 268–279 301–314 strong-induction-and-the-well-ordering-principle
19 Recursive Definitions L15 (wk 6) 290–304 325–340 recursive-definitions
20 Solving Recurrence Relations L16 (wk 6) 304–328 340–364 solving-recurrence-relations
21 Set Theory Definitions L17 (wk 6) 336–351 377–391 sets-and-set-notation
22 More Definitions and Examples of Sets L17 (wk 6) 336–351 377–391 sets-and-set-notation
23 Properties of Sets L18 (wk 7) 352–366 391–407 proving-set-properties
24 Functions Defined on General Sets L18 (wk 7) 383–396 425–439 functions-on-general-sets
25 one-to-one, Onto, and Inverse Functions L19 (wk 7) 397–416 439–461
26 Composition of Functions L20 (wk 7) 416–427 461–472
27 Cardinalities L21 (wk 8) 428–441 473–486
28 Countable and Uncountable Sets L22 (wk 8) 428–441 473–486
29 Relations on Sets L23 (wk 8) 442–449 487–494
30 Reflexivity, Symmetry, Transitivity L23 (wk 8) 449–459 495–505
31 Equivalence Relations L24 (wk 9) 459–478 505–523
32 Partial Order Relations L25 (wk 9) 498–515 546–563
33 Definitions and Examples of Groups L26 (wk 9) — —
34 Elementary Properties of Groups L27 (wk 10) — —
35 Group Isomorphisms not scheduled — —
36 Definitions and Examples of Fields L28 (wk 10) — —
37 Introduction to Counting L29 (wk 11) 516–553 564–604
38 Counting Selections L30 (wk 11) 516–553 564–604
39 Introduction to Probability L31 (wk 12) 516–553 564–604
40 Binomial Coefficients L31 (wk 12) 565–591 617–641
41 Inclusion Exclusion L32 (wk 12) 545–549 595–599
42 The Pigeonhole Principle L33 (wk 12) 554–565 604–616
43 Introduction to Graph Theory L33 (wk 12) 625–660 677–697
44 Walks, Trails and Circuits L34 (wk 12) 625–660 677–697
45 Matrix Representations of Graphs L35 (wk 13) 661–675 698–712
46 Trees L36 (wk 13) 683–701 720–742

§35 Group Isomorphisms has no lecture in the published schedule — the Week 10 row jumps from L27 (V34) to L28 (V36). The section itself carries no practice problems and is marked optional and will not be assessed, so the gap is deliberate.

§28 Countable and Uncountable Sets carries its own note in the source: this content will not be assessed in MATH1061. It still has a full set of problems and solutions.

Sections 33–36 (groups and fields) list no textbook page ranges.

Non-assessable application blocks

Eight sections carry an extra block of problems, not assessable for MATH1061, mostly drawn from K. Rosen, Discrete Mathematics and its Applications, 7th ed. (1991), with two from Epp. They connect the pure material to computer science, and are reproduced in full alongside their sections.

After § Block Source
2 Applications to Computer Science — bit strings and logic circuits Rosen pp. 20–24
5 Applications of Valid and Invalid Arguments Rosen pp. 48–68
9 Applications of Proofs to Computer Science — resolution and automated theorem proving Rosen pp. 74–80
15 Trace Tables for Algorithms — the Division and Euclidean algorithms Epp §4.8 (4th ed.) / §4.10 (5th ed.)
21 Computer Representations of Sets Rosen pp. 134–138
28 Computability of Functions Rosen pp. 170–177
31 Identifiers — C identifiers as an equivalence-relation example Rosen pp. 608–618
37 Runtime — counting basic instructions Epp 5th ed. pp. 787–799

A note on the diagrams: the source PDF draws logic circuits, graphs and Hasse diagrams as figures. Those are reproduced here as explicit gate netlists and edge lists, which carry the same information in a form that renders everywhere.

§1 — Logical Form

Epp 4th ed. pp. 023–029 · 5th ed. (metric) pp. 037–043. Lecture 2 (week 1). Notes: logical-connectives.

Problems

  1. Write a truth table for the statement form \(p \wedge (\sim q \vee r)\).

  2. Suppose you know that \((\sim p \wedge q) \vee p\) is false. What can you conclude about the truth values of each of the two variables?

  3. Let \(p\) be the statement “DATAENDFLAG is off,” \(q\) the statement “ERROR equals 0,” and \(r\) the statement “SUM is less than 1,000.” Express the following sentences in symbolic notation.

    1. DATAENDFLAG is off, ERROR equals 0, and SUM is less than 1,000.

    2. DATAENDFLAG is off but ERROR is not equal to 0.

    3. DATAENDFLAG is off; however, ERROR is not 0 or SUM is greater than or equal to 1,000.

    4. DATAENDFLAG is on and ERROR equals 0 but SUM is greater than or equal to 1,000.

    5. Either DATAENDFLAG is on or it is the case that both ERROR equals 0 and SUM is less than 1,000.

Solutions

  1. The truth table is

    \(p\) \(q\) \(r\) \(\sim q\) \(\sim q \vee r\) \(p \wedge (\sim q \vee r)\)
    T T T F T T
    T T F F F F
    T F T T T T
    T F F T T T
    F T T F T F
    F T F F F F
    F F T T T F
    F F F T T F
  2. We know \((\sim p \wedge q) \vee p\) is false. A disjunction is false only when both parts are false, so we must have

    \[p \text{ is false} \quad \text{and} \quad (\sim p \wedge q) \text{ is false.}\]

    From \(p\) being false, it follows that \(\sim p\) is true. Now \((\sim p \wedge q)\) can be false only if \(q\) is false (since \(\sim p\) is already true). Hence \(q\) is false.

    Therefore \(p\) is false and \(q\) is false.

    1. \(p \wedge q \wedge r\).

    2. \(p \wedge \sim q\).

    3. \(p \wedge (\sim q \vee \sim r)\).

    4. \(\sim p \wedge q \wedge \sim r\).

    5. \(\sim p \vee (q \wedge r)\).

§2 — Logical Equivalence

Epp 4th ed. pp. 030–038 · 5th ed. (metric) pp. 043–053. Lecture 2 (week 1). Notes: logical-equivalence-laws.

Problems

    1. Use the laws of logical equivalence to show that \(p \wedge q \equiv \; \sim (\sim p \vee \sim q)\).

    2. Use the laws of logical equivalence to show that \(\sim (p \vee \sim q) \vee (\sim p \wedge \sim q) \equiv \; \sim p\).

  1. For each of the following, write down a truth table for the statement, and determine whether the statement is a tautology, a contradiction, or neither.

    1. \(((p \wedge q) \vee (q \wedge r)) \vee \sim q\)

    2. \((\sim p \vee q) \vee (p \wedge \sim q)\)

  2. Use De Morgan’s laws to write sentences equivalent to the following:

    1. It is not true that I am studying Computer Science and I am studying Engineering.

    2. I am not going to the movies this weekend or I am not going swimming this weekend.

  3. Let \(\oplus\) denote exclusive or, which means ‘… or … but not both’. The truth table for \(p \oplus q\) is as follows:

    \(p\) \(q\) \(p \oplus q\)
    T T F
    T F T
    F T T
    F F F
    1. Use a truth table to show that \((p \vee q) \wedge \sim p \equiv q \wedge \sim p\).

    2. Use a truth table to show that \((p \oplus q) \wedge r \equiv (p \wedge r) \oplus (q \wedge r)\).

    3. Use the laws of logical equivalence and the equivalence \(p \oplus q \equiv (p \vee q) \wedge \sim (p \wedge q)\) (which is the definition of \(\oplus\)) to show that

    \[(p \oplus q) \wedge r \equiv (p \wedge r) \oplus (q \wedge r).\]

Solutions

    1. We have

    \[\begin{aligned} \sim\big(\sim p \vee \sim q\big) &\equiv \; \sim(\sim p) \wedge \sim(\sim q) && \text{by De Morgan's law} \\ &\equiv p \wedge q && \text{by the double negative law (used twice).} \end{aligned}\]

    1. We have

    \[\begin{aligned} \sim (p \vee \sim q) \vee (\sim p \wedge \sim q) &\equiv (\sim p \wedge \sim \sim q) \vee (\sim p \wedge \sim q) && \text{by De Morgan's law} \\ &\equiv (\sim p \wedge q) \vee (\sim p \wedge \sim q) && \text{by the double negative law} \\ &\equiv \; \sim p \wedge (q \vee \sim q) && \text{by the distributive law} \\ &\equiv \; \sim p \wedge \mathbf{t} && \text{by the negation law} \\ &\equiv \; \sim p && \text{by the identity law.} \end{aligned}\]

    1. We have
    \(p\) \(q\) \(r\) \(p \wedge q\) \(q \wedge r\) \((p \wedge q) \vee (q \wedge r)\) \(\sim q\) \(\big((p \wedge q) \vee (q \wedge r)\big) \vee \sim q\)
    T T T T T T F T
    T T F T F T F T
    T F T F F F T T
    T F F F F F T T
    F T T F T T F T
    F T F F F F F F
    F F T F F F T T
    F F F F F F T T

    Thus the statement is neither a tautology nor a contradiction.

    1. We have
    \(p\) \(q\) \(\sim p\) \(\sim p \vee q\) \(\sim q\) \(p \wedge \sim q\) \((\sim p \vee q) \vee (p \wedge \sim q)\)
    T T F T F F T
    T F F F T T T
    F T T T F F T
    F F T T T F T

    Hence the statement is a tautology.

    1. Let \(C\) be the statement “I am studying Computer Science” and \(E\) be the statement “I am studying Engineering.” The given statement is \(\sim (C \wedge E)\). By De Morgan’s law,

    \[\sim (C \wedge E) \equiv (\sim C) \vee (\sim E).\]

    So an equivalent statement is: “I am not studying Computer Science or I am not studying Engineering.”

    1. Let \(M\) be the statement “I am going to the movies this weekend” and \(S\) be the statement “I am going swimming this weekend.” The given sentence is \((\sim M) \vee (\sim S)\). By De Morgan’s law (in reverse),

    \[(\sim M) \vee (\sim S) \equiv \; \sim (M \wedge S).\]

    So an equivalent statement is: “It is not the case that I am both going to the movies this weekend and going swimming this weekend.”

    1. We have
    \(p\) \(q\) \(p \vee q\) \(\sim p\) \((p \vee q) \wedge \sim p\) \(q \wedge \sim p\)
    T T T F F F
    T F T F F F
    F T T T T T
    F F F T F F

    Since the last two columns agree on every row, we have \((p \vee q) \wedge \sim p \equiv q \wedge \sim p\).

    1. We have
    \(p\) \(q\) \(r\) \(p \oplus q\) \((p \oplus q) \wedge r\) \(p \wedge r\) \(q \wedge r\) \((p \wedge r) \oplus (q \wedge r)\)
    T T T F F T T F
    T T F F F F F F
    T F T T T T F T
    T F F T F F F F
    F T T T T F T T
    F T F T F F F F
    F F T F F F F F
    F F F F F F F F

    The columns for \((p \oplus q) \wedge r\) and \((p \wedge r) \oplus (q \wedge r)\) match on all rows, hence \((p \oplus q) \wedge r \equiv (p \wedge r) \oplus (q \wedge r)\).

    1. To save space, multiple laws have been applied in some steps. There are many different ways to apply the laws to prove this, so your proof may look different to this one. We have

    \[\begin{aligned} &(p \wedge r) \oplus (q \wedge r) \\ &\equiv \big((p \wedge r) \vee (q \wedge r)\big) \wedge \sim \big((p \wedge r) \wedge (q \wedge r)\big) && \text{by the given equivalence (def of } \oplus) \\ &\equiv (p \vee q) \wedge r \wedge \sim \big((p \wedge r) \wedge (q \wedge r)\big) && \text{by distributive and commutative laws (for } \wedge r) \\ &\equiv (p \vee q) \wedge r \wedge \sim \big((p \wedge q) \wedge (r \wedge r)\big) && \text{by associative and commutative laws (for } \wedge) \\ &\equiv (p \vee q) \wedge r \wedge \sim \big((p \wedge q) \wedge r\big) && \text{by idempotent law} \\ &\equiv (p \vee q) \wedge r \wedge \big(\sim (p \wedge q) \vee \sim r\big) && \text{by De Morgan's law} \\ &\equiv (p \vee q) \wedge \Big(\big(r \wedge \sim (p \wedge q)\big) \vee (r \wedge \sim r)\Big) && \text{by distributive law (for } r \wedge) \\ &\equiv (p \vee q) \wedge \Big(\big(r \wedge \sim (p \wedge q)\big) \vee \mathbf{c}\Big) && \text{by negation law} \\ &\equiv (p \vee q) \wedge r \wedge \sim (p \wedge q) && \text{by identity law} \\ &\equiv (p \vee q) \wedge \sim (p \wedge q) \wedge r && \text{by associative and commutative laws (for } \wedge) \\ &\equiv (p \oplus q) \wedge r && \text{by the given equivalence (def of } \oplus) \end{aligned}\]

Application — Applications to Computer Science

Not assessable for MATH1061. Taken from K. Rosen, Discrete Mathematics and its Applications, 7th edition, 1991, pages 20–24.

Information is often represented using binary or bit strings, which are sequences consisting of zeros and ones. The length of a string is the total number of bits in it — for example \(0110110\) is a string of length 7. The logical operators \(\wedge\), \(\vee\) and \(\oplus\) extend to binary strings. The operators AND, OR and XOR are:

\(p\) \(q\) \(p \vee q\) \(p \wedge q\) \(p \oplus q\)
0 0 0 0 0
0 1 1 0 1
1 0 1 0 1
1 1 1 1 0

Propositional logic also applies to the design of computer hardware. A digital logic circuit receives input signals (either 0, the wire being off, or 1, on) and produces output signals which are also bits. Complicated digital logic circuits can be constructed from three basic gates: the OR gate, the NOT gate and the AND gate.

A combinational circuit uses these gates and obeys the following rules.

  1. Two input wires cannot be combined without a gate.
  2. A single input wire can be split partway and directed into two separate gates.
  3. An output wire from a gate can be used as input for another gate.
  4. No output wire from a gate can eventually feed back into that gate.

For example, the logical expression \((p \wedge \sim q) \vee \sim r\) has the combinational circuit

  • \(\text{NOT}(q) \to n_1\)
  • \(\text{AND}(p,\, n_1) \to a_1\)
  • \(\text{NOT}(r) \to n_2\)
  • \(\text{OR}(a_1,\, n_2) \to\) output \((p \wedge \sim q) \vee \sim r\)
Application problems
  1. Find the bitwise OR, bitwise AND and bitwise XOR of each of these pairs of binary strings.

    1. \(1011110\), \(0100001\)

    2. \(11110000\), \(10101010\)

  2. Construct a combinational circuit using the NOT, OR and AND gates that produces the outputs:

    1. \((\sim p \vee \sim r) \wedge q\)

    2. \((p \wedge \sim r) \vee (\sim q \wedge r)\)

  3. Find the output of the following combinational circuit, simplify the expression using logical laws and construct a new circuit for the simplified expression. The circuit is

    • \(\text{NOT}(p) \to n_1\)
    • \(\text{NOT}(q) \to n_2\)
    • \(\text{OR}(p,\, n_2) \to o_1\)
    • \(\text{OR}(n_1,\, n_2) \to o_2\)
    • \(\text{AND}(o_1,\, o_2) \to a_1\)
    • \(\text{OR}(a_1,\, r) \to\) output

    (both \(p\) and \(q\) have their wires split, feeding two gates each).

  4. The logical binary operation NAND, denoted \(\mid\), is defined by

    \[p \mid q \equiv \; \sim (p \wedge q).\]

    1. Write down the truth table for \(\sim (p \mid q)\). Show all working.

    2. Prove, using truth tables or otherwise, that \((p \mid p) \equiv (\sim p)\). Show all working.

    3. Prove, using truth tables or otherwise, that \(((p \mid p) \mid (q \mid q)) \equiv p \vee q\). Show all working.

Application solutions
    1. \[1011110 \text{ OR } 0100001 = 1111111,\] \[1011110 \text{ AND } 0100001 = 0000000,\] \[1011110 \text{ XOR } 0100001 = 1111111.\]

    2. \[11110000 \text{ OR } 10101010 = 11111010,\] \[11110000 \text{ AND } 10101010 = 10100000,\] \[11110000 \text{ XOR } 10101010 = 01011010.\]

    1. The circuit is
    • \(\text{NOT}(p) \to n_1\)
    • \(\text{NOT}(r) \to n_2\)
    • \(\text{OR}(n_1,\, n_2) \to o_1\)
    • \(\text{AND}(o_1,\, q) \to\) output \((\sim p \vee \sim r) \wedge q\)
    1. The circuit is
    • \(\text{NOT}(r) \to n_1\)
    • \(\text{AND}(p,\, n_1) \to a_1\)
    • \(\text{NOT}(q) \to n_2\)
    • \(\text{AND}(n_2,\, r) \to a_2\)
    • \(\text{OR}(a_1,\, a_2) \to\) output \((p \wedge \sim r) \vee (\sim q \wedge r)\)
  1. The given digital logic circuit has output \(\big((p \vee \sim q) \wedge (\sim p \vee \sim q)\big) \vee r\). Using laws of logical equivalence we have

    \[\begin{aligned} \big((p \vee \sim q) \wedge (\sim p \vee \sim q)\big) \vee r &\equiv \big(\sim q \vee (p \wedge \sim p)\big) \vee r && \text{by distributive and commutative laws} \\ &\equiv (\sim q \vee \mathbf{c}) \vee r && \text{by negation law} \\ &\equiv \; \sim q \vee r && \text{by identity law} \end{aligned}\]

    The simplification is \(\sim q \vee r\). The simplified logic circuit is

    • \(\text{NOT}(q) \to n_1\)
    • \(\text{OR}(r,\, n_1) \to\) output \(\sim q \vee r\)
    1. We have
    \(p\) \(q\) \(p \wedge q\) \(p \mid q\) \(\sim (p \mid q)\)
    T T T F T
    T F F T F
    F T F T F
    F F F T F

    Hence \(\sim (p \mid q)\) has the same truth values as \(p \wedge q\) (it is true exactly when both \(p\) and \(q\) are true).

    1. Since \(p \mid p \equiv \; \sim (p \wedge p)\), we have

    \[p \mid p \equiv \; \sim (p \wedge p) \equiv \; \sim p\]

    because \(p \wedge p \equiv p\) (idempotent law). Alternatively, by truth table:

    \(p\) \(p \wedge p\) \(p \mid p\) \(\sim p\)
    T T F F
    F F T T

    The last two columns agree, so \(p \mid p \equiv \; \sim p\).

    1. First note from part (b) that \(p \mid p \equiv \; \sim p\) and \(q \mid q \equiv \; \sim q\). Therefore

    \[\begin{aligned} (p \mid p) \mid (q \mid q) &\equiv (\sim p) \mid (\sim q) && \text{by part (b)} \\ &\equiv \; \sim\big((\sim p) \wedge (\sim q)\big) && \text{by def of NAND} \\ &\equiv p \vee q && \text{by De Morgan's law.} \end{aligned}\]

    Alternatively, by truth table:

    \(p\) \(q\) \(p \mid p\) \(q \mid q\) \((p \mid p) \mid (q \mid q)\) \(p \vee q\)
    T T F F T T
    T F F T T T
    F T T F T T
    F F T T F F

    The last two columns match, hence \((p \mid p) \mid (q \mid q) \equiv p \vee q\).

§3 — Conditional Statements

Epp 4th ed. pp. 039–050 · 5th ed. (metric) pp. 053–066. Lecture 3 (week 1). Notes: conditional-statements.

Problems

  1. Let \(p\), \(q\) and \(r\) be statements. Use the laws of logical equivalence to show that

    \[p \to (q \vee r) \equiv (p \wedge \sim q) \to r.\]

  2. Let \(p\) and \(q\) be statement variables. Are the following statements logically equivalent? Justify your answer.

    • Statement A: \((p \to q) \wedge \sim q\)
    • Statement B: \(\sim (p \wedge q)\)
  3. Write each of the following statements in the form ‘if … then …’.

    1. A sufficient condition for the warranty to be good is that you bought the computer less than a year ago.

    2. Jane gets seasick whenever she is on a boat.

  4. For each of the following, write down a truth table for the statement, and determine whether the statement is a tautology, a contradiction, or neither.

    1. \((\sim p \wedge (p \to q)) \to \sim q\)

    2. \((p \to (q \vee r)) \leftrightarrow ((p \wedge \sim q) \to r)\)

  5. Negate the following two statements.

    1. If it rains, then Sue takes her umbrella.

    2. The cakes burn if the oven temperature is too high.

  6. Recall “a sufficient condition for \(s\) is \(r\)” means \(r\) is a sufficient condition for \(s\), and that “a necessary condition for \(s\) is \(r\)” means \(r\) is a necessary condition for \(s\). Rewrite the following statements in ‘if … then …’ form:

    1. A sufficient condition for John’s team to win the championship is that it wins the rest of its games.

    2. A necessary condition for this computer program to be correct is that it not produce error messages during translation.

Solutions

  1. We have

    \[\begin{aligned} p \to (q \vee r) &\equiv \; \sim p \vee (q \vee r) && \text{by definition of implication} \\ &\equiv (\sim p \vee q) \vee r && \text{by the associative law} \\ &\equiv \; \sim (p \wedge \sim q) \vee r && \text{by De Morgan's law and double negative law} \\ &\equiv (p \wedge \sim q) \to r && \text{by definition of implication.} \end{aligned}\]

  2. Let \(A\) be the statement \((p \to q) \wedge \sim q\) and let \(B\) be the statement \(\sim (p \wedge q)\). Then

    \[\begin{aligned} A &\equiv (\sim p \vee q) \wedge \sim q && \text{by definition of implication} \\ &\equiv (\sim p \wedge \sim q) \vee (q \wedge \sim q) && \text{by distributive and commutative laws} \\ &\equiv (\sim p \wedge \sim q) \vee \mathbf{c} && \text{by negation law} \\ &\equiv \; \sim p \wedge \sim q && \text{by identity law,} \end{aligned}\]

    while

    \[B \equiv \; \sim (p \wedge q) \equiv \; \sim p \vee \sim q \qquad \text{by De Morgan's law.}\]

    Thus Statement A and Statement B are not logically equivalent because, for example, if \(p\) is false and \(q\) is true, then \(A\) is false and \(B\) is true. Hence \(A \not\equiv B\).

    1. If you bought the computer less than a year ago, then the warranty is good.

    2. If Jane is on a boat, then Jane gets seasick.

    1. Solution Option 1. The truth table for the statement is
    \(p\) \(q\) \(\sim p\) \(p \to q\) \(\sim p \wedge (p \to q)\) \(\sim q\) \((\sim p \wedge (p \to q)) \to \sim q\)
    T T F T F F T
    T F F F F T T
    F T T T T F F
    F F T T T T T

    The statement is neither a tautology nor a contradiction.

    Solution Option 2. The same truth table, written under the parts of the statement:

    \(p\) \(q\) \(\sim p\) \(\wedge\) \((p \to q)\) \(\to\) \(\sim q\)
    T T F F T T F
    T F F F F T T
    F T T T T F F
    F F T T T T T
    order ① ④ ② ⑤ ③

    The small numbers show the order in which the columns were completed. Column ④ comes from comparing columns ① and ②. Column ⑤ comes from comparing columns ③ and ④, and shows the truth values for the entire statement. This shows that this statement is neither a tautology nor a contradiction. (You don’t have to number the columns — this is just to help your understanding!)

    1. Solution Option 1. The truth tables for the LHS and RHS are:
    \(p\) \(q\) \(r\) \(q \vee r\) \(p \to (q \vee r)\) \(\sim q\) \(p \wedge \sim q\) \((p \wedge \sim q) \to r\)
    T T T T T F F T
    T T F T T F F T
    T F T T T T T T
    T F F F F T T F
    F T T T T F F T
    F T F T T F F T
    F F T T T T F T
    F F F F T T F T

    Therefore the truth table for the statement is:

    \(p\) \(q\) \(r\) \(\big(p \to (q \vee r)\big) \leftrightarrow \big((p \wedge \sim q) \to r\big)\)
    T T T T
    T T F T
    T F T T
    T F F T
    F T T T
    F T F T
    F F T T
    F F F T

    The statement is a tautology.

    Solution Option 2. Another way to represent the truth table for \((p \to (q \vee r)) \leftrightarrow ((p \wedge \sim q) \to r)\):

    \(p\) \(q\) \(r\) \(p \to\) \((q \vee r)\) \(\leftrightarrow\) \((p \wedge \sim q)\) \(\to r\)
    T T T T T T F T
    T T F T T T F T
    T F T T T T T T
    T F F F F T T F
    F T T T T T F T
    F T F T T T F T
    F F T T T T F T
    F F F T F T F T
    order ② ① ⑤ ③ ④

    Here column ⑤ comes from comparing columns ② and ④. We see (from column ⑤) that this is a tautology.

    1. Let \(R\) be the statement “it rains” and \(U\) be the statement “Sue takes her umbrella.” The given statement is \(R \to U\). Its negation is

    \[R \wedge \sim U,\]

    i.e. “It rains and Sue does not take her umbrella.”

    1. Let \(H\) be the statement “the oven temperature is too high” and \(B\) be the statement “the cakes burn.” The given statement is \(H \to B\). Its negation is

    \[H \wedge \sim B,\]

    i.e. “The oven temperature is too high and the cakes do not burn.”

    1. If John’s team wins the rest of its games, then John’s team will win the championship.

    2. If this computer program is correct, then it does not produce error messages during translation.

§4 — Valid and Invalid Arguments

Epp 4th ed. pp. 051–063 · 5th ed. (metric) pp. 066–079. Lecture 4 (week 2). Notes: valid-argument-forms.

Problems

  1. State truth values for the statement variables \(p\) and \(q\) that demonstrate that the following argument is invalid, and justify your answer.

    Premise 1: \(q\) Premise 2: \(p \to q\) Conclusion: \(p\)

  2. Find values of the statement variables \(a\), \(b\) and \(c\) that show that the following argument is invalid.

    1. \(a \to b\)
    2. \(\sim a \to c\)
    3. \(\sim b \to c\) \(\therefore\ b\)
  3. Use a truth table to show that the following argument forms are valid.

    1. Elimination:
    1. \(p \vee q\)
    2. \(\sim q\) \(\therefore\ p\)
    1. Generalisation:
    1. \(p\) \(\therefore\ p \vee q\)
  4. Use a truth table to show that the following argument is invalid.

    1. \((p \wedge q) \to \sim r\)
    2. \(p \vee \sim q\)
    3. \(\sim q \to p\) \(\therefore\ \sim r\)

Solutions

  1. To show the argument is invalid, we must find truth values for \(p\) and \(q\) such that the premises are true and the conclusion is false.

    Suppose \(p\) is false and \(q\) is true. Then Premise 1 (\(q\)) is true and Premise 2 (\(p \to q\)) is true (by definition of implication). However, the conclusion (\(p\)) is false. Therefore the argument is invalid.

  2. To show the argument is invalid, we must find truth values for \(a\), \(b\) and \(c\) such that the premises are true and the conclusion is false.

    Let \(a\) be false, \(b\) be false, and \(c\) be true. By definition of implication, all three premises are true, but the conclusion (\(b\)) is false, so the argument is invalid.

    \(p\) \(q\) premise 1: \(p \vee q\) premise 2: \(\sim q\) conclusion: \(p\)
    T T T F T
    T F T T T
    F T T F F
    F F F T F

    There is only one row of the truth table in which all premises are true (the second row), and here the conclusion is true. Thus, whenever all the premises are true, the conclusion is true, so the argument is valid.

    \(p\) \(q\) premise 1: \(p\) conclusion: \(p \vee q\)
    T T T T
    T F T T
    F T F T
    F F F F

    There are two rows for which all premises (here just the one premise) are true — the first two rows — and for each of these rows the conclusion is true. Thus, whenever all the premises are true, the conclusion is true, so the argument is valid.

  3. We compute the truth table:

    \(p\) \(q\) \(r\) \((p \wedge q) \to \sim r\) \(p \vee \sim q\) \(\sim q \to p\) Prem. 1 \(\wedge\) Prem. 2 \(\wedge\) Prem. 3 \(\sim r\)
    T T T F T T F F
    T T F T T T T T
    T F T T T T T F
    T F F T T T T T
    F T T T F T F F
    F T F T F T F T
    F F T T T F F F
    F F F T T F F T

    There are three rows for which all the premises are true, however for one of these rows the conclusion is false. In particular, taking \(p\) to be true, \(q\) to be false, and \(r\) to be true makes all the premises true but the conclusion false, so the argument is invalid.

§5 — Methods for Determining Validity

Epp 4th ed. pp. 051–063 · 5th ed. (metric) pp. 066–079. Lecture 5 (week 2). Notes: determining-argument-validity.

Problems

  1. Determine whether the following arguments are valid. If the argument is valid, prove it using rules of inference. If the argument is invalid, demonstrate it with a choice of truth values for the variables.

    \(p \to q\) \(q \to p\) \(\therefore\ p \vee q\)

    \(p\) \(p \to q\) \(\sim q \vee r\) \(\therefore\ r\)

    \(p \vee q\) \(p \to \sim q\) \(p \to r\) \(\therefore\ r\)

  2. Write the following argument in symbolic form. Then determine whether the argument is valid or not.

    An error message is received if the output is flagged. The system is running normally only if the output is not flagged. An error message is received. Therefore, the system is not running normally.

    Please use the following variables:

    \(p\) the output is flagged
    \(q\) an error message is received
    \(r\) the system is running normally
  3. Give truth values for the variables \(p\), \(q\), \(r\), \(s\) and \(w\) which demonstrate that the following argument is invalid.

    \[\big[(p \vee q) \wedge (q \to r) \wedge ((p \wedge s) \to w) \wedge (\sim r) \wedge (q \to s)\big] \to w\]

Solutions

    1. The argument is invalid. When \(p\) and \(q\) are both false, all the premises are true, but the conclusion is false. One could arrive at this example by computing the following truth table:
    \(p\) \(q\) \(p \to q\) \(q \to p\) \(p \vee q\)
    T T T T T
    T F F T T
    F T T F T
    F F T T F

    (In the last row, the premises are true, but the conclusion is false.)

    1. The argument is valid, and we prove this with rules of inference:
    1. \(p\) Premise
    2. \(p \to q\) Premise
    3. \(\sim q \vee r\) Premise
    4. \(q\) from 1 and 2 by Modus Ponens
    5. \(r\) from 3 and 4 by elimination (and the double negative law)

    Therefore the argument is valid.

    1. Suppose the argument is invalid. Then it is possible to find truth values for the statement variables that make all the premises true and the conclusion false. For the conclusion to be false, we have \(r\) is false. Since \(r\) is false but the third premise (\(p \to r\)) is true, we must have \(p\) is false. Since \(p\) is false, this means the second premise (\(p \to \sim q\)) is also true. Now since \(p\) is false but the first premise (\(p \vee q\)) is true, we must have \(q\) is true.

    Thus, with these truth values for the three variables, we have that all the premises are true, but the conclusion is false. So the argument is invalid.

  1. The argument has the symbolic form:

    1. \(p \to q\) (An error message is received if the output is flagged.)
    2. \(r \to \sim p\) (The system is running normally only if the output is not flagged.)
    3. \(q\) (An error message is received.)
    \(\therefore\) \(\sim r\) (The system is not running normally.)

    Suppose the argument is invalid. Then it is possible to find truth values for the statement variables that make all the premises true and the conclusion false. For the conclusion to be false, we have \(r\) is true. Since \(r\) is true, and premise 2 is true, we must have \(\sim p\) true, that is \(p\) is false. Since premise 3 is true, we have \(q\) is true.

    Thus, with these truth values for the three variables, we have that all the premises are true, but the conclusion is false. So the argument is invalid.

  2. The argument being valid is the same as this overall implication statement being a tautology.

    \[\big[(p \vee q) \wedge (q \to r) \wedge ((p \wedge s) \to w) \wedge (\sim r) \wedge (q \to s)\big] \to w\]

    We can write the argument as follows:

    1. \(p \vee q\)
    2. \(q \to r\)
    3. \((p \wedge s) \to w\)
    4. \(\sim r\)
    5. \(q \to s\) \(\therefore\ w\)

    Suppose the argument is invalid. Then it is possible to find truth values for the statement variables that make all the premises true and the conclusion false. For the conclusion to be false, we have \(w\) is false. Since premise 4 is true we have \(r\) is false. Since premise 2 is true and \(r\) is false, we must have \(q\) is false. Since premise 1 is true and \(q\) is false we have \(p\) is true. Since premise 3 is true and \(w\) is false, we must have \((p \wedge s)\) false, and since \(p\) is true, this means \(s\) is false. These truth values also mean premise 5 is true.

    Thus with these truth values for the variables, we have that all the premises are true, but the conclusion is false. So the argument is invalid.

Application — Applications of Valid and Invalid Arguments

Not assessable for MATH1061. Taken from K. Rosen, Discrete Mathematics and its Applications, 7th edition, 1991, pages 48–68.

Application problems
  1. Translating English sentences into logical expressions is an essential part of specifying both hardware and software systems. Software engineers must take requirements in English and produce precise, unambiguous specifications that can be used as the basis for system development.

    Consider the following predicates and domains:

    Predicates:

    \(F(x, y)\) disk \(x\) has more than \(y\) kilobytes of free space.
    \(S(m)\) mail message \(m\) can be saved.
    \(A(x)\) alert \(x\) is active.
    \(Q(m)\) message \(m\) is queued.
    \(T(m)\) message \(m\) is transmitted.
    \(D(s)\) the diagnostic monitor tracks the status of system \(s\).

    Domains:

    \(M\) the nonempty set of all messages.
    \(X\) the set of all disks.
    \(Y\) the set of all alerts.
    \(S\) the set of all systems.

    Express the system specifications below using quantifiers and logical connectives:

    1. At least one mail message, among the nonempty set of messages, can be saved if there is a disk with more than 10 kilobytes of free space.

    2. Whenever there is an active alert, all queued messages are transmitted.

    3. The diagnostic monitor tracks the status of all systems except the main console.

  2. System specifications should be consistent; that is, they should not contain conflicting requirements that could be used to derive a contradiction. When specifications are not consistent, there would be no way to develop a system that satisfies all specifications. Logically, a system is consistent if it is possible for all the premises to be simultaneously true.

    Use logical expressions and methods of logical deduction to determine whether the following specifications are consistent.

    1. The system is in multiuser state if and only if it is operating normally. If the system is operating normally, the kernel is functioning. The kernel is not functioning or the system is in interrupt mode. If the system is not in multiuser state, then it is in interrupt mode. The system is not in interrupt mode.

    2. The router can send packets to the edge system only if it supports the new address space. For the router to support the new address space it is necessary that the latest software release be installed. The router can send packets to the edge system if the latest software release is installed. The router does not support the new address space.

Application solutions
    1. \((\exists x \in X \text{ s.t. } F(x, 10)) \to (\exists m \in M \text{ s.t. } S(m))\).

    2. \((\exists y \in Y \text{ s.t. } A(y)) \to (\forall m \in M,\ (Q(m) \to T(m)))\).

    3. Let \(c \in S\) denote the main console. \(\forall s \in S,\ (s \neq c \to D(s))\).

    1. Define the following variables:
    \(m\) the system is in multiuser state
    \(n\) the system is operating normally
    \(k\) the kernel is functioning
    \(i\) the system is in interrupt mode

    Now we can express the specifications as follows:

    1. \(m \leftrightarrow n\)
    2. \(n \to k\)
    3. \(\sim k \vee i\)
    4. \(\sim m \to i\)
    5. \(\sim i\)

    Solution Option 1. If these specifications are consistent, then we must be able to find truth values for each of the variables that makes all of the specifications true. Assume the specifications are consistent (so we assume each specification is true). For specification 5 to be true we know \(i\) is false. For specification 4 to be true, with \(i\) false, we know \(\sim m\) is false so \(m\) is true. For premise 1 to be true, with \(m\) true, we have \(n\) is true. For specification 2 to be true, and since \(n\) is true, we have \(k\) is true. However, now \(\sim k\) is false and \(i\) is false, which means specification 3 is false. This contradicts our assumption that all the specifications are true, and so it is impossible for the system to be consistent. Therefore, the system is not consistent.

    Solution Option 2. Suppose the specifications are all true. Using laws of inference, we have

    1. \(m \leftrightarrow n\) specification
    2. \(n \to k\) specification
    3. \(\sim k \vee i\) specification
    4. \(\sim m \to i\) specification
    5. \(\sim i\) specification
    6. \(\sim k\) from 3 and 5 by elimination (and double negative law)
    7. \(\sim n\) from 2 and 6 by Modus Tollens
    8. \((m \to n) \wedge (n \to m)\) from 1 by definition of \(\leftrightarrow\)
    9. \(m \to n\) from 8 by specialisation
    10. \(\sim m\) from 9 and 7 by Modus Tollens
    11. \(i\) from 4 and 10 by Modus Ponens

    But now we have both \(i\) and \(\sim i\) being true, which is a contradiction. This contradicts our assumption that all the specifications are true, and so it is impossible for the system to be consistent. Therefore, the system is not consistent.

    1. Define the following variables:
    \(r\) the router can send packets to the edge system
    \(a\) the router supports the new address space
    \(s\) the latest software release is installed

    Now we can express the specifications as follows:

    1. \(r \to a\)
    2. \(a \to s\)
    3. \(s \to r\)
    4. \(\sim a\)

    Assume the specifications are consistent, so each specification is true. For specification 4 to be true, \(\sim a\) is true, so \(a\) is false. For specification 1 to be true with \(a\) false we must have \(r\) is false. Now, since \(r\) is false and specification 3 is true, we have \(s\) is false. We double-check that specification 2 is true, which it is because \(a\) is false. Thus, we have found truth values of the statement variables that make all of the specifications true, so the system is consistent.

§6 — Quantified Statements

Epp 4th ed. pp. 096–108 · 5th ed. (metric) pp. 108–121. Lecture 5 (week 2). Notes: quantified-statements.

Problems

  1. Rewrite each of the following in the form “\(\forall \underline{\phantom{xx}},\ \underline{\phantom{xx}}\).” Here, let \(D\) be the set of all dinosaurs, and let \(L\) be the set of all logicians.

    1. All dinosaurs are extinct.

    2. Every real number is positive, negative, or zero.

    3. No irrational numbers are integers.

    4. No logicians are lazy.

    5. The number \(2{,}147{,}581{,}953\) is not equal to the square of any integer.

    6. The number \(-1\) is not equal to the square of any real number.

  2. Rewrite each of the following in the form “\(\exists \underline{\phantom{xx}}\) such that \(\underline{\phantom{xx}}\).” Here, let \(X\) be the set of all exercises.

    1. Some exercises have answers.

    2. Some real numbers are rational.

  3. Determine whether the following statements are true or false. Justify your reasoning.

    1. \(\forall x \in \mathbb{R},\ x \geq \frac{1}{x}\).

    2. \(\forall a \in \mathbb{Z},\ \frac{a-1}{a} \notin \mathbb{Z}\).

    3. \(\forall m, n \in \mathbb{Z}^+,\ mn \geq m + n\).

Solutions

    1. \(\forall x \in D,\ x\) is extinct.

    2. \(\forall x \in \mathbb{R},\ (x > 0 \vee x < 0 \vee x = 0)\).

    3. \(\forall x \in \mathbb{R},\ (x \text{ is irrational} \to x \notin \mathbb{Z})\).

    4. \(\forall x \in L,\ x\) is not lazy.

    5. \(\forall x \in \mathbb{Z},\ x^2 \neq 2147581953\).

    6. \(\forall x \in \mathbb{R},\ -1 \neq x^2\).

    1. \(\exists x \in X\) such that \(x\) has an answer.

    2. \(\exists x \in \mathbb{R}\) such that \(x \in \mathbb{Q}\).

    1. False. Counterexample: take \(x = \frac{1}{2}\). Then \(x = \frac{1}{2}\) but \(\frac{1}{x} = 2\), so \(x \geq \frac{1}{x}\) is false.

    2. False. Counterexample: take \(a = 1\). Then \(\frac{a-1}{a} = \frac{0}{1} = 0 \in \mathbb{Z}\).

    3. False. Counterexample: take \(n = m = 1\). Then \(mn = 1\) but \(m + n = 2\).

§7 — Negation of Quantified Statements

Epp 4th ed. pp. 108–117 · 5th ed. (metric) pp. 122–131. Lecture 6 (week 2). Notes: quantified-statements.

Problems

  1. Write the negation of each of the following statements, and then determine whether the original or the negation is true:

    1. \(\forall x \in \mathbb{R}\), if \(x^2 > 9\) then \(x \geq 2\).

    2. \(\forall n \in \mathbb{Z}^+\), \(n\) is either prime or composite.

    3. \(\exists x, y \in \mathbb{R}\) such that \(x \notin \mathbb{Z}\) and \(y \notin \mathbb{Z}\) but \(xy \in \mathbb{Z}\).

  2. Write down the negation of each of the following statements:

    1. All fire engines are red.

    2. Some apples are ripe.

    3. A triangle with equal sides exists.

Solutions

    1. Original: \(\forall x \in \mathbb{R}\), if \(x^2 > 9\) then \(x \geq 2\). Negation: \(\exists x \in \mathbb{R}\) such that \(x^2 > 9\) and \(x < 2\).

    The original statement is false. For instance, take \(x = -4\). Then \(x^2 > 9\) but \(x \not\geq 2\).

    1. Original: \(\forall n \in \mathbb{Z}^+\), \(n\) is either prime or composite. Negation: \(\exists n \in \mathbb{Z}^+\) such that \(n\) is neither prime nor composite.

    The original statement is false. For instance, consider \(n = 1\), which is neither prime nor composite.

    1. Original: \(\exists x, y \in \mathbb{R}\) such that \(x \notin \mathbb{Z}\) and \(y \notin \mathbb{Z}\) but \(xy \in \mathbb{Z}\). Negation: \(\forall x, y \in \mathbb{R}\), if \(x \notin \mathbb{Z}\) and \(y \notin \mathbb{Z}\) then \(xy \notin \mathbb{Z}\).

    The original statement is true. For instance, take \(x = \sqrt{2}\) and \(y = \sqrt{2}\). Then \(x \notin \mathbb{Z}\) and \(y \notin \mathbb{Z}\), but \(xy = 2 \in \mathbb{Z}\).

    1. Original: All fire engines are red. Negation: There exists a fire engine which is not red.

    2. Original: Some apples are ripe. Negation: All apples are not ripe.

    3. Original: A triangle with equal sides exists. Negation: There does not exist a triangle with equal sides.

§8 — Statements with Multiple Quantifiers

Epp 4th ed. pp. 117–131 · 5th ed. (metric) pp. 131–146. Lecture 6 (week 2). Notes: quantified-statements.

Problems

  1. Write down the negation of each of the following statements. Then determine whether the statement is true, or whether its negation is true.

    1. \(\forall x, y \in \mathbb{Z}\), \(\exists z \in \mathbb{Z}\) such that \(z = x - y\).

    2. \(\forall x \in \mathbb{R}\), \(\exists y \in \mathbb{R}\) such that \(x^2 - xy - 2y^2 = 0\).

    3. \(\exists x \in \mathbb{Z}^+\) such that \(\forall y \in \mathbb{Z}^+\), \(x \geq y\).

    4. \(\forall r, s \in \mathbb{R}\), if \(rs \in \mathbb{Q}\), then \(r \in \mathbb{Q}\) and \(s \in \mathbb{Q}\).

  2. For each of the following English sentences:

    1. For any rational number \(x\) and any negative integer \(y\) there exists a real number that is greater than the product of \(x\) and \(y\).

    2. For every real number \(x\), there is a real number \(y\) for which \(3y = x\).

    do the following:

    1. Write the statement in symbolic form.

    2. Write the negation of the statement in symbolic form.

    3. Determine which of the statement or its negation is true.

Solutions

    1. Original: \(\forall x, y \in \mathbb{Z}\), \(\exists z \in \mathbb{Z}\) such that \(z = x - y\). Negation: \(\exists x \in \mathbb{Z}\ \exists y \in \mathbb{Z}\) such that \(\forall z \in \mathbb{Z}\), \(z \neq x - y\).

    The original statement is true (and the negation is false). For any given \(x, y \in \mathbb{Z}\), choose \(z = x - y \in \mathbb{Z}\).

    1. Original: \(\forall x \in \mathbb{R}\), \(\exists y \in \mathbb{R}\) such that \(x^2 - xy - 2y^2 = 0\). Negation: \(\exists x \in \mathbb{R}\) s.t. \(\forall y \in \mathbb{R}\), \(x^2 - xy - 2y^2 \neq 0\).

    The original statement is true (and the negation is false). For any \(x \in \mathbb{R}\), taking \(y = \frac{x}{2}\) gives

    \[x^2 - x\left(\frac{x}{2}\right) - 2\left(\frac{x}{2}\right)^2 = x^2 - \frac{x^2}{2} - 2 \cdot \frac{x^2}{4} = x^2 - \frac{x^2}{2} - \frac{x^2}{2} = 0.\]

    1. Original: \(\exists x \in \mathbb{Z}^+\) such that \(\forall y \in \mathbb{Z}^+\), \(x \geq y\). Negation: \(\forall x \in \mathbb{Z}^+\ \exists y \in \mathbb{Z}^+\) such that \(x < y\).

    The negation is true (and the original is false). For any \(x \in \mathbb{Z}^+\), take \(y = x + 1\). Then \(y\) is an integer satisfying the inequality \(x < y\).

    1. Original: \(\forall r, s \in \mathbb{R}\), if \(rs \in \mathbb{Q}\), then \(r \in \mathbb{Q}\) and \(s \in \mathbb{Q}\). Negation: \(\exists r, s \in \mathbb{R}\) such that \(rs \in \mathbb{Q}\) and (\(r \notin \mathbb{Q}\) or \(s \notin \mathbb{Q}\)).

    The negation is true (and the original is false). Take \(r = \sqrt{2}\) and \(s = \sqrt{2}\). Then \(rs = 2 \in \mathbb{Q}\) but \(r, s \notin \mathbb{Q}\).

      1. Original: \(\forall x \in \mathbb{Q},\ \forall y \in \mathbb{Z}^{<0},\ \exists z \in \mathbb{R}\) s.t. \(z > xy\).
    1. Negation: \(\exists x \in \mathbb{Q}\ \exists y \in \mathbb{Z}^{<0}\) s.t. \(\forall z \in \mathbb{R}\), \(z \leq xy\).

    2. The original statement is true. For any given \(x \in \mathbb{Q}\) and \(y \in \mathbb{Z}^{<0}\), take \(z = xy + 1\). This is a real number satisfying the inequality \(z > xy\).

      1. Original: \(\forall x \in \mathbb{R},\ \exists y \in \mathbb{R}\) s.t. \(3y = x\).
    1. Negation: \(\exists x \in \mathbb{R}\) s.t. \(\forall y \in \mathbb{R}\), \(3y \neq x\).

    2. The original statement is true. Given any \(x \in \mathbb{R}\), choose \(y = \frac{x}{3}\). This is a real number satisfying the equality \(3y = x\).

§9 — Direct Proofs and Counterexamples

Epp 4th ed. pp. 145–163 · 5th ed. (metric) pp. 160–182. Lecture 7 (week 3). Notes: proof-techniques.

Problems

  1. For each of the following statements, either prove the statement or disprove it using a counterexample.

    1. The sum of any pair of even integers is even.

    2. There exist positive integers \(a\), \(b\) and \(c\) such that \(a^2 + b^2 = c^2\).

    3. There is an even integer \(n\) such that \(5n - 4\) is prime.

    4. \(\forall n \in \mathbb{Z}^+\), if \(n \geq 4\) then \(2n^2 - 5n + 2\) is composite.

    5. \(\forall x, y \in \mathbb{R}\), if \(y^2 > x^2\) then \(y > x\).

  2. Find a counterexample that disproves the following statement:

    \[\text{For all } x \in \mathbb{R},\ \lfloor x^2 \rfloor = \lfloor x \rfloor^2.\]

    (Recall that \(\lfloor \cdot \rfloor\) denotes the floor function.)

  3. Find the errors in the following proofs and rewrite the proof correctly. Here is a list of common errors made when writing proofs:

    • Arguing from examples
    • Using the same variable to mean two different things
    • Jumping to the conclusion
    • Assuming what is to be proved
    • Confusion between what is known and what is to be shown
    • Use of the word any instead of some
    • Misuse of the word if
    1. The difference between any odd integer and any even integer is odd.

    Proof. Let \(m\) be any odd integer, then \(m = 2k + 1\) where \(k \in \mathbb{Z}\) and let \(n\) be any even integer, so \(n = 2k\) for any \(k \in \mathbb{Z}\). Then \(m - n = 2k + 1 - 2k = 1\) and 1 is odd so the statement is true.

    1. If \(a\) and \(b\) are even integers, then \(4a + 2b\) is even.

    Proof. We have \(a = 2k\) and \(b = 2j\) for some \(k, j \in \mathbb{Z}\). Then we want to show that \(4a + 2b\) is even, so \[\begin{aligned} 4a + 2b &= 2\ell \\ 4(2k) + 2(2j) &= 2\ell \\ 8k + 4j &= 2\ell \\ 2(4k + 2j) &= 2\ell \end{aligned}\] therefore the statement is true.

    1. \(\forall x \in \mathbb{R}\), \(\exists y \in \mathbb{R}\) such that \(xy = 1\).

    Proof. This is true, for example if \(x = 10\) then let \(y = \frac{1}{10}\) and \(xy = 10 \cdot \frac{1}{10} = 1\). Therefore the statement is true.

    1. \(\forall n \in \mathbb{Z}\), if \(n > 0\) then \(n^2 + 2n + 1\) is composite.

    Proof. Suppose \(n\) is an integer and \(n > 0\), then we want to show that \(n^2 + 2n + 1\) is composite, so this means \[n^2 + 2n + 1 = rs\] where \[1 < r, s < n^2 + 2n + 1\] Clearly neither \(r\) nor \(s\) are equal to 1 so \(n^2 + 2n + 1\) is not prime and hence must be composite.

Solutions

    1. The sum of any pair of even integers is even.

    Proof. Let \(m, n \in \mathbb{Z}\) be even. Then \(m = 2k\) and \(n = 2\ell\) for some \(k, \ell \in \mathbb{Z}\). Hence \[m + n = 2k + 2\ell = 2(k + \ell),\] which is even. Therefore, the sum of any pair of even integers is even. \(\square\)

    1. There exist positive integers \(a\), \(b\) and \(c\) such that \(a^2 + b^2 = c^2\).

    Proof. Consider the positive integers \(a = 3\), \(b = 4\), \(c = 5\). Then \(a^2 + b^2 = 9 + 16 = 25 = c^2\). \(\square\)

    1. There is an even integer \(n\) such that \(5n - 4\) is prime. This statement is false. We prove the negation: \(\forall n \in \mathbb{Z}^{\text{even}}\), \(5n - 4\) is not prime.

    Proof. Suppose that \(n\) is an even integer. Thus \(n = 2k\) for some integer \(k\). Hence \[5n - 4 = 5(2k) - 4 = 10k - 4 = 2(5k - 2).\] By definition, this means that \(5n - 4\) is even. The only even prime is 2. If \(2(5k-2) = 2\), then \(5k - 2 = 1\) and hence \(k = \frac{3}{5}\) which is not an integer. Therefore \(5n - 4\) is not a prime number. \(\square\)

    1. \(\forall n \in \mathbb{Z}^+\), if \(n \geq 4\) then \(2n^2 - 5n + 2\) is composite.

    Proof. Suppose \(n\) is a positive integer and \(n \geq 4\). Let us factor \[2n^2 - 5n + 2 = (2n - 1)(n - 2).\] Since \(n \geq 4\), we have \(2n - 1 \geq 7\) and \(n - 2 \geq 2\), so both factors are integers \(> 1\). Hence the expression is composite for all \(n \geq 4\). \(\square\)

    1. \(\forall x, y \in \mathbb{R}\), if \(y^2 > x^2\) then \(y > x\). This statement is false. Take \(x = 1\) and \(y = -2\). Then \(y^2 = 4 > x^2 = 1\), but \(y = -2 \not> 1 = x\).
  1. Take \(x = \frac{3}{2}\). Then \(\lfloor x^2 \rfloor = \left\lfloor \frac{9}{4} \right\rfloor = 2\), while \(\lfloor x \rfloor^2 = \left\lfloor \frac{3}{2} \right\rfloor^2 = 1^2 = 1\). Hence \(\lfloor x^2 \rfloor \neq \lfloor x \rfloor^2\), so the statement is false.

    1. Statement. The difference between any odd integer and any even integer is odd.

    Proof. Let \(m\) be any odd integer, then \(m = 2k + 1\) where \(k \in \mathbb{Z}\) and let \(n\) be any even integer, so \(n = 2k\) for any \(k \in \mathbb{Z}\). [Error: using \(k\) for both \(m\) and \(n\) means that the statement is only proved when \(m = n + 1\).] Then \(m - n = 2k + 1 - 2k = 1\) and 1 is odd so the statement is true.

    Correct proof. Let \(m\) be any odd integer and let \(n\) be any even integer. Then there exist \(k, \ell \in \mathbb{Z}\) such that \(m = 2k + 1\) and \(n = 2\ell\). Hence \[m - n = (2k + 1) - 2\ell = 2(k - \ell) + 1,\] which is an odd integer by definition. \(\square\)

    1. Statement. If \(a\) and \(b\) are even integers, then \(4a + 2b\) is even. [Error: the proof forgets to state its assumptions.]

    Proof. We have \(a = 2k\) and \(b = 2j\) for some \(k, j \in \mathbb{Z}\). Then we want to show that \(4a + 2b\) is even, so \[\begin{aligned} 4a + 2b &= 2\ell && \textbf{[Error: this line assumes what is to be proved.]} \\ 4(2k) + 2(2j) &= 2\ell \\ 8k + 4j &= 2\ell \\ 2(4k + 2j) &= 2\ell \end{aligned}\] therefore the statement is true.

    Correct proof. Suppose \(a\) and \(b\) are even integers. Then \(a = 2k\) and \(b = 2j\) for some \(k, j \in \mathbb{Z}\). Therefore \[4a + 2b = 4(2k) + 2(2j) = 8k + 4j = 2(4k + 2j).\] Since \(k, j \in \mathbb{Z}\), we have \((4k + 2j) \in \mathbb{Z}\) and so, by definition, \(4a + 2b\) is even. \(\square\)

    1. Statement. \(\forall x \in \mathbb{R}\), \(\exists y \in \mathbb{R}\) such that \(xy = 1\).

    Proof. This is true, for example if \(x = 10\) then let \(y = \frac{1}{10}\) and \(xy = 10 \cdot \frac{1}{10} = 1\). Therefore the statement is true. [Error: the proof argues from a single example, which cannot establish a universal statement. Moreover, the statement is false.]

    Correct disproof. The statement is false. As a counterexample, let \(x = 0\). We know that \(0 \cdot y = 0\) for all \(y \in \mathbb{R}\) and so there is no real number \(y\) such that \(0 \cdot y = 1\). \(\square\)

    1. Statement. \(\forall n \in \mathbb{Z}\), if \(n > 0\) then \(n^2 + 2n + 1\) is composite.

    Proof. Suppose \(n\) is an integer and \(n > 0\), then we want to show that \(n^2 + 2n + 1\) is composite, so this means \[n^2 + 2n + 1 = rs \qquad \textbf{[Error: forgets to define } r, s \textbf{ — e.g. for some } r, s \in \mathbb{Z}\textbf{]}\] where \[1 < r, s < n^2 + 2n + 1\] Clearly neither \(r\) nor \(s\) are equal to 1 so \(n^2 + 2n + 1\) is not prime and hence must be composite. [Error: this is jumping to the conclusion without showing what needs to be shown. The expression we need to show (the definition of composite) has been given but it has not been shown.]

    Correct proof. Let \(n \in \mathbb{Z}\) with \(n > 0\). Then we have \[n^2 + 2n + 1 = (n+1)(n+1).\] Now since \(0 < n < n^2 + 2n\), we must have that \(1 < n + 1 < n^2 + 2n + 1\). So by letting \(r = s = n + 1\) we have shown that \[n^2 + 2n + 1 = rs\] for integers \(1 < r, s < n^2 + 2n + 1\). By definition this means \(n^2 + 2n + 1\) is composite. \(\square\)

Application — Applications of Proofs to Computer Science

Not assessable for MATH1061. Taken from K. Rosen, Discrete Mathematics and its Applications, 7th edition, 1991, pages 74–80.

Computer programs have been developed to automate the task of reasoning and proving theorems. Many of these programs make use of a rule of inference known as resolution. This rule of inference is based on the tautology

\[\big((p \vee q) \wedge (\sim p \vee r)\big) \to (q \vee r) \tag{1}\]

Resolution can be used to build automatic theorem proving systems.

To construct proofs using resolution as the only rule of inference, the hypotheses and conclusion must be expressed as clauses. Clauses are logical expressions which only feature disjunctions of variables or their negations (i.e. expressions which only use \(\vee\), \(\sim\), so they do not use \(\wedge\), \(\to\)). To express any statement as a clause, we can use the laws of logical equivalence.

Application problems
    1. Using a truth table, show that expression (1) is a tautology.

    2. Write the following expressions as clauses.

    1. \(\sim (\sim r \wedge w)\)

    2. \(r \to u\)

    3. \(\sim (u \wedge w)\)

    1. Use resolution only to show that the following argument is valid:

    If it is raining, then Yvette has their umbrella. It is not the case that Yvette has their umbrella and gets wet. It is not the case that Yvette gets wet while it is not raining. Therefore, Yvette does not get wet.

Application solutions
    1. The truth table is
    \(p\) \(q\) \(r\) \(p \vee q\) \(\sim p \vee r\) \((p \vee q) \wedge (\sim p \vee r)\) \(q \vee r\) \(\big((p \vee q) \wedge (\sim p \vee r)\big) \to (q \vee r)\)
    T T T T T T T T
    T T F T F F T T
    T F T T T T T T
    T F F T F F F T
    F T T T T T T T
    F T F T T T T T
    F F T F T F T T
    F F F F T F F T

    Since the last column is always T, the statement is a tautology.

    1. Clause forms:
    1. \(\sim (\sim r \wedge w) \equiv r \vee \sim w\) (by De Morgan’s law and double negative law).

    2. \(r \to u \equiv \; \sim r \vee u\) (by definition of implication).

    3. \(\sim (u \wedge w) \equiv \; \sim u \vee \sim w\) (by De Morgan’s law).

    1. Define the following variables:
    \(r\) it is raining
    \(u\) Yvette has their umbrella
    \(w\) Yvette gets wet

    Then we can express the argument as

    1. \(r \to u\)
    2. \(\sim (u \wedge w)\)
    3. \(\sim (\sim r \wedge w)\) \(\therefore\ \sim w\)

    Using the equivalences in (b), we can rewrite this as

    1. \(\sim r \vee u\)
    2. \(\sim u \vee \sim w\)
    3. \(r \vee \sim w\)
    4. \(\sim r \vee \sim w\) From 1 and 2 by Resolution
    5. \(\sim w\) From 3 and 4 by Resolution.

§10 — Proof by Contradiction

Epp 4th ed. pp. 198–207 · 5th ed. (metric) pp. 218–228. Lecture 8 (week 3). Notes: proof-techniques.

Problems

  1. Prove the following statement using a proof by contradiction:

    \[\text{For integers } a \text{ and } b,\ \text{if } 6a + 3b \text{ is odd, then } b \text{ is odd.}\]

  2. Use a proof by contradiction to prove the following statement:

    \[\text{For all positive integers } x \text{ and } y,\ x^2 - y^2 \neq 1.\]

Solutions

  1. For integers \(a\) and \(b\), if \(6a + 3b\) is odd, then \(b\) is odd.

    Proof. Suppose the statement is false, that is, suppose that there exist integers \(a\) and \(b\) for which \(6a + 3b\) is odd and \(b\) is even.

    Since \(b\) is even, \(b = 2k\) for some integer \(k\). Thus \[6a + 3b = 6a + 3(2k) = 6a + 6k = 2(3a + 3k).\] Since \(a, k \in \mathbb{Z}\), \(3a + 3k \in \mathbb{Z}\) so \(6a + 3b\) is even. This contradicts our assumption that \(6a + 3b\) is odd, so our assumption was incorrect and hence the original statement is true. \(\square\)

  2. For all positive integers \(x\) and \(y\), \(x^2 - y^2 \neq 1\).

    Proof. Suppose, for contradiction, that \(x\) and \(y\) are positive integers satisfying \(x^2 - y^2 = 1\). Now, by factoring, we have \[1 = x^2 - y^2 = (x - y)(x + y).\] Since \(x\) and \(y\) are positive integers, we must have \(x - y = 1\) and \(x + y = 1\). This means \(2x = 2\), so \(x = 1\). But then we must have \(y = 0\) which is not a positive integer; this is a contradiction. Therefore the original statement is true. \(\square\)

§11 — Proof by Contraposition

Epp 4th ed. pp. 198–207 · 5th ed. (metric) pp. 218–228. Lecture 9 (week 3). Notes: proof-techniques.

Problems

  1. Use proof by contraposition to prove the following statement:

    \[\text{For all integers } a \text{ and } b,\ \text{if } (ab)^2 \text{ is odd then } a \text{ is odd and } b \text{ is odd.}\]

  2. Prove the following statement:

    \[\text{For any integer } n,\ n^2 \text{ is odd if and only if } n \text{ is odd.}\]

    Use a direct proof for the reverse implication (\(\Leftarrow\)), and a proof by contraposition for the forward implication (\(\Rightarrow\)).

  3. Use proof by contraposition to prove the following statement:

    \[\text{For all integers } x \text{ and } y,\ \text{if } x^2(y^2 - 2y) \text{ is odd then } x \text{ and } y \text{ are odd.}\]

Solutions

  1. For all integers \(a\) and \(b\), if \((ab)^2\) is odd then \(a\) is odd and \(b\) is odd.

    Proof. We prove the contrapositive statement, which is:

    For all integers \(a\) and \(b\), if \(a\) is even or \(b\) is even, then \((ab)^2\) is even.

    Let \(a\) and \(b\) be integers and suppose \(a\) is even or \(b\) is even. Without loss of generality, suppose \(a\) is even, so \(a = 2k\) for some \(k \in \mathbb{Z}\). Then \(ab = 2kb\) is even, and therefore \((ab)^2 = 4k^2b^2 = 2(2k^2b^2)\). Since \(k, b \in \mathbb{Z}\), we know \(2k^2b^2 \in \mathbb{Z}\), so \((ab)^2\) is even by definition. \(\square\)

  2. For any integer \(n\), \(n^2\) is odd if and only if \(n\) is odd.

    Proof. To prove this “if and only if” statement, we must prove each of the following two statements:

    (\(\Leftarrow\)) For any integer \(n\), if \(n\) is odd, then \(n^2\) is odd. (\(\Rightarrow\)) For any integer \(n\), if \(n^2\) is odd, then \(n\) is odd.

    (\(\Leftarrow\)) Let \(n\) be an integer and suppose \(n\) is odd. Then \(n = 2k + 1\) for some \(k \in \mathbb{Z}\). Hence \(n^2 = (2k+1)^2 = 2(2k^2 + 2k) + 1\). Since \(k \in \mathbb{Z}\), we have \(2k^2 + 2k \in \mathbb{Z}\) and hence \(n^2\) is odd by definition.

    (\(\Rightarrow\)) The contrapositive is: for any integer \(n\), if \(n\) is even, then \(n^2\) is even. Let \(n\) be an integer and suppose \(n\) is even. Then \(n = 2k\) for some \(k \in \mathbb{Z}\). Hence \(n^2 = 4k^2 = 2(2k^2)\). Since \(k \in \mathbb{Z}\), we have \(2k^2 \in \mathbb{Z}\), so \(n^2\) is even by definition.

    Therefore, for all integers \(n\), \(n^2\) is odd if and only if \(n\) is odd. \(\square\)

  3. For all integers \(x\) and \(y\), if \(x^2(y^2 - 2y)\) is odd then \(x\) and \(y\) are odd.

    Proof. We will prove the contrapositive statement, which is:

    For all integers \(x\) and \(y\), if \(x\) is even or \(y\) is even, then \(x^2(y^2 - 2y)\) is even.

    Let \(x, y \in \mathbb{Z}\) and suppose \(x\) or \(y\) is even. We consider the cases of \(x\) even or \(y\) even individually.

    Case 1. Suppose \(x\) is even. Then \(x = 2k\) for some \(k \in \mathbb{Z}\). Hence \[x^2(y^2 - 2y) = (2k)^2(y^2 - 2y) = 2(2k^2)(y^2 - 2y).\] Since \(k, y \in \mathbb{Z}\), we have \((2k^2)(y^2 - 2y) \in \mathbb{Z}\), and hence \(x^2(y^2-2y)\) is even, by definition.

    Case 2. Suppose \(y\) is even. Then \(y = 2\ell\) for some \(\ell \in \mathbb{Z}\). Hence \[x^2(y^2 - 2y) = x^2\big((2\ell)^2 - 2(2\ell)\big) = 2\big(x^2(2\ell^2 - 2\ell)\big).\] Since \(x, \ell \in \mathbb{Z}\), we have \(x^2(2\ell^2 - 2\ell) \in \mathbb{Z}\) and thus \(x^2(y^2-2y)\) is even, by definition. \(\square\)

§12 — Rational Numbers

Epp 4th ed. pp. 163–170 · 5th ed. (metric) pp. 183–190. Lecture 9 (week 3). Notes: rational-and-irrational-numbers.

Problems

  1. Prove the following statement:

    \[\text{For real numbers } m \text{ and } n,\ \text{if } m \text{ is irrational and } n \text{ is rational, then } m - n \text{ is irrational.}\]

  2. Prove the following statement:

    \[\text{For a real number } r,\ \text{if } 2r^2 - 3r \text{ is irrational, then } r \text{ is irrational.}\]

  3. Prove the following statement:

    \[\text{For all real numbers } r, s, n,\ \text{if } r \text{ is irrational, } s \text{ is rational and } n \text{ is a positive integer, then } n(r-s) \text{ is irrational.}\]

  4. Recall that \(\sqrt{2}\) is irrational. Consider the statement

    \[\text{For every rational number } r,\ \text{if } r \neq 0,\ \text{then } r\sqrt{2} \text{ is an irrational number.}\]

    1. Write down the negation of the statement.

    2. Use a proof by contradiction to prove the statement.

Solutions

  1. For real numbers \(m\) and \(n\), if \(m\) is irrational and \(n\) is rational, then \(m - n\) is irrational.

    Proof. Let \(m\) and \(n\) be real numbers and suppose (for a contradiction) that \(m\) is an irrational number and \(n\) is rational, but \(m - n\) is rational. Since \(n\) is rational, we have \(n = \frac{a}{b}\) for some \(a, b \in \mathbb{Z}\) where \(b \neq 0\). Since \(m - n\) is rational, we have \(m - n = \frac{c}{d}\) for some \(c, d \in \mathbb{Z}\) where \(d \neq 0\).

    Notice that \(m = (m-n) + n\). Hence \[m = (m-n) + n = \frac{c}{d} + \frac{a}{b} = \frac{bc + ad}{bd}.\] Since \(a, b, c, d \in \mathbb{Z}\) we have \(bc + ad \in \mathbb{Z}\) and \(bd \in \mathbb{Z}\). Further, since \(b \neq 0\) and \(d \neq 0\), we have \(bd \neq 0\). Thus, by definition, \(m\) is rational. But this contradicts the assumption that \(m\) is irrational. \(\square\)

  2. For a real number \(r\), if \(2r^2 - 3r\) is irrational, then \(r\) is irrational.

    Proof. We prove the contrapositive statement: for a real number \(r\), if \(r\) is rational, then \(2r^2 - 3r\) is rational.

    Let \(r\) be a real number and suppose \(r \in \mathbb{Q}\). Then \(r = \frac{a}{b}\) for some \(a, b \in \mathbb{Z}\) where \(b \neq 0\). So \[2r^2 - 3r = 2\left(\frac{a}{b}\right)^2 - 3\left(\frac{a}{b}\right) = \frac{2a^2 - 3ab}{b^2}.\] Since \(a, b \in \mathbb{Z}\), we have \(2a^2 - 3ab \in \mathbb{Z}\) and \(b^2 \in \mathbb{Z}\). Further, \(b^2 \neq 0\) because \(b \neq 0\). This means \(2r^2 - 3r\) is rational, by definition. \(\square\)

  3. For all real numbers \(r\), \(s\), and \(n\), if \(r\) is irrational, \(s\) is rational and \(n\) is a positive integer, then \(n(r-s)\) is irrational.

    Proof. Suppose, for a contradiction, that the statement is false. Then there exist real numbers \(r\), \(s\) and \(n\) such that \(r\) is an irrational number, \(s\) is rational, \(n\) is a positive integer, but \(n(r-s)\) is rational. Since \(s\) is rational, \(s = \frac{a}{b}\) for some \(a, b \in \mathbb{Z}\) where \(b \neq 0\). Since \(n(r-s)\) is rational, we have \(n(r-s) = \frac{c}{d}\) for some \(c, d \in \mathbb{Z}\) where \(d \neq 0\).

    Now, \[\begin{aligned} r - s &= n(r-s)\left(\frac{1}{n}\right) && \text{noting that } n \neq 0 \\ r - s &= \frac{c}{d} \cdot \frac{1}{n} \\ r &= s + \frac{c}{d} \cdot \frac{1}{n} \\ &= \frac{a}{b} + \frac{c}{dn} \\ &= \frac{adn + bc}{bdn} \end{aligned}\] Since \(a, b, c, d \in \mathbb{Z}\), we have \(adn + bc \in \mathbb{Z}\) and \(bdn \in \mathbb{Z}\). Further, since \(b \neq 0\), \(d \neq 0\) and \(n \neq 0\), we have \(bdn \neq 0\). This means \(r\) is rational, by definition, which is a contradiction. \(\square\)

  4. For every rational number \(r\), if \(r \neq 0\), then \(r\sqrt{2}\) is an irrational number.

    1. Negation: There exists \(r \in \mathbb{Q}\) such that \(r \neq 0\) and \(r\sqrt{2} \in \mathbb{Q}\).

    Proof. Suppose, for a contradiction, that the original statement is false. Then there exists \(r \in \mathbb{Q}\) such that \(r \neq 0\) and \(r\sqrt{2} \in \mathbb{Q}\). Since \(r\) is rational, \(r = \frac{a}{b}\) for some \(a, b \in \mathbb{Z}\) with \(b \neq 0\), and since \(r \neq 0\), we have \(a \neq 0\). Since we have assumed \(r\sqrt{2}\) is rational, \(r\sqrt{2} = \frac{c}{d}\) for some \(c, d \in \mathbb{Z}\) where \(d \neq 0\).

    Now \[\sqrt{2} = \frac{r\sqrt{2}}{r} = \left(r\sqrt{2}\right)\left(\frac{1}{r}\right) = \frac{c}{d} \cdot \frac{b}{a} = \frac{bc}{ad}.\] Since \(a, b, c, d \in \mathbb{Z}\), we know \(bc \in \mathbb{Z}\) and \(ad \in \mathbb{Z}\). Further, since \(a \neq 0\) and \(d \neq 0\), we have \(ad \neq 0\). Thus, by definition, \(\sqrt{2}\) is rational. However, this is a contradiction, as \(\sqrt{2}\) is irrational. Thus, the original statement is true. \(\square\)

§13 — Divisibility

Epp 4th ed. pp. 170–179 · 5th ed. (metric) pp. 190–199. Lecture 10 (week 4). Notes: divisibility-and-factorisation.

Problems

  1. Find the unique factorisation (in standard form) of the integer \(27\,720\).

  2. Prove or disprove the statement: for any integers \(m\), \(n\), and \(q\), if \(m \mid n\) and \(n \mid q\) then \(m^2 \mid nq\).

  3. Prove the following statements.

    1. If \(k\) is any even integer and \(m\) is any odd integer, then \((k+2)^2 - (m-3)^2\) is divisible by 4.

    2. The difference between the cube of two consecutive integers leaves the remainder 1 when it is divided by 6.

  4. Use proof by contradiction to prove this statement:

    \[\text{For all integers } c, d \text{ and } e,\ \text{if } c \mid d \text{ and } c \nmid e,\ \text{then } c \nmid (d+e).\]

Solutions

  1. By repeated division, we have \(27\,720 = 2^3 \cdot 3^2 \cdot 5 \cdot 7 \cdot 11\).

  2. For any integers \(m\), \(n\), and \(q\), if \(m \mid n\) and \(n \mid q\) then \(m^2 \mid nq\). The statement is true, and we will prove it.

    Proof. Let \(m\), \(n\), and \(q\) be non-zero integers. Assume \(m \mid n\) and \(n \mid q\). Then, by definition, there exist integers \(a, b\) such that \[n = ma \quad \text{and} \quad q = nb.\] Then observe \[nq = (ma)(nb) = ma \cdot (mab) = m^2(a^2 b).\] Since \(a^2 b \in \mathbb{Z}\), we conclude that \(m^2 \mid nq\). \(\square\)

    1. If \(k\) is any even integer and \(m\) is any odd integer, then \((k+2)^2 - (m-3)^2\) is divisible by 4.

    Proof. Let \(k\) be even and \(m\) be odd. Then \(k = 2a\) and \(m = 2b + 1\) for some \(a, b \in \mathbb{Z}\). Hence \[(k+2)^2 = (2a+2)^2 = 4(a+1)^2 \quad \text{and} \quad (m-3)^2 = (2b-2)^2 = 4(b-1)^2.\] Therefore \[(k+2)^2 - (m-3)^2 = 4\big[(a+1)^2 - (b-1)^2\big].\] Since \(a, b \in \mathbb{Z}\) we have \((a+1)^2 - (b-1)^2 \in \mathbb{Z}\) so \(4 \mid \big[(k+2)^2 - (m-3)^2\big]\). \(\square\)

    1. The difference between the cube of two consecutive integers leaves the remainder 1 when it is divided by 6.

    Proof. Let \(n\) and \(n+1\) be two consecutive integers. The difference between their cubes is \[(n+1)^3 - n^3 = 3n^2 + 3n + 1 = 3n(n+1) + 1.\] Now whether \(n\) is odd or even (so whatever integer \(n\) is), one of \(n\), \(n+1\) will be odd and the other one will be even. Thus the product \(n(n+1)\) is even.

    Since \(n(n+1)\) is even, write \(n(n+1) = 2t\) for some \(t \in \mathbb{Z}\). Then \[(n+1)^3 - n^3 = 3(2t) + 1 = 6t + 1 \equiv 1 \pmod 6.\] Hence the remainder is 1 upon division by 6. \(\square\)

  3. For all integers \(c\), \(d\) and \(e\), if \(c \mid d\) and \(c \nmid e\), then \(c \nmid (d+e)\).

    Proof. Suppose for a contradiction that the statement is false. Then there exist integers \(c\), \(d\) and \(e\) such that \(c \mid d\) and \(c \nmid e\) and \(c \mid (d+e)\).

    Now \(c \mid d\) means that \(d = cx\) for some integer \(x\). And \(c \mid (d+e)\) means that \(d + e = cy\) for some integer \(y\).

    Hence \(d + e = cx + e = cy\), so \(e = cy - cx = c(y-x)\), and we have \(y - x \in \mathbb{Z}\), because \(x\) and \(y\) are both integers. Thus \(c \mid e\), a contradiction. So the original statement is true. \(\square\)

§14 — Modular Arithmetic

Epp 4th ed. pp. 482–485 · 5th ed. (metric) pp. 528–531. Lecture 11 (week 4). Notes: modular-arithmetic.

Problems

  1. Prove the following statement:

    \[\text{If } y \text{ is any integer, then when } y^2 \text{ is divided by 5, the remainder is always 0, 1 or 4, and never 2 or 3.}\]

    1. Evaluate \(58 \operatorname{div} 11\) and \(58 \bmod 11\).

    2. Find integers \(q\) and \(r\), with \(0 \leqslant r < d\) and \(n = dq + r\), in the case that \(n = -95\) and \(d = 11\).

    1. An integer \(n\), when divided by 7, leaves remainder 4. What is the remainder (between 0 and 6 inclusive) when \(3n\) is divided by 7?

    2. When the integer \(m\) is divided by 11, the remainder 9 is left. What remainder is left when \(4m\) is divided by 22?

  2. Prove or disprove the following statements.

    1. For all real numbers \(x\), \(\lfloor x^2 \rfloor = \lfloor x \rfloor^2\).

    2. For all odd integers \(n\), \(\left\lfloor \dfrac{n^2}{4} \right\rfloor = \left(\dfrac{n-1}{2}\right)\left(\dfrac{n+1}{2}\right)\).

Solutions

  1. If \(y\) is any integer, then when \(y^2\) is divided by 5, the remainder is always 0, 1 or 4, and never 2 or 3.

    Proof. Fix \(y \in \mathbb{Z}\). By the division algorithm, \(y = 5m + r\) for some \(m \in \mathbb{Z}\) and \(r \in \{0, 1, 2, 3, 4\}\). Then \[y^2 = (5m+r)^2 = 25m^2 + 10mr + r^2 \equiv r^2 \pmod 5.\] Let us compute the possible values for \(r^2 \bmod 5\): \[0^2 \equiv 0, \quad 1^2 \equiv 1, \quad 2^2 \equiv 4, \quad 3^2 \equiv 9 \equiv 4, \quad 4^2 \equiv 16 \equiv 1 \pmod 5.\] Thus \(y^2 \equiv 0, 1\), or \(4 \pmod 5\), and never 2 or 3. \(\square\)

    1. Since \(58 = 11 \cdot 5 + 3\) with \(0 \leq 3 < 11\), we have

    \[58 \operatorname{div} 11 = 5 \quad \text{and} \quad 58 \bmod 11 = 3.\]

    1. We want \(-95 = 11q + r\) with \(0 \leq r < 11\). Taking \(q = -9\) gives

    \[11(-9) = -99 \quad \Rightarrow \quad -95 = -99 + 4 = 11(-9) + 4,\]

    so \(q = -9\) and \(r = 4\).

    1. If \(n\) leaves remainder 4 upon division by 7, then \(n = 7k + 4\) for some integer \(k\). Hence

    \[3n = 21k + 12 = 7(3k + 1) + 5,\]

    so the remainder when \(3n\) is divided by 7 is 5. (The source PDF prints this step as \(7(3k) + 5\), which equals \(21k + 5\), not \(21k + 12\) — the quotient is \(3k + 1\), as above. The answer of 5 is unaffected.)

    1. If \(m\) leaves remainder 9 upon division by 11, then \(m = 11k + 9\) for some integer \(k\). Then

    \[4m = 44k + 36 = 22(2k+1) + 14,\]

    so the remainder when \(4m\) is divided by 22 is 14.

    1. For all real numbers \(x\), \(\lfloor x^2 \rfloor = \lfloor x \rfloor^2\). The statement is false. For a counterexample, take \(x = \frac{3}{2}\). Then

    \[\lfloor x^2 \rfloor = \left\lfloor \frac{9}{4} \right\rfloor = 2 \quad \text{but} \quad \lfloor x \rfloor^2 = \left\lfloor \frac{3}{2} \right\rfloor^2 = 1^2 = 1.\]

    1. For all odd integers \(n\), \(\left\lfloor \dfrac{n^2}{4} \right\rfloor = \left(\dfrac{n-1}{2}\right)\left(\dfrac{n+1}{2}\right)\). The statement is true.

    Proof. Let \(n\) be an odd integer, so \(n = 2k+1\) for some integer \(k\). Then \[\frac{n^2}{4} = \frac{(2k+1)^2}{4} = \frac{4k^2 + 4k + 1}{4} = k(k+1) + \frac{1}{4},\] so \(\left\lfloor \frac{n^2}{4} \right\rfloor = k(k+1)\). Moreover, notice \[\left(\frac{n-1}{2}\right)\left(\frac{n+1}{2}\right) = \left(\frac{2k}{2}\right)\left(\frac{2k+2}{2}\right) = k(k+1).\] Hence \(\left\lfloor \frac{n^2}{4} \right\rfloor = \left(\frac{n-1}{2}\right)\left(\frac{n+1}{2}\right)\) for all odd integers \(n\). \(\square\)

§15 — The Euclidean Algorithm

Epp 4th ed. pp. 220–224 · 5th ed. (metric) pp. 250–254. Lecture 12 (week 4). Notes: gcd-lcm-and-euclidean-algorithm.

Problems

  1. Use the Euclidean Algorithm to determine \(\gcd(14, 3003)\).

  2. Compute

    1. \(\operatorname{lcm}(12, 18)\)

    2. \(\operatorname{lcm}(2^2 \cdot 3 \cdot 5,\ 2^3 \cdot 3^2)\)

    3. \(\operatorname{lcm}(2800, 6125)\)

  3. Prove that for all positive integers \(a\) and \(b\), we have

    \[\gcd(a,b) = \operatorname{lcm}(a,b) \iff a = b.\]

  4. Prove that for all positive integers \(a\) and \(b\), we have

    \[a \mid b \iff \operatorname{lcm}(a,b) = b.\]

    1. Use the Euclidean algorithm to determine \(\gcd(931, 301)\). Show your working.

    2. Write down the unique prime factorisation of 931 in standard form.

    3. How many positive divisors does 931 have? (Hint: use part (b).)

    1. Using the Euclidean algorithm, compute \(\gcd(116, 88)\). Show your working.

    2. Using your answer to part (a), compute \(\operatorname{lcm}(116, 88)\).

Solutions

  1. We apply the Euclidean Algorithm:

    \[3003 = 14 \cdot 214 + 7, \qquad 14 = 7 \cdot 2 + 0.\]

    Hence \(\gcd(14, 3003) = 7\).

    1. We factor \(12 = 2^2 \cdot 3\) and \(18 = 2 \cdot 3^2\) so \(\operatorname{lcm}(12, 18) = 2^2 \cdot 3^2 = 36\).

    2. Using the maximum exponents of each prime, we have

    \[\operatorname{lcm}(2^2 \cdot 3 \cdot 5,\ 2^3 \cdot 3^2) = 2^3 \cdot 3^2 \cdot 5 = 360.\]

    1. We factor \(2800 = 2^4 \cdot 5^2 \cdot 7\) and \(6125 = 5^3 \cdot 7^2\). Thus

    \[\operatorname{lcm}(2800, 6125) = 2^4 \cdot 5^3 \cdot 7^2 = 98000.\]

  2. For all positive integers \(a\) and \(b\), \(\gcd(a,b) = \operatorname{lcm}(a,b)\) if and only if \(a = b\).

    Proof. Let \(a\) and \(b\) be positive integers.

    (\(\Longleftarrow\)) It is clear that if \(a = b\) then \(\gcd(a,b) = \operatorname{lcm}(a,b)\).

    (\(\Longrightarrow\)) Assume \(\gcd(a,b) = \operatorname{lcm}(a,b)\). Denoting \(g = \gcd(a,b)\), we have \(a = gm\) and \(b = gn\) for some positive integers \(m\) and \(n\) with \(\gcd(m,n) = 1\).

    Now notice \(gm\) and \(gn\) have no common prime factors beyond those in \(g\), so the least common multiple uses all primes from both \(m\) and \(n\). Therefore \[\operatorname{lcm}(a,b) = gmn.\] Now, we have \(g = gmn\). Therefore \(mn = 1\), and so \(m = n = 1\). This lets us conclude \(a = b\). \(\square\)

  3. For all positive integers \(a\) and \(b\), \(a \mid b\) if and only if \(\operatorname{lcm}(a,b) = b\).

    Proof. Let us denote \(g = \gcd(a,b)\) so that \(a = gm\) and \(b = gn\) for some positive integers \(m\) and \(n\) with \(\gcd(m,n) = 1\). As explained in the previous question, we have \[\operatorname{lcm}(a,b) = gmn.\]

    (\(\Longrightarrow\)) If \(a \mid b\), then \(gm \mid gn\), hence \(m \mid n\). Since \(\gcd(m,n) = 1\), we conclude \(m = 1\). Therefore \(\operatorname{lcm}(a,b) = gmn = gn = b\).

    (\(\Longleftarrow\)) If \(\operatorname{lcm}(a,b) = b\), then \(gmn = gn\) which means \(m = 1\). Therefore \(a = gm = g\) and so \(a \mid b\). \(\square\)

    1. Using the Euclidean Algorithm:

    \[\begin{aligned} 931 &= 301 \cdot 3 + 28 \\ 301 &= 28 \cdot 10 + 21 \\ 28 &= 21 \cdot 1 + 7 \\ 21 &= 7 \cdot 3 + 0. \end{aligned}\]

    Hence \(\gcd(931, 301) = 7\).

    1. Since \(\gcd(931, 301) = 7\), we know \(7 \mid 931\), and we have \(931 = 7 \cdot 133\). The number 133 has prime factorisation \(7 \cdot 19\). Therefore the unique prime factorisation of 931 is \(7^2 \cdot 19\).

    2. We have \(931 = 7^2 \cdot 19^1\) from part (b). Any divisor of 931 must have a prime factorisation of the form \(7^a \cdot 19^b\) where \(a \in \{0,1,2\}\) and \(b \in \{0,1\}\). Thus the number of positive divisors is

    \[(2+1)(1+1) = 6.\]

    (There are 3 choices for the exponent \(a\) and 2 choices for the exponent \(b\).)

    1. Using the Euclidean Algorithm:

    \[\begin{aligned} 116 &= 88 \cdot 1 + 28 \\ 88 &= 28 \cdot 3 + 4 \\ 28 &= 4 \cdot 7 + 0 \end{aligned}\]

    Hence \(\gcd(116, 88) = 4\).

    1. Using \(\gcd(a,b) \cdot \operatorname{lcm}(a,b) = ab\) for \(a, b \in \mathbb{Z}^+\),

    \[\operatorname{lcm}(116, 88) = \frac{116 \cdot 88}{\gcd(116,88)} = \frac{116 \cdot 88}{4} = 116 \cdot 22 = 2552.\]

Application — Trace Tables for Algorithms

Not assessable for MATH1061. For reference, see S. Epp, Discrete Mathematics with Applications, Section 4.8 (4th edition) or Section 4.10 (5th edition), Brooks/Cole, Belmont, CA.

The following pseudo-code describes the Division Algorithm, with annotations to explain what each step is doing.

Input: \(a\) (a non-negative integer), \(d\) (a positive integer). This algorithm will take two inputs, \(a\) and \(d\), and will find the quotient \(a \operatorname{div} d\) and the remainder \(a \bmod d\).

Algorithm body:

r := a          -- defining our variables
q := 0
                -- starting with a, repeatedly subtract lots of d until we reach
                -- a number less than d; each round adds 1 to the quotient count
while r >= d,
    r := r - d
    q := q + 1
end while

Output: \(q\), \(r\). Once we have obtained a number less than \(d\), we are finished. The output \(q\) is the quotient (\(a \operatorname{div} d\)) and the output \(r\) is the remainder (\(a \bmod d\)).

The following pseudo-code describes the Euclidean Algorithm, with annotations to explain what each step is doing.

Input: \(A\), \(B\) (integers with \(A > B \geq 0\)).

Algorithm body:

a := A          -- defining our variables
b := B
r := B
                -- if b /= 0, compute a mod b (call the algorithm above for this
                -- step) and set r to be this value; replace a with b and b with r
                -- and repeat until b = 0
while b /= 0,
    r := a mod b
    a := b
    b := r
end while
gcd := a

Output: \(\text{gcd}\) (a positive integer).

A trace table keeps track of the value of each variable at each step of an algorithm. For example, the trace table for the Division Algorithm with input variables \(a = 19\) and \(d = 4\) is:

Variable / iteration 0 1 2 3 4
\(a\) 19
\(d\) 4
\(r\) 19 15 11 7 3
\(q\) 0 1 2 3 4

Output: \(q = 4\), \(r = 3\).

Application problems
  1. Construct a trace table to trace the action of the Division Algorithm for the following inputs.

    1. \(a = 26\), \(d = 7\)

    2. \(a = 15\), \(d = 6\)

  2. Construct a trace table to trace the action of the Euclidean Algorithm with inputs \(A = 1001\) and \(B = 871\).

Application solutions
    1. Trace table for the Division Algorithm with \(a = 26\), \(d = 7\):
    Variable / iteration 0 1 2 3
    \(a\) 26
    \(d\) 7
    \(r\) 26 19 12 5
    \(q\) 0 1 2 3

    Output: \(q = 3\), \(r = 5\).

    1. Trace table for the Division Algorithm with \(a = 15\), \(d = 6\):
    Variable / iteration 0 1 2
    \(a\) 15
    \(d\) 6
    \(r\) 15 9 3
    \(q\) 0 1 2

    Output: \(q = 2\), \(r = 3\).

  1. Trace table for the Euclidean Algorithm with inputs \(A = 1001\) and \(B = 871\):

    Variable / iteration 0 1 2 3 4 5
    \(A\) 1001
    \(B\) 871
    \(a\) 1001 871 130 91 39 13
    \(b\) 871 130 91 39 13 0
    \(r\) 871 130 91 39 13 0
    \(\text{gcd}\) 13

    Output: \(\text{gcd} = 13\).

§16 — Sequences

Epp 4th ed. pp. 227–244 · 5th ed. (metric) pp. 258–275. Lecture 13 (week 5).

Problems

  1. Write down the first four terms in the sequence defined by \(a_k = \dfrac{k}{10+k}\) for \(k = 1, 2, \dots\).

  2. Consider the sequences \(a_k = 2k+1\) and \(b_k = (k-1)^3 + k + 2\) where \(k = 0, 1, 2, \dots\). Show that the first three terms of these sequences are identical, but that their fourth terms differ.

    1. Find an explicit formula for a sequence \(c_n\) whose first six terms are \(-1, 1, -1, 1, -1, 1\).

    2. Find an explicit formula for a sequence \(d_n\) whose first six terms are \(\frac{1}{4}, \frac{2}{9}, \frac{3}{16}, \frac{4}{25}, \frac{5}{36}, \frac{6}{49}\).

    1. Write \((x^2 - 1) + (2x^3 - 1) + (3x^4 - 1) + (4x^5 - 1) + (5x^6 - 1)\) in summation notation.

    2. Evaluate \(\displaystyle\prod_{i=1}^{5} (-1)^i \cdot i\).

    3. Write \((y-1)(y^2 - 2)(y^3 - 3)(y^4 - 4)\) in product notation.

    4. Evaluate \(\displaystyle\sum_{j=1}^{n} (-1)^j \frac{1}{2^j}\) when \(n = 5\).

Solutions

  1. We have

    \[a_1 = \frac{1}{11}, \qquad a_2 = \frac{2}{12} = \frac{1}{6}, \qquad a_3 = \frac{3}{13}, \qquad a_4 = \frac{4}{14} = \frac{2}{7}.\]

  2. We compute

    \[a_0 = 1,\quad a_1 = 3,\quad a_2 = 5,\quad a_3 = 7;\]

    \[b_0 = (-1)^3 + 0 + 2 = 1, \quad b_1 = 0^3 + 1 + 2 = 3, \quad b_2 = 1^3 + 2 + 2 = 5, \quad b_3 = 2^3 + 3 + 2 = 13.\]

    Hence the first three terms agree (\(a_0 = b_0\), \(a_1 = b_1\) and \(a_2 = b_2\)), but the fourth terms differ (\(a_3 = 7\) while \(b_3 = 13\)).

    1. One explicit formula is

    \[c_n = (-1)^n \qquad n = 1, 2, 3, \dots\]

    1. The denominators are square numbers, so an explicit formula is

    \[d_n = \frac{n}{(n+1)^2} \qquad n = 1, 2, 3, \dots\]

    1. We have

    \[(x^2 - 1) + (2x^3 - 1) + (3x^4 - 1) + (4x^5 - 1) + (5x^6 - 1) = \sum_{k=1}^{5} \left(kx^{k+1} - 1\right).\]

    1. We have

    \[\prod_{i=1}^{5} (-1)^i \cdot i = \left(\prod_{i=1}^{5} (-1)^i\right)\left(\prod_{i=1}^{5} i\right) = (-1)^{1+2+3+4+5} \cdot 5! = (-1)^{15} \cdot 120 = -120.\]

    1. We have

    \[(y-1)(y^2-2)(y^3-3)(y^4-4) = \prod_{k=1}^{4} \left(y^k - k\right).\]

    1. When \(n = 5\), we have

    \[\sum_{j=1}^{5} (-1)^j \frac{1}{2^j} = -\frac{1}{2} + \frac{1}{4} - \frac{1}{8} + \frac{1}{16} - \frac{1}{32} = -\frac{11}{32}.\]

§17 — Mathematical Induction

Epp 4th ed. pp. 244–268 · 5th ed. (metric) pp. 275–301. Lecture 13 (week 5).

Problems

  1. Use mathematical induction to prove the following statements.

    1. For every positive integer \(n\), \(n(n+1)\) is even.

    2. For every integer \(n \geq 2\), \(n^3 - n\) is a multiple of 6.

    3. For all integers \(n \geq 7\), \(3^n < n!\).

    4. For each integer \(n \geq 1\),

    \[\sum_{i=1}^{n} i^3 = \frac{n^2(n+1)^2}{4}.\]

    1. For each integer \(n \geq 0\), \(1 + 4n \leq 5^n\).

Solutions

    1. For every positive integer \(n\), \(n(n+1)\) is even.

    Proof. Let \(P(n)\) be the predicate “\(n(n+1)\) is even”.

    Basis: when \(n = 1\), \(n(n+1) = 2\) which is even, so \(P(1)\) is true.

    (The inductive step is to show that for all integers \(k \geq 1\), \(P(k)\) implies \(P(k+1)\).)

    Inductive hypothesis: suppose that for some integer \(k \geq 1\), \(P(k)\) is true. Thus \(k(k+1) = 2a\) for some integer \(a\).

    Now consider \(n = k+1\). Then \[(k+1)(k+2) = k(k+1) + 2(k+1) = 2a + 2(k+1) = 2(a + k + 1).\] Since \(a\) and \(k\) are integers, \(a + k + 1\) is an integer, so \((k+1)(k+2)\) is even and \(P(k+1)\) is true.

    Thus, by the principle of mathematical induction, \(n(n+1)\) is even for every positive integer \(n\). \(\square\)

    1. For every integer \(n \geq 2\), \(n^3 - n\) is a multiple of 6.

    Proof. Let \(P(n)\) be the predicate \(6 \mid (n^3 - n)\).

    Basis: when \(n = 2\), \(n^3 - n = 8 - 2 = 6\). Since \(6 \mid 6\), \(P(2)\) is true.

    (The inductive step is to show that for all integers \(k \geq 2\), \(P(k)\) implies \(P(k+1)\).)

    Inductive hypothesis: suppose that for some integer \(k \geq 2\), \(P(k)\) is true. Thus \(k^3 - k = 6a\) for some integer \(a\).

    Now consider \(n = k+1\). Then \[(k+1)^3 - (k+1) = k^3 + 3k^2 + 3k + 1 - k - 1 = k^3 - k + 3k^2 + 3k = 6a + 3k(k+1).\] For the integer \(k\), \(k(k+1)\) is even from part (a), so let \(k(k+1) = 2b\) for some integer \(b\). Hence \((k+1)^3 - (k+1) = 6a + 3(2b) = 6(a+b)\). Thus \(6 \mid \big((k+1)^3 - (k+1)\big)\) and so \(P(k+1)\) is true.

    Thus, by the principle of mathematical induction, \(n^3 - n\) is a multiple of 6 for each integer \(n \geq 2\). \(\square\)

    1. For all integers \(n \geq 7\), \(3^n < n!\).

    Proof. Let \(P(n)\) be the predicate \(3^n < n!\).

    Basis: \(P(7)\) is the statement \(3^7 < 7!\). The left-hand side is \(3^7 = 2187\) and the right-hand side is \(7! = 5040\), so \(P(7)\) holds.

    (The inductive step is to show that for all integers \(k \geq 7\), \(P(k)\) implies \(P(k+1)\).)

    Inductive hypothesis: let \(k \geq 7\) be an integer and assume that \(P(k)\) is true; that is \(3^k < k!\). We will now show that \(P(k+1)\) holds, so we will show \(3^{k+1} < (k+1)!\).

    Taking the right-hand side, we have \[\begin{aligned} (k+1)! &= (k+1) \cdot k! \\ &> (k+1) \cdot 3^k && \text{by the inductive hypothesis} \\ &> 3 \cdot 3^k && \text{since } k \geq 7 \\ &= 3^{k+1} \end{aligned}\] Hence \(P(k+1)\) is true.

    Thus by the principle of mathematical induction, for all integers \(n \geq 7\), \(3^n < n!\). \(\square\)

    1. For each integer \(n \geq 1\), \(\displaystyle\sum_{i=1}^{n} i^3 = \frac{n^2(n+1)^2}{4}\).

    Proof. Let \(P(n)\) be the predicate \[\sum_{i=1}^{n} i^3 = \frac{n^2(n+1)^2}{4}.\]

    Basis: consider \(n = 1\). We have \(\sum_{i=1}^{1} i^3 = 1^3 = 1\) and \(\frac{1^2(1+1)^2}{4} = \frac{4}{4} = 1\) so \(P(1)\) is true.

    (The inductive step is to show that for all integers \(k \geq 1\), \(P(k)\) implies \(P(k+1)\).)

    Inductive hypothesis: suppose \(k \geq 1\) is an integer and \(P(k)\) is true. That is, \[\sum_{i=1}^{k} i^3 = \frac{k^2(k+1)^2}{4}.\]

    Now consider \(n = k+1\). \[\begin{aligned} \sum_{i=1}^{k+1} i^3 &= \sum_{i=1}^{k} i^3 + (k+1)^3 && \text{(splitting off the last term)} \\ &= \frac{k^2(k+1)^2}{4} + (k+1)^3 && \text{by the inductive hypothesis} \\ &= \frac{k^2(k+1)^2}{4} + \frac{4(k+1)^3}{4} \\ &= \frac{(k+1)^2}{4}\big(k^2 + 4(k+1)\big) \\ &= \frac{(k+1)^2}{4}\big(k^2 + 4k + 4\big) \\ &= \frac{(k+1)^2}{4}(k+2)^2 \end{aligned}\] Thus \(P(k+1)\) is true.

    Therefore, by the principle of mathematical induction, \(\sum_{i=1}^{n} i^3 = \frac{n^2(n+1)^2}{4}\) for every integer \(n \geq 1\). \(\square\)

    1. For each integer \(n \geq 0\), \(1 + 4n \leq 5^n\).

    Proof. Let \(P(n)\) be the predicate \(1 + 4n \leq 5^n\).

    Basis: when \(n = 0\) we have \(1 + 4 \cdot 0 \leq 5^0\), so \(P(0)\) is true.

    (The inductive step is to show that for all integers \(k \geq 0\), \(P(k)\) implies \(P(k+1)\).)

    Inductive hypothesis: suppose \(k \geq 0\) is an integer and \(P(k)\) is true; that is, \(1 + 4k \leq 5^k\).

    Now consider \(n = k+1\). Starting with the left-hand side of \(P(k+1)\), we have \[\begin{aligned} 1 + 4(k+1) &= 1 + 4k + 4 \\ &\leq 5^k + 4 && \text{by the inductive hypothesis} \\ &\leq 5^k + 4 \cdot 5^k && \text{since } k \geq 0 \text{ so } 5^k \geq 1 \\ &= 5(5^k) \\ &= 5^{k+1} \end{aligned}\] so \(P(k+1)\) is true.

    Thus, by the principle of mathematical induction, \(1 + 4n \leq 5^n\) for every integer \(n \geq 0\). \(\square\)

§18 — Strong Induction and the Well Ordering Principle

Epp 4th ed. pp. 268–279 · 5th ed. (metric) pp. 301–314. Lecture 14 (week 5).

Problems

  1. Consider the statement

    \[\text{For every integer } n \geq 2, \text{ there exist integers } a, b \geq 0 \text{ such that } n = 2a + 3b.\]

    1. Prove the statement using the well-ordering principle.

    2. Prove the statement using strong mathematical induction.

  2. The Lucas sequence \(a_1, a_2, a_3, \dots\) is defined by \(a_1 = 1\), \(a_2 = 3\) and \(a_k = a_{k-1} + a_{k-2}\) for all integers \(k \geqslant 3\). Use strong mathematical induction to prove that \(a_n \leqslant \left(\dfrac{7}{4}\right)^n\) for all \(n \in \mathbb{Z}^+\).

  3. Suppose \(x\) is a real number and \(x + \frac{1}{x}\) is an integer. Use strong mathematical induction to prove \(a_n := x^n + \frac{1}{x^n}\) is an integer for all \(n \geq 0\).

Solutions

  1. For every integer \(n \geq 2\), there exist integers \(a, b \geq 0\) such that \(n = 2a + 3b\).

    Proof (using the well-ordering principle). Consider the set of counterexamples \[S = \{\, n \in \mathbb{Z}^{\geq 2} : \text{there do not exist integers } a, b \geq 0 \text{ with } n = 2a + 3b \,\}.\] If \(S\) is empty, we are done. Otherwise, by the well-ordering principle, \(S\) has a least element. Let us call it \(N\).

    Notice that \(2 \notin S\) because \(2 = 2 \cdot 1 + 3 \cdot 0\), and \(3 \notin S\) because \(3 = 2 \cdot 0 + 3 \cdot 1\). Also \(4 \notin S\) because \(4 = 2 \cdot 2 + 3 \cdot 0\). Hence \(N \geq 5\). By minimality of \(N\), we must have \(N - 2 \notin S\). Therefore there exist integers \(a, b \geq 0\) such that \[N - 2 = 2a + 3b.\] In other words, we have \[N = 2(a+1) + 3b.\] This means \(N \notin S\), contradicting the assumption that \(N \in S\). Hence \(S\) is empty, and the statement holds for all \(n \geq 2\). \(\square\)

    Proof (using strong induction). Let \(P(n)\) be the predicate: “there exist integers \(a, b \geq 0\) such that \(n = 2a + 3b\).”

    Basis: we have \[2 = 2 \cdot 1 + 3 \cdot 0, \qquad 3 = 2 \cdot 0 + 3 \cdot 1, \qquad 4 = 2 \cdot 2 + 3 \cdot 0,\] so \(P(2)\), \(P(3)\), \(P(4)\) are true.

    (The inductive step is to show that for every integer \(k \geq 4\), if \(P(2), P(3), \dots, P(k)\) are all true then \(P(k+1)\) is true.)

    Inductive hypothesis: suppose \(k \geq 4\) is an integer and assume \(P(m)\) is true for every integer \(m = 2, 3, \dots, k\).

    Consider \(n = k+1\). Since \(k \geq 4\), we have \(k - 1 \geq 3\), so by the inductive hypothesis \(P(k-1)\) is true. Thus there exist \(a, b \geq 0\) such that \[k - 1 = 2a + 3b.\] Adding 2 gives \[k + 1 = 2(a+1) + 3b,\] so \(P(k+1)\) holds. \(\square\)

  2. The Lucas sequence \(a_1, a_2, a_3, \dots\) is defined by \(a_1 = 1\), \(a_2 = 3\) and \(a_k = a_{k-1} + a_{k-2}\) for all integers \(k \geqslant 3\). Prove that \(a_n \leqslant \left(\frac{7}{4}\right)^n\) for all \(n \in \mathbb{Z}^+\).

    Proof. Let \(P(n)\) be the predicate \(a_n \leqslant \left(\frac{7}{4}\right)^n\).

    Basis: when \(n = 1\), \(a_1 = 1 \leqslant \left(\frac{7}{4}\right)^1 = 1\frac{3}{4}\) is true, so \(P(1)\) is true. When \(n = 2\), \(P(2)\) holds because \(a_2 = 3 \leqslant \left(\frac{7}{4}\right)^2 = \frac{49}{16} = 3\frac{1}{16}\).

    (The inductive step is to show that for every integer \(k \geq 1\), if \(P(1), \dots, P(k)\) are all true then \(P(k+1)\) is true.)

    Inductive hypothesis: suppose \(k \geq 1\) is an integer and that \(P(m)\) is true for all \(m\) with \(1 \leq m \leq k\).

    Now consider \(P(k+1)\), which we want to prove holds. We have \[\begin{aligned} a_{k+1} &= a_k + a_{k-1} \\ &\leqslant \left(\frac{7}{4}\right)^k + \left(\frac{7}{4}\right)^{k-1} && \text{by the inductive hypothesis} \\ &= \left(\frac{7}{4}\right)^{k-1}\left(\frac{7}{4} + 1\right) \\ &= \left(\frac{7}{4}\right)^{k-1}\left(\frac{11}{4}\right) \\ &\leqslant \left(\frac{7}{4}\right)^{k-1}\left(\frac{49}{16}\right) && \text{noting } \tfrac{11}{4} = \tfrac{44}{16} < \tfrac{49}{16} \\ &= \left(\frac{7}{4}\right)^{k-1}\left(\frac{7}{4}\right)^2 \\ &= \left(\frac{7}{4}\right)^{k+1}. \end{aligned}\] Hence \(P(k+1)\) holds. Therefore, by the principle of strong mathematical induction, \(a_n \leqslant \left(\frac{7}{4}\right)^n\) for all integers \(n \geqslant 1\). \(\square\)

  3. Suppose \(x\) is a real number and \(x + \frac{1}{x}\) is an integer. Prove \(a_n := x^n + \frac{1}{x^n}\) is an integer for all \(n \geq 0\).

    Proof. Suppose \(x\) is a real number and \(x + \frac{1}{x}\) is an integer. Let \(P(n)\) be the predicate “\(a_n\) is an integer.”

    Basis: \(P(0)\) holds since \(a_0 = x^0 + \frac{1}{x^0} = 2\). \(P(1)\) holds since \(a_1 = x + \frac{1}{x}\), which is an integer by assumption.

    (The inductive step is to show that for every integer \(k \geq 1\), if \(P(0), P(1), \dots, P(k)\) are all true then \(P(k+1)\) is true.)

    Inductive hypothesis: suppose \(k \geq 1\) is an integer and \(P(m)\) holds for all integers \(m = 0, 1, \dots, k\).

    We now aim to show that \(P(k+1)\) is true. Notice that \[\begin{aligned} a_1 a_k &= \left(x + \frac{1}{x}\right)\left(x^k + \frac{1}{x^k}\right) \\ &= \left(x^{k+1} + \frac{1}{x^{k+1}}\right) + \left(x^{k-1} + \frac{1}{x^{k-1}}\right) = a_{k+1} + a_{k-1}. \end{aligned}\] Therefore \(a_{k+1} = a_1 a_k - a_{k-1}\). By the inductive hypothesis, \(a_k\) and \(a_{k-1}\) are integers, and \(a_1\) is an integer, so \(a_{k+1}\) is an integer. Hence \(P(k+1)\) holds.

    Therefore, by the principle of strong mathematical induction, \(a_n\) is an integer for all \(n \geq 0\). \(\square\)

§19 — Recursive Definitions

Epp 4th ed. pp. 290–304 · 5th ed. (metric) pp. 325–340. Lecture 15 (week 6).

Problems

  1. Define the sequence \(b_1 = 2\), \(b_2 = 6\) and

    \[b_k = b_{k-1} + b_{k-2} + \gcd(b_{k-1}, b_{k-2})\]

    for all \(k \geq 3\). Use strong induction to prove \(b_n\) is even for all \(n \geq 1\).

  2. A sequence is defined by \(a_1 = 2\) and

    \[a_k = k(a_{k-1} + k - 1)\]

    for all integers \(k \geq 2\). Use mathematical induction to prove \(a_n\) is even for all \(n \geq 1\).

  3. Recursively define the set \(S\) of binary strings as follows.

    1. (Base rule) \(1 \in S\).

    2. (Recursion rule) If \(s \in S\) then \(0s \in S\) and \(1s \in S\) (e.g. if \(s = 1\) then \(0s = 01\)).

    3. (Restriction rule) No other strings lie in \(S\) besides those given by the rules above.

    What are the strings contained in \(S\) arising from the first three applications of the recursive rule?

Solutions

  1. Proof. Let \(P(n)\) be the predicate “\(b_n\) is even”.

    Base case. The base cases are \(n = 1, 2\). We have \(b_1 = 2\) and \(b_2 = 6\) so \(P(1)\) and \(P(2)\) are true.

    Inductive hypothesis. Assume \(P(m)\) is true for all integers \(m\) with \(1 \leq m \leq k\), where \(k \geq 2\) is an integer. That is, we assume \(b_1, b_2, \dots, b_k\) are all even.

    Inductive step. We need to show that \(P(k+1)\) is true, i.e. \(b_{k+1}\) is even. Using the recursive definition, we have \[b_{k+1} = b_k + b_{k-1} + \gcd(b_k, b_{k-1}).\] By the inductive hypothesis, \(b_k\) and \(b_{k-1}\) are even, so \(2 \mid b_k\) and \(2 \mid b_{k-1}\), hence \(2 \mid \gcd(b_k, b_{k-1})\) as well. Therefore each term on the right-hand side is even, so their sum \(b_{k+1}\) is even. Thus \(P(k+1)\) holds.

    Therefore, \(b_n\) is even for all integers \(n \geq 1\). \(\square\)

  2. Proof. Let \(P(n)\) be the predicate “\(a_n\) is even”.

    Base case. The base case is \(n = 1\). We have \(a_1 = 2\), which is even, so \(P(1)\) is true.

    Inductive hypothesis. Assume \(P(k)\) is true for some integer \(k \geq 1\). That is, assume \(a_k\) is even.

    Inductive step. We need to show that \(P(k+1)\) is true, i.e. \(a_{k+1}\) is even. By the recursive definition, we have \[a_{k+1} = (k+1)\big(a_k + (k+1) - 1\big) = (k+1)(a_k + k).\] Note that \(a_k\) is even by the inductive hypothesis. We split into cases:

    • If \(k\) is even, then \(a_k + k\) is even, hence \(a_{k+1}\) is even.
    • If \(k\) is odd, then \(k + 1\) is even, hence \(a_{k+1}\) is even.

    In all cases, \(a_{k+1}\) is even, so \(P(k+1)\) is true.

    Therefore, \(a_n\) is even for all integers \(n \geq 1\). \(\square\)

  3. By the base rule, we only know \(1 \in S\).

    First application of the recursive rule:

    \[01 \in S, \qquad 11 \in S.\]

    Second application of the recursive rule:

    \[001 \in S, \quad 101 \in S, \quad 011 \in S, \quad 111 \in S.\]

    Third application of the recursive rule:

    \[0001 \in S, \quad 1001 \in S, \quad 0101 \in S, \quad 1101 \in S,\] \[0011 \in S, \quad 1011 \in S, \quad 0111 \in S, \quad 1111 \in S.\]

§20 — Solving Recurrence Relations

Epp 4th ed. pp. 304–328 · 5th ed. (metric) pp. 340–364. Lecture 16 (week 6).

Problems

  1. Define a sequence \(a_n = 2^n + 5 \cdot 3^n\) for all integers \(n \geq 0\). Show that

    \[a_n = 5a_{n-1} - 6a_{n-2}\]

    for all integers \(n \geq 2\).

  2. Let \(\{b_n\}_{n \geq 0}\) be the sequence defined by

    \[b_0 = 2, \qquad b_1 = 5, \qquad \text{and} \qquad b_k = 2b_{k-1} - b_{k-2} \text{ for each integer } k \geq 2.\]

    1. Write down the values of \(b_0, b_1, b_2, b_3, b_4, b_5\).

    2. Based on your answers for part (a), guess an explicit formula for this sequence.

    3. Prove that your guess from part (b) is correct.

  3. Let \(T_n\) be a sequence defined by \(T_1 = 1\) and

    \[T_{n+1} = 1 + (T_1 \times T_2 \times T_3 \times \cdots \times T_n)\]

    for all \(n \geq 1\). Prove that \(T_{n+1} = (T_n)^2 - T_n + 1\) for all \(n \geq 2\).

  4. Consider the Fibonacci sequence defined recursively by

    \[\begin{aligned} a_0 &= 0 \\ a_1 &= 1 \\ a_n &= a_{n-2} + a_{n-1} \qquad \text{for all integers } n \geq 2. \end{aligned}\]

    Use strong mathematical induction to prove the explicit formula

    \[a_n = \frac{\phi^n - \psi^n}{\sqrt{5}}\]

    for all integers \(n \geq 0\), where \(\phi = \frac{1+\sqrt{5}}{2}\) and \(\psi = \frac{1-\sqrt{5}}{2}\). (You may use the fact that \(\phi^2 = \phi + 1\) and \(\psi^2 = \psi + 1\).)

Solutions

  1. Define a sequence \(a_n = 2^n + 5 \cdot 3^n\) for all integers \(n \geq 0\). Show that \(a_n = 5a_{n-1} - 6a_{n-2}\) for all integers \(n \geq 2\).

    Proof. For \(n \geq 2\), we have \(n - 2 \geq 0\) so all three terms \(a_n\), \(a_{n-1}\) and \(a_{n-2}\) are defined in the given formula. Then, \[\begin{aligned} 5a_{n-1} - 6a_{n-2} &= 5\left(2^{n-1} + 5 \cdot 3^{n-1}\right) - 6\left(2^{n-2} + 5 \cdot 3^{n-2}\right) \\ &= \left(5 \cdot 2^{n-1} - 6 \cdot 2^{n-2}\right) + \left(25 \cdot 3^{n-1} - 30 \cdot 3^{n-2}\right) \\ &= (10 - 6)2^{n-2} + (75 - 30)3^{n-2} \\ &= 4 \cdot 2^{n-2} + 45 \cdot 3^{n-2} \\ &= 2^n + 5 \cdot 3^n \\ &= a_n. \end{aligned}\] Hence \(a_n = 5a_{n-1} - 6a_{n-2}\) for all integers \(n \geq 2\). \(\square\)

    \[\begin{aligned} b_0 &= 2 \\ b_1 &= 5 \\ b_2 &= 2(5) - 2 = 8 \\ b_3 &= 2(8) - 5 = 11 \\ b_4 &= 2(11) - 8 = 14 \\ b_5 &= 2(14) - 11 = 17 \end{aligned}\]

    1. Guess that \(b_n = 3n + 2\) for each integer \(n \geq 0\).

    Proof. The guess gives \(b_0 = 3(0) + 2 = 2\) and \(b_1 = 3(1) + 2 = 5\) so it gives the correct initial conditions. Now we substitute the guess into the recurrence relation. The LHS of the recurrence relation is \(b_k = 3k + 2\). The RHS of the recurrence relation is \[\begin{aligned} 2b_{k-1} - b_{k-2} &= 2\big(3(k-1) + 2\big) - \big(3(k-2) + 2\big) \\ &= 2(3k - 1) - (3k - 4) \\ &= 6k - 2 - 3k + 4 \\ &= 3k + 2 \end{aligned}\] So the guess also satisfies the recurrence relation. Thus the guess is correct. \(\square\)

    Alternative proof (by strong induction). Let \(P(n)\) be the predicate \(b_n = 3n + 2\).

    Basis: \(b_0 = 2\) from the recursive definition and \(3(0) + 2 = 2\) so \(P(0)\) is true. \(b_1 = 5\) from the recursive definition and \(3(1) + 2 = 5\) so \(P(1)\) is true.

    Inductive hypothesis: suppose \(k \geq 1\) is an integer and \(P(0), P(1), \dots, P(k)\) are true.

    Now \[\begin{aligned} b_{k+1} &= 2b_k - b_{k-1} && \text{by the rec. rel. since } k+1 \geq 2 \\ &= 2(3k+2) - \big(3(k-1) + 2\big) && \text{since } P(k) \text{ and } P(k-1) \text{ are true} \\ &= 6k + 4 - (3k - 1) \\ &= 3k + 5 \\ &= 3(k+1) + 2 \end{aligned}\] So \(P(k+1)\) is true. Therefore, by the principle of (strong) mathematical induction, the guess is correct for this recursively defined sequence. \(\square\)

  2. Let \(T_n\) be a sequence defined by \(T_1 = 1\) and \(T_{n+1} = 1 + (T_1 \times \cdots \times T_n)\) for all \(n \geq 1\). Prove that \(T_{n+1} = (T_n)^2 - T_n + 1\) for all \(n \geq 2\).

    Proof. We can prove this simply by applying the recursive formula twice, once for \(T_{n+1}\) and once for \(T_n\): \[\begin{aligned} T_{n+1} &= 1 + (T_1 \cdot T_2 \cdot T_3 \cdots T_n) && \text{by the recursive definition} \\ &= 1 + T_n \cdot (T_1 \cdot T_2 \cdot T_3 \cdots T_{n-1}) \\ &= 1 + T_n \cdot (T_n - 1) && \text{since } T_n = 1 + (T_1 \cdots T_{n-1}) \text{ by rec. def.} \\ &= 1 + (T_n)^2 - T_n \\ &= (T_n)^2 - T_n + 1. \end{aligned}\] Note that both applications of the recursive definition are valid, since \(n - 1 \geq 1\). \(\square\)

  3. Fibonacci explicit formula.

    Proof. Let \(P(n)\) be the predicate \[a_n = \frac{\phi^n - \psi^n}{\sqrt{5}}.\]

    Basis. The base cases are \(n = 0, 1\), which are true because \[\frac{\phi^0 - \psi^0}{\sqrt{5}} = \frac{1-1}{\sqrt{5}} = 0 = a_0,\] and \[\frac{\phi - \psi}{\sqrt{5}} = \frac{\frac{1+\sqrt{5}}{2} - \frac{1-\sqrt{5}}{2}}{\sqrt{5}} = \frac{\sqrt{5}}{\sqrt{5}} = 1 = a_1.\]

    Inductive hypothesis. Now suppose that for some integer \(m \geq 1\), we have that \(P(i)\) is true for all \(0 \leq i \leq m\). In particular, \(P(m)\) and \(P(m-1)\) are true, that is, \[a_m = \frac{\phi^m - \psi^m}{\sqrt{5}} \qquad \text{and} \qquad a_{m-1} = \frac{\phi^{m-1} - \psi^{m-1}}{\sqrt{5}}.\]

    We need to show that \(P(m+1)\) is true. That is, we aim to show \[a_{m+1} = \frac{\phi^{m+1} - \psi^{m+1}}{\sqrt{5}}.\]

    Starting with the right-hand side of \(P(m+1)\), we have \[\begin{aligned} \frac{\phi^{m+1} - \psi^{m+1}}{\sqrt{5}} &= \frac{\phi^2 \phi^{m-1} - \psi^2 \psi^{m-1}}{\sqrt{5}} \\ &= \frac{(\phi+1)\phi^{m-1} - (\psi+1)\psi^{m-1}}{\sqrt{5}} && \text{since } \phi^2 = \phi+1 \text{ and } \psi^2 = \psi+1 \\ &= \frac{(\phi^m + \phi^{m-1}) - (\psi^m + \psi^{m-1})}{\sqrt{5}} \\ &= \frac{\phi^m - \psi^m}{\sqrt{5}} + \frac{\phi^{m-1} - \psi^{m-1}}{\sqrt{5}} \\ &= a_m + a_{m-1} && \text{by the inductive hypothesis} \\ &= a_{m+1} && \text{by definition.} \end{aligned}\] Hence \(P(m+1)\) is true. \(\square\)

§21 — Set Theory Definitions

Epp 4th ed. pp. 336–351 · 5th ed. (metric) pp. 377–391. Lecture 17 (week 6).

Problems

  1. Define the sets

    \[\begin{aligned} V &= \{x \in \mathbb{Z} \mid 2 \leqslant x \leqslant 11 \text{ and } x \text{ is odd}\}, \\ X &= \{x \in \mathbb{Z} \mid -1 \leqslant x \leqslant 4\}, \\ Y &= \{2, 3, 7, 9\}, \\ Z &= \{x \in \mathbb{Z} \mid x \neq 0\}. \end{aligned}\]

    Write down the following sets, listing their elements and using brackets as appropriate.

    1. \(V \cap X\). (b) \(V \cup Y\). (c) \(X \cap Z\).

    2. \(X \cap (Z - Y)\). (e) \((X \cup Y) \cap V\). (f) \(X \cup (Y \cap V)\).

  2. For each of the following sets, write down its elements and then state the size of the set; that is, find \(|A|\), \(|B|\), \(|C|\) and \(|D|\).

    1. \(A = \{x \in \mathbb{Z} \mid \exists\, i \in \mathbb{Z} \text{ such that } x = 1 + (-1)^i\}\).

    2. \(B = \{y \in \mathbb{Z} \mid -8 < y \leqslant -5\}\).

    3. \(C = \{s \in \mathbb{Z} \mid s^2 + 1 = 0\}\).

    4. \(D = \{t \in \mathbb{Z} \mid t^4 = 1\}\).

  3. Define the sets \(A = \{0,1,2,3,4\}\), \(B = \{3,4,5,6\}\) and \(C = \{1,2,4,5,7\}\). List the elements of the set

    \[R = \big((A - B) \cap (A - C)\big) \cup (B \cap C \cap A).\]

  4. Are the following statements true or false? Justify in one or two sentences.

    1. \(x \in \{x\}\) (b) \(\{x\} \subseteq \{x\}\) (c) \(\{\emptyset\} \in \{\{\emptyset\}\}\)

    2. \(\emptyset \subseteq \{x\}\) (e) \(\emptyset \in \{x\}\)

Solutions

    1. \(V \cap X = \{3\}\)

    2. \(V \cup Y = \{2, 3, 5, 7, 9, 11\}\)

    3. \(X \cap Z = \{-1, 1, 2, 3, 4\}\)

    4. \(X \cap (Z - Y) = \{-1, 1, 4\}\)

    5. \((X \cup Y) \cap V = \{3, 7, 9\}\)

    6. \(X \cup (Y \cap V) = \{-1, 0, 1, 2, 3, 4, 7, 9\}\)

    1. Note that \((-1)^i\) is either \(+1\) or \(-1\) (according as \(i\) is even or odd). So \(x\) is either \(1+1\) or \(1-1\), that is, 2 or 0. So \(A = \{2, 0\}\) and \(|A| = 2\).

    2. \(B = \{-7, -6, -5\}\), so \(|B| = 3\).

    3. If \(s^2 + 1 = 0\), we have \(s^2 = -1\), and there are no such real numbers \(s\), let alone integers! So \(C = \emptyset\) and \(|C| = 0\).

    4. \(D = \{-1, 1\}\), since \(t^4 = 1\) with \(t \in \mathbb{Z}\) implies \(t = \pm 1\). Hence \(|D| = 2\).

  1. We have \(A - B = \{0,1,2\}\) and \(A - C = \{0,3\}\), so \((A-B) \cap (A-C) = \{0\}\). Also \(B \cap C \cap A = \{4\}\). Thus \(R = \{0\} \cup \{4\} = \{0, 4\}\).

    1. True: \(\{x\}\) is the set whose only element is \(x\), so \(x \in \{x\}\).

    2. True: every set is a subset of itself, so \(\{x\} \subseteq \{x\}\).

    3. True: \(\{\{\emptyset\}\}\) has the single element \(\{\emptyset\}\), so \(\{\emptyset\} \in \{\{\emptyset\}\}\).

    4. True: \(\emptyset \subseteq S\) for every set \(S\), in particular \(\emptyset \subseteq \{x\}\).

    5. False (in general): \(\emptyset \in \{x\}\) would require \(\emptyset = x\).

Application — Computer Representations of Sets

Not assessable for MATH1061. Taken from K. Rosen, Discrete Mathematics and its Applications, 7th edition, 1991, pages 134–138.

There are various ways to represent sets using a computer. One such method involves the use of binary strings.

If we fix a finite universal set \(U\) with an arbitrary ordering, for example \(U = \{a_1, a_2, \dots, a_n\}\), we can represent subsets of \(U\) using binary strings of length \(n\). A 1 in the \(i\)th position of the string indicates that element \(a_i\) is in the subset and a 0 indicates it is not.

For example if \(U = \{1,2,3\}\) then the string \(000\) represents the empty set, \(101\) represents the subset \(\{1,3\}\) and \(111\) represents the entire set \(U\).

The set operations union and intersection can be computed at the level of binary strings by utilising the bitwise OR and AND operations. The union of two sets corresponds to the bitwise OR operation on the strings and the intersection corresponds to bitwise AND.

Application problems
  1. Define the universal set \(U = \{1,2,3,4,5,6,7,8,9,10\}\).

    1. Express each of the following subsets of \(U\) using binary strings:
    1. \(\{3,4,5\}\)

    2. \(\{1,3,6,10\}\)

    3. \(\{a \in U \mid a \text{ is prime}\}\).

    1. Using the same \(U\), find the set represented by the following strings:
    1. \(1111001111\)

    2. \(0101111000\)

    3. \(1000000001\)

    1. Use binary strings to compute the following set operations.
    1. \(\{1,3,4,6,10\} \cap \{2,3,6,9,10\}\)

    2. \(\{2,4,6,8\} \cup \{3,4,5\}\)

Application solutions
      1. \(0011100000\)
    1. \(1010010001\)

    2. \(0110101000\) — the primes in \(U\) are \(2, 3, 5, 7\). (The source PDF prints \(1010111000\) here, which decodes to \(\{1,3,5,6,7\}\) and is not the set of primes; the string above is the correct one.)

      1. \(\{1,2,3,4,7,8,9,10\}\)
    1. \(\{2,4,5,6,7\}\)

    2. \(\{1,10\}\)

      1. \(1011010001 \text{ AND } 0110010011 = 0010010001\), so the intersection is \(\{3,6,10\}\).
    1. \(0101010100 \text{ OR } 0011100000 = 0111110100\), so the union is \(\{2,3,4,5,6,8\}\).

§22 — More Definitions and Examples of Sets

Epp 4th ed. pp. 336–351 · 5th ed. (metric) pp. 377–391. Lecture 17 (week 6).

Problems

  1. Let \(A = \{1,2\}\), \(B = \{2,3\}\) and \(C = \{3,4\}\). List the elements of the following sets.

    1. \(A \cap C\)

    2. \((A \cup C) - B\)

    3. \((A \times B) \cap (A \times C)\)

    4. \(\mathcal{P}(B \cap C)\)

    5. \(\{x : (x \subseteq A) \wedge (x \in \mathcal{P}(B))\}\)

  2. Are the following statements true or false?

    1. \(\mathbb{Q} \in \mathbb{R}\).

    2. \(\{1,2\} \subseteq \mathbb{Z}\).

    3. \(\{-3, 0\} \subseteq \mathcal{P}(\mathbb{Z})\).

    4. \(\{(1,2), (0,1)\} \in \mathbb{Z} \times \mathbb{Z}\).

  3. Let \(X = \mathcal{P}(\emptyset)\) and \(Y = \mathcal{P}(\mathcal{P}(\emptyset))\). List the elements of each of the following sets.

    1. \(X\) (b) \(Y\) (c) \(X \cap Y\) (d) \(X \cup Y\)
  4. Define sets \(S = \{3,5\}\) and \(T = \{2,3,7\}\). Are the following statements true or false?

    1. \(S \subseteq T\)

    2. \(\emptyset \in \mathcal{P}(S)\)

    3. \(\emptyset \in T\)

    4. \((3,3,3) \in S \times T\)

Solutions

    1. \(A \cap C = \emptyset\).

    2. \((A \cup C) - B = \{1, 4\}\).

    3. \((A \times B) \cap (A \times C) = \{(1,3), (2,3)\}\).

    4. \(\mathcal{P}(B \cap C) = \{\emptyset, \{3\}\}\).

    5. \(\{x : (x \subseteq A) \wedge (x \in \mathcal{P}(B))\} = \{\emptyset, \{2\}\}\).

    1. False. (\(\mathbb{Q}\) is a subset of \(\mathbb{R}\), not an element of it.)

    2. True.

    3. False. (The elements of \(\mathcal{P}(\mathbb{Z})\) are sets of integers; \(-3\) and \(0\) are integers, not sets.)

    4. False. (It is a subset of \(\mathbb{Z} \times \mathbb{Z}\), not an element — the elements of \(\mathbb{Z} \times \mathbb{Z}\) are ordered pairs.)

    1. \(X = \mathcal{P}(\emptyset) = \{\emptyset\}\).

    2. \(Y = \mathcal{P}(\mathcal{P}(\emptyset)) = \{\emptyset, \{\emptyset\}\}\).

    3. \(X \cap Y = \{\emptyset\}\).

    4. \(X \cup Y = \{\emptyset, \{\emptyset\}\}\).

    1. False. (\(5 \in S\) but \(5 \notin T\).)

    2. True. (The empty set is an element of every power set.)

    3. False. (\(T\)’s elements are \(2\), \(3\) and \(7\).)

    4. False. (The elements of \(S \times T\) are ordered pairs, not triples.)

§23 — Properties of Sets

Epp 4th ed. pp. 352–366 · 5th ed. (metric) pp. 391–407. Lecture 18 (week 7).

Problems

    1. Use the element method to prove that for all sets \(A\) and \(B\)

    \[A \cup (B - A) = A \cup B.\]

    1. Use the element method to prove that for all sets \(X\), \(Y\) and \(Z\),

    \[(X - Z) \cap (Y - Z) \cap (X - Y) = \emptyset.\]

    1. Use set identities to prove that for all sets \(A\), \(B\) and \(C\)

    \[(A - B) - C = A - (B \cup C).\]

  1. Prove or disprove the following:

    1. For all sets \(A\) and \(B\), we have \(\mathcal{P}(A \cup B) = \mathcal{P}(A) \cup \mathcal{P}(B)\).

    2. For all sets \(A\) and \(B\), we have \(\mathcal{P}(A) = \mathcal{P}(B) \iff A = B\).

    3. There exists a set \(A\) such that \(\mathcal{P}(A) = \{\{a\}, \{b\}, \{c\}, \{a,b\}, \{b,c\}, \{a,c\}, \{a,b,c\}\}\).

    4. Let \(A\), \(B\) and \(C\) be sets. If \(B \subseteq C\) then \(A \times B \subseteq A \times C\).

  2. Let \(A\), \(B\) and \(C\) be sets. Prove that

    \[A - (B \cup C) = (A - B) - C.\]

Solutions

    1. For all sets \(A\) and \(B\), \(A \cup (B - A) = A \cup B\).

    Proof. First we show that \(A \cup (B-A) \subseteq A \cup B\).

    Consider an arbitrary element \(x \in A \cup (B-A)\). Then we know \(x \in A\) or \(x \in B - A\). This means \(x \in A\), or \(x \in B\) and \(x \notin A\). If \(x \in A\), then \(x \in A \cup B\). Similarly if \(x \in B - A\), then it must be in \(B\) and so \(x \in A \cup B\). In either case, we have \(x \in A \cup B\), thus \(A \cup (B-A) \subseteq A \cup B\).

    Now we show that \(A \cup B \subseteq A \cup (B-A)\).

    Let \(x \in A \cup B\), then \(x \in A\) or \(x \in B\) or it is in both \(A\) and \(B\). If \(x \in A\), then \(x \in A \cup (B-A)\). If \(x \notin A\) then \(x \in B\) since \(x \in A \cup B\). So we have that \(x \in B\) and \(x \notin A\), hence \(x \in B - A\) and thus \(x \in A \cup (B-A)\). This proves that \(A \cup B \subseteq A \cup (B-A)\), and together this proves that \(A \cup B = A \cup (B-A)\). \(\square\)

    1. For all sets \(X\), \(Y\) and \(Z\), \((X-Z) \cap (Y-Z) \cap (X-Y) = \emptyset\).

    Proof. Assume that \((X-Z) \cap (Y-Z) \cap (X-Y) \neq \emptyset\).

    So let \(x \in (X-Z) \cap (Y-Z) \cap (X-Y)\). Then, in particular, \(x \in Y - Z\) and \(x \in X - Y\). The former implies \(x \in Y\), but the latter implies \(x \notin Y\). No such \(x\) exists. Therefore the set is empty. \(\square\)

    1. For all sets \(A\), \(B\) and \(C\), \((A-B) - C = A - (B \cup C)\).

    Proof. Using \(A - B = A \cap B^c\), \((S - T) = S \cap T^c\), and De Morgan’s law, we have \[(A-B) - C = (A \cap B^c) \cap C^c = A \cap (B^c \cap C^c) = A \cap (B \cup C)^c = A - (B \cup C). \qquad \square\]

    1. For all sets \(A\) and \(B\), \(\mathcal{P}(A \cup B) = \mathcal{P}(A) \cup \mathcal{P}(B)\). False. Counterexample: \(A = \{1\}\), \(B = \{2\}\). Then

    \[\mathcal{P}(A \cup B) = \{\emptyset, \{1\}, \{2\}, \{1,2\}\} \neq \{\emptyset, \{1\}\} \cup \{\emptyset, \{2\}\} = \{\emptyset, \{1\}, \{2\}\}.\]

    1. For all sets \(A\) and \(B\), \(\mathcal{P}(A) = \mathcal{P}(B) \iff A = B\). True.

    (\(\Longleftarrow\)) If \(A = B\) then clearly \(\mathcal{P}(A) = \mathcal{P}(B)\).

    (\(\Longrightarrow\)) We next show that if \(\mathcal{P}(A) = \mathcal{P}(B)\), then \(A = B\). Suppose (for a contradiction) that \(\mathcal{P}(A) = \mathcal{P}(B)\) but \(A \neq B\). Then (without loss of generality), there exists an element \(x\) in \(A\) that is not in \(B\). Therefore, there is an element \(\{x\}\) in \(\mathcal{P}(A)\) that is not in \(\mathcal{P}(B)\). Therefore \(\mathcal{P}(A) \neq \mathcal{P}(B)\) — a contradiction. Thus, for all sets \(A\) and \(B\), if \(\mathcal{P}(A) = \mathcal{P}(B)\) then \(A = B\). \(\square\)

    1. There exists a set \(A\) such that \(\mathcal{P}(A) = \{\{a\}, \{b\}, \{c\}, \{a,b\}, \{b,c\}, \{a,c\}, \{a,b,c\}\}\). False.

    No such set exists. The power set of any set must contain the empty set — that is, \(\emptyset \in \mathcal{P}(A)\), where \(A\) is any set. Also, the power set of any set of size \(n\) has cardinality \(2^n\). The power set above contains 7 elements, and there does not exist any integer \(n\) such that \(2^n = 7\).

    1. Let \(A\), \(B\) and \(C\) be sets. If \(B \subseteq C\) then \(A \times B \subseteq A \times C\). True.

    First note that if \(A = \emptyset\) then \(A \times B = A \times C = \emptyset\) and the statement is true. Also, if \(B = \emptyset\) then \(A \times B = \emptyset\) and \(\emptyset \subseteq A \times C\) for any \(C\), so the statement is true.

    Suppose \(A\), \(B\) and \(C\) are non-empty sets with \(B \subseteq C\). Thus if \(b \in B\), then \(b \in C\). Consider any element \((a,b) \in A \times B\), where \(a \in A\) and \(b \in B\). Since \(b \in B\), \(b \in C\). As such, \(a \in A\) and \(b \in C\) for all \(a\) or \(b\), so \((a,b) \in A \times C\). Therefore \(A \times B \subseteq A \times C\). \(\square\)

  1. For all sets \(A\), \(B\) and \(C\), \(A - (B \cup C) = (A-B) - C\).

    Proof Option 1. For any \(x\), we have \[x \in A - (B \cup C) \iff x \in A \wedge x \notin (B \cup C) \iff x \in A \wedge (x \notin B) \wedge (x \notin C).\] Also, we have \[x \in (A-B) - C \iff x \in (A-B) \wedge x \notin C \iff (x \in A \wedge x \notin B) \wedge x \notin C.\] The statements on the right-hand side are equivalent. Hence \(A - (B \cup C) = (A-B) - C\). \(\square\)

    Proof Option 2. \[\begin{aligned} A - (B \cup C) &= A \cap (B \cup C)^c && \text{by the Set Difference Law} \\ &= A \cap (B^c \cap C^c) && \text{by De Morgan's law} \\ &= (A \cap B^c) \cap C^c && \text{by Associative law} \\ &= (A - B) \cap C^c && \text{by the Set Difference Law} \\ &= (A - B) - C && \text{by the Set Difference Law} \qquad \square \end{aligned}\]

§24 — Functions Defined on General Sets

Epp 4th ed. pp. 383–396 · 5th ed. (metric) pp. 425–439. Lecture 18 (week 7).

Problems

  1. Consider the function \(f : \mathbb{R} \to \mathbb{R}\) defined by \(f(x) = x^2 - 1\).

    1. What is the domain, codomain and range of \(f\)?

    2. What is the image of \(\{x \in \mathbb{R} \mid -1 \leq x \leq 2\}\) under \(f\)?

    3. What is the preimage of \(\{y \in \mathbb{R} \mid 0 \leq y \leq 3\}\) under \(f\)?

  2. Let \(X = \{1,3,5\}\) and \(Y = \{s,t,u,v\}\) be sets. Define \(f : X \to Y\) by the arrow diagram

    \[1 \mapsto v, \qquad 3 \mapsto s, \qquad 5 \mapsto v\]

    (the elements \(t\) and \(u\) of \(Y\) receive no arrows).

    1. Write the domain of \(f\) and the co-domain of \(f\).

    2. Determine \(f(1)\), \(f(3)\), and \(f(5)\).

    3. Determine the range of \(f\).

    4. Is 3 an inverse image of \(s\)? Is 1 an inverse image of \(u\)?

    5. What is the inverse image of \(s\)? of \(u\)? of \(v\)?

    6. Represent \(f\) as a set of ordered pairs.

  3. Let \(D\) be the set of all finite subsets of \(\mathbb{Z}^+\). Consider the function \(T : \mathbb{Z}^+ \to D\) defined by \(T(n) := \{d \mid d \in \mathbb{Z}^+ \text{ and } d \text{ divides } n\}\). Find the following:

    1. \(T(1)\) (b) \(T(15)\) (c) \(T(17)\) (d) \(T(5)\) (e) \(T(18)\) (f) \(T(21)\)
  4. Let \(S\) be the set of all strings of \(a\)’s and \(b\)’s.

    1. Define the function \(f : S \to \mathbb{Z}\) by

    \[f(s) = \begin{cases} \text{the number of } b\text{'s to the left of the left-most } a \text{ in } s, & \text{if } s \text{ contains an } a, \\ 0, & \text{if } s \text{ contains no } a\text{'s.} \end{cases}\]

    Find \(f(aba)\), \(f(bbab)\), and \(f(b)\). What is the range of \(f\)?

    1. Define the function \(g : S \to S\) by

    \[g(s) = \text{the string obtained by writing the characters of } s \text{ in reverse order.}\]

    Find \(g(aba)\), \(g(bbab)\), and \(g(b)\). What is the range of \(g\)?

Solutions

    1. The domain is \(\mathbb{R}\), the codomain is \(\mathbb{R}\), and the range is \([-1, \infty)\).

    2. \(f\big(\{x \in \mathbb{R} \mid -1 \leq x \leq 2\}\big) = [-1, 3]\).

    3. \(f^{-1}\big(\{y \in \mathbb{R} \mid 0 \leq y \leq 3\}\big) = \{x \in \mathbb{R} \mid 1 \leq x^2 \leq 4\} = [-2,-1] \cup [1,2]\).

    1. The domain is \(X = \{1,3,5\}\) and the codomain is \(Y = \{s,t,u,v\}\).

    2. \(f(1) = v\), \(f(3) = s\), \(f(5) = v\).

    3. \(\operatorname{range}(f) = \{s, v\}\).

    4. Yes, 3 is an inverse image of \(s\). No, 1 is not an inverse image of \(u\).

    5. \(f^{-1}(s) = \{3\}\), \(f^{-1}(u) = \emptyset\), \(f^{-1}(v) = \{1, 5\}\).

    6. \(f = \{(1,v), (3,s), (5,v)\}\).

    1. \(T(1) = \{1\}\).

    2. \(T(15) = \{1, 3, 5, 15\}\).

    3. \(T(17) = \{1, 17\}\).

    4. \(T(5) = \{1, 5\}\).

    5. \(T(18) = \{1, 2, 3, 6, 9, 18\}\).

    6. \(T(21) = \{1, 3, 7, 21\}\).

    1. \(f(aba) = 0\), \(f(bbab) = 2\), \(f(b) = 0\), \(\operatorname{range}(f) = \mathbb{Z}^{\geq 0}\).

    2. \(g(aba) = aba\), \(g(bbab) = babb\), \(g(b) = b\), \(\operatorname{range}(g) = S\).

§25 — one-to-one, Onto, and Inverse Functions

Epp 4th ed. pp. 397–416 · 5th ed. (metric) pp. 439–461. Lecture 19 (week 7).

Problems

  1. Consider the function \(f : \mathbb{Z} \to \mathbb{Z}\) defined by \(f(n) = \left\lfloor \dfrac{1-6n}{3} \right\rfloor\).

    1. Is \(f\) one-to-one?

    2. Is \(f\) onto?

    3. Is \(f\) bijective?

    4. Prove or disprove that for all \(m, n \in \mathbb{Z}\), we have \(f(mn) = f(m)f(n)\).

  2. Define the set \(A = \{0,1,2,3,4\}\) and the functions \(f : A \to A\) and \(g : A \to A\) by

    \[f(a) = (a+4)^2 \pmod 5, \qquad g(a) = (a^2 + 3a + 1) \pmod 5.\]

    Is \(f = g\)?

  3. Consider the function \(f : \mathbb{Z}^+ \times \mathbb{Z}^+ \to \mathbb{Z}^+\) defined by \(f((m,n)) = 2^m 3^n\).

    1. Determine \(f((3,1))\).

    2. Determine the pre-image of \(\{36\}\).

    3. Prove \(f\) is one-to-one.

  4. Consider the function \(f : \mathbb{R} \to \mathbb{Z}\) defined by \(f(x) = \lfloor x^2 \rfloor\).

    1. What is the image \(f([-1,1])\)?

    2. What is the preimage \(f^{-1}(\{1\})\)?

    3. Is \(f\) one-to-one?

    4. Is \(f\) onto?

    5. Is \(f\) bijective?

Solutions

  1. First, we note that, for \(n \in \mathbb{Z}\),

    \[f(n) = \left\lfloor \frac{1-6n}{3} \right\rfloor = \left\lfloor -2n + \frac{1}{3} \right\rfloor = -2n\]

    since \(-2n \in \mathbb{Z}\) and \(\frac{1}{3} \in [0,1)\).

    1. Yes. If \(f(m) = f(n)\) then \(-2m = -2n\), hence \(m = n\). Therefore \(f\) is one-to-one.

    2. No. Since \(f(n) = -2n\) we see that \(f\) never maps to any odd integer; for example, there is no integer \(w\) with \(f(w) = 1\). Therefore \(f\) is not onto.

    3. No, \(f\) is not a bijection. (It is injective but not surjective.)

    4. This is false; take \(m = n = 1\). Then \(f(mn) = -2\) but \(f(m)f(n) = 4\).

  2. Yes. First note \(f\) and \(g\) have the same domain and codomain. This means we just need to show \(f(a) = g(a)\) for all \(a \in A\). Notice that we have

    \[(a+4)^2 = a^2 + 8a + 16 \equiv a^2 + 3a + 1 \pmod 5,\]

    so \(f(a) = g(a)\) for all \(a \in A\), and hence \(f = g\).

    1. \(f((3,1)) = 2^3 3^1 = 24\).

    2. \(36 = 2^2 3^2\), so \(f((m,n)) = 36\) if and only if \((m,n) = (2,2)\). Hence \(f^{-1}(\{36\}) = \{(2,2)\}\).

    3. If \(f((m,n)) = f((m',n'))\), then \(2^m 3^n = 2^{m'} 3^{n'}\), so \(2^{m-m'} = 3^{n'-n}\). By unique prime factorisation, this forces \(m - m' = 0\) and \(n' - n = 0\). That is, \(m = m'\) and \(n = n'\), and hence \((m,n) = (m',n')\). So \(f\) is one-to-one.

    1. Note that if \(x \in [-1,1]\) then \(x^2 \in [0,1]\). This means we have \(\lfloor x^2 \rfloor \in \{0,1\}\). Both values occur because \(\lfloor 0^2 \rfloor = 0\) and \(\lfloor 1^2 \rfloor = 1\). Thus \(f([-1,1]) = \{0,1\}\).

    2. \(f^{-1}(\{1\}) = \{x \in \mathbb{R} : 1 \leq x^2 < 2\} = (-\sqrt{2}, -1] \cup [1, \sqrt{2})\).

    3. No, \(f\) is not one-to-one. For example, \(f\!\left(\frac{1}{2}\right) = 0 = f\!\left(-\frac{1}{2}\right)\) but \(\frac{1}{2} \neq -\frac{1}{2}\).

    4. No, \(f\) is not onto since \(f(x) = \lfloor x^2 \rfloor \geq 0\) for all \(x\). This means, for instance, that \(-1\) is not in the image of \(f\).

    5. No, \(f\) is not bijective.

§26 — Composition of Functions

Epp 4th ed. pp. 416–427 · 5th ed. (metric) pp. 461–472. Lecture 20 (week 7).

Problems

  1. Let \(f, g : \mathbb{Z} \to \mathbb{Z}\) be defined by \(f(x) = x^3 - 1\) and \(g(x) = x + 5\).

    1. Determine the functions \((g \circ f)(x)\) and \((f \circ g)(x)\).

    2. Is \(g \circ f\) a bijection?

  2. Define \(f : \mathbb{Q} \to \mathbb{Z}\) by \(f(x) = 2\lfloor x + 1 \rfloor\).

    1. What is the range of \(f\)?

    2. Is \(f\) one-to-one?

    3. Is \(f\) onto?

    4. Define \(g : \mathbb{Z} \to \mathbb{Z}\) by \(g(x) = 2x - 1\). Determine the composition \((g \circ f)(x)\).

    5. Calculate \((g \circ f)\left(-\frac{3}{4}\right)\).

    6. What is the range of \(g \circ f\)?

  3. Let \(f\) and \(g\) be functions. Prove or disprove the following statements:

    1. If \(f\) and \(f \circ g\) are one-to-one, then \(g\) is one-to-one.

    2. If \(f\) and \(f \circ g\) are onto, then \(g\) is onto.

Solutions

    1. We have

    \[(g \circ f)(x) = g(x^3 - 1) = x^3 + 4\]

    and

    \[(f \circ g)(x) = f(x+5) = (x+5)^3 - 1 = x^3 + 15x^2 + 75x + 124.\]

    1. No, \(g \circ f\) is not a bijection. The map \(x \mapsto x^3 + 4\) is one-to-one on \(\mathbb{Z}\), but it is not onto (e.g. 0 is not in its image since \(x^3 = -4\) has no integer solution).
    1. The range is a subset of the even integers because \(\lfloor x+1 \rfloor \in \mathbb{Z}\). If \(2m\) is an even integer then \(f(m-1) = 2\lfloor m-1+1 \rfloor = 2m\), so the range is the set of all even integers.

    2. No, \(f\) is not one-to-one. For example,

    \[f(0) = 2\lfloor 0+1 \rfloor = 2\lfloor 1 \rfloor = 2 \times 1 = 2, \quad \text{and}\] \[f\!\left(\tfrac{1}{2}\right) = 2\left\lfloor \tfrac{1}{2} + 1 \right\rfloor = 2\left\lfloor \tfrac{3}{2} \right\rfloor = 2 \times 1 = 2,\]

    but \(0 \neq \frac{1}{2}\).

    1. No, \(f\) is not onto since \(\operatorname{range}(f) \neq \mathbb{Z}\) (e.g. there is no \(x \in \mathbb{Q}\) such that \(f(x) = 1\) because 1 is odd).

    2. We have

    \[(g \circ f)(x) = g(f(x)) = 2\big(2\lfloor x+1 \rfloor\big) - 1 = 4\lfloor x+1 \rfloor - 1.\]

    1. We have

    \[(g \circ f)\left(-\frac{3}{4}\right) = 4 \cdot 0 - 1 = -1.\]

    1. The range is a subset of \(\{4n - 1 \mid n \in \mathbb{Z}\}\) because \(\lfloor x+1 \rfloor \in \mathbb{Z}\). This is the range because, given an integer \(4n-1\), we have \(f(n-1) = 4\lfloor n-1+1 \rfloor - 1 = 4n-1\). Note that you can also write the range as \(\{n \in \mathbb{Z} \mid n \equiv -1 \pmod 4\}\) which is the same as \(\{n \in \mathbb{Z} \mid n \equiv 3 \pmod 4\}\).
    1. The statement is true.

    Proof. Suppose that \(f\) and \(f \circ g\) are one-to-one functions. Suppose for a contradiction that \(g\) is not one-to-one. Thus there exist two distinct elements in the domain of \(g\), say \(x_1\) and \(x_2\), for which \(g(x_1) = g(x_2)\). Let \(g(x_1) = y = g(x_2)\). Thus \(f(g(x_1)) = f(y) = f(g(x_2))\), but \(x_1 \neq x_2\), which contradicts the fact that \(f \circ g\) is one-to-one. Therefore \(g\) must be one-to-one. \(\square\)

    1. The statement is false.

    Counterexample. Consider the functions \(g : \{1\} \to \{2,3\}\) defined by \(g(1) = 2\), and \(f : \{2,3\} \to \{4\}\) defined by \(f(2) = 4\) and \(f(3) = 4\). Then \(f\) is onto, and \(f \circ g\) is onto (since \(f \circ g : \{1\} \to \{4\}\) and \(f(g(1)) = 4\)). However, \(g\) is not onto. \(\square\)

§27 — Cardinalities

Epp 4th ed. pp. 428–441 · 5th ed. (metric) pp. 473–486. Lecture 21 (week 8).

Problems

  1. Let \(A = \{-2,-1,0,1\}\) and \(B = \{3,4,5,6\}\). Prove that \(|A| = |B|\) by providing a bijection between \(A\) and \(B\).

  2. Prove that \(|\{1,2,3\}| \leq |\mathbb{Z}|\) by providing an appropriate injective (one-to-one) function.

  3. Consider the intervals

    \[[6,12] = \{x \in \mathbb{R} \mid 6 \leq x \leq 12\} \quad \text{and} \quad (6,12) = \{x \in \mathbb{R} \mid 6 < x < 12\}.\]

    Prove that \(|[6,12]| \leq |(6,12)|\) by providing an appropriate injective (one-to-one) function.

  4. Let \(A = \{x \in \mathbb{R} \mid -1 \leq x \leq 1\}\) and \(B = \{x \in \mathbb{R} \mid 3 \leq x \leq 7\}\). Prove that \(|A| = |B|\) by providing a bijection between \(A\) and \(B\).

Solutions

  1. Define the function \(f : A \to B\) by

    \[f(-2) = 3, \qquad f(-1) = 4, \qquad f(0) = 5, \qquad f(1) = 6.\]

    Then \(f\) is bijective, hence \(|A| = |B|\).

  2. Define the function \(g : \{1,2,3\} \to \mathbb{Z}\) by \(g(1) = 1\), \(g(2) = 2\), and \(g(3) = 3\). Then \(g\) is injective (one-to-one), so \(|\{1,2,3\}| \leq |\mathbb{Z}|\).

  3. Define \(h : [6,12] \to (6,12)\) by

    \[h(x) = \frac{x}{2} + 4.\]

    For each \(x \in [6,12]\) we have

    \[\begin{aligned} \frac{6}{2} + 4 \leq\ &h(x) \leq \frac{12}{2} + 4 \\ 7 \leq\ &h(x) \leq 10. \end{aligned}\]

    Thus \(h\) is indeed a function from \([6,12]\) to \((6,12)\), and \(h\) has range \([7,10] \subseteq (6,12)\).

    Suppose \(x, y \in [6,12]\) and \(h(x) = h(y)\). Then \(\frac{x}{2} + 4 = \frac{y}{2} + 4\), so \(x = y\). Therefore \(h\) is injective (one-to-one).

    Thus \(|[6,12]| \leq |(6,12)|\).

  4. Let \(A = \{x \in \mathbb{R} \mid -1 \leq x \leq 1\}\) and \(B = \{x \in \mathbb{R} \mid 3 \leq x \leq 7\}\).

    Consider the function \(f : A \to B\) where \(f(x) = 2x + 5\). Since \(x \geq -1\), \(f(x) = 2x+5 \geq 2(-1)+5 = 3\), and since \(x \leq 1\), \(f(x) = 2x+5 \leq 2(1)+5 = 7\). Also each value of \(x\) gives a unique value of \(2x+5\). Thus \(f\) is a function with the correct domain and codomain.

    We now prove that \(f\) is one-to-one. Suppose that \(x_1, x_2 \in A\) and \(f(x_1) = f(x_2)\). Thus \(2x_1 + 5 = 2x_2 + 5\), and so \(x_1 = x_2\). Thus \(f\) is one-to-one.

    We now prove that \(f\) is onto. Suppose \(y \in [3,7]\). Let \(x = \frac{y-5}{2}\). Since \(y \leq 7\), we have \(x \leq 1\) and since \(y \geq 3\), we have \(x \geq -1\). So \(x \in [-1,1]\). Furthermore \(f(x) = f\!\left(\frac{y-5}{2}\right) = 2\left(\frac{y-5}{2}\right) + 5 = y - 5 + 5 = y\). Thus \(f\) is onto.

    Hence \(f\) is a bijection and so \(|A| = |B|\).

§28 — Countable and Uncountable Sets

Epp 4th ed. pp. 428–441 · 5th ed. (metric) pp. 473–486. Lecture 22 (week 8).

Note that this content will not be assessed in MATH1061.

Problems

  1. Determine whether each of the following sets is countable or uncountable.

    1. \(\{1, 2, \dots, n\}\).

    2. \(\mathbb{Z} \times \mathbb{Z}\).

    3. \(\mathbb{Q} \times \mathbb{Q}\).

    4. \(\mathbb{Z} \times \mathbb{Z} \times \cdots\) (infinitely many copies of \(\mathbb{Z}\)).

    1. Is \(|\mathbb{Z}| = |\mathbb{Q}|\)?

    2. Is \(|\mathbb{Z}| = |\mathbb{Z} \times \mathbb{Z}|\)?

    3. Is \(|\mathbb{R}| = |\mathbb{Q} \times \mathbb{Q}|\)?

    1. Show that the union of a countable number of countable sets is countable.

    2. Show that the set of all finite binary strings is countable.

    3. Use (a) to show that the set of all irrational numbers is uncountable.

Solutions

    1. The set is countable because the function \(f : \{1,2,\dots,n\} \to \mathbb{N}\) defined by \(f(i) = i\) is injective.

    2. The set \(\mathbb{Z} \times \mathbb{Z}\) is countable. Define the function \(f : \mathbb{Z} \times \mathbb{Z} \to \mathbb{N}\) by

    \[f(a,b) = \begin{cases} 2^a 3^b, & a > 0,\ b > 0, \\ 5^{-a} 7^{-b}, & a < 0,\ b < 0, \\ 11^{a} 13^{-b}, & a > 0,\ b < 0, \\ 17^{-a} 19^{b}, & a < 0,\ b > 0, \\ 23^{b}, & a = 0,\ b > 0, \\ 29^{a}, & a > 0,\ b = 0, \\ 31^{-b}, & a = 0,\ b < 0, \\ 37^{-a}, & a < 0,\ b = 0, \\ 1, & (a,b) = (0,0). \end{cases}\]

    We claim \(f\) is injective. Suppose \(f(a,b) = f(c,d)\). By unique prime factorisation, the set of primes dividing \(f(a,b)\) and the set of primes dividing \(f(c,d)\) must be equal. Hence \((a,b)\) and \((c,d)\) lie in the same case. Again, by unique prime factorisation, the corresponding exponents must agree, so \(a = c\) and \(b = d\). Therefore \(f\) is injective.

    1. The set \(\mathbb{Q} \times \mathbb{Q}\) is countable. First define an injection \(e : \mathbb{Z} \to \mathbb{N}\) by

    \[e(z) = \begin{cases} 2z, & z \geq 0, \\ -2z - 1, & z < 0. \end{cases}\]

    Next define an injection \(c : \mathbb{N} \times \mathbb{N} \to \mathbb{N}\) by

    \[c(m,n) = 2^m 3^n.\]

    Write each rational in lowest terms with positive denominator: \(x = \frac{p}{q}\), \(y = \frac{r}{s}\) with \(p, r \in \mathbb{Z}\) and \(q, s \in \mathbb{N}\). Define \(F : \mathbb{Q} \times \mathbb{Q} \to \mathbb{N}\) by

    \[F\left(\frac{p}{q}, \frac{r}{s}\right) = 2^{\,c(e(p),\,q)}\, 3^{\,c(e(r),\,s)}.\]

    If \(F\left(\frac{p}{q}, \frac{r}{s}\right) = F\left(\frac{p'}{q'}, \frac{r'}{s'}\right)\), then by unique prime factorisation we get \(c(e(p),q) = c(e(p'),q')\) and \(c(e(r),s) = c(e(r'),s')\). Since \(c\) is injective, we have \(e(p) = e(p')\), \(q = q'\), \(e(r) = e(r')\), \(s = s'\). Moreover, since \(e\) is injective, we have \(p = p'\) and \(r = r'\). Therefore \(\frac{p}{q} = \frac{p'}{q'}\) and \(\frac{r}{s} = \frac{r'}{s'}\), so \(F\) is injective.

    1. The set \(\mathbb{Z} \times \mathbb{Z} \times \dots\) is not countable. We will prove this by giving an injection \(\mathbb{R} \to \mathbb{Z} \times \mathbb{Z} \times \cdots\).

    For each \(x \in \mathbb{R}\), write \(x = n + r\) with \(n = \lfloor x \rfloor \in \mathbb{Z}\) and \(r \in [0,1)\). Choose the unique decimal expansion of \(r\) that does not end with an infinite tail of 9’s. Write this decimal expansion as

    \[r = 0.d_1 d_2 d_3 \cdots, \qquad d_k \in \{0,1,\dots,9\}.\]

    Then define the function \(\iota : \mathbb{R} \to \mathbb{Z} \times \mathbb{Z} \times \cdots\) by

    \[\iota(x) = (n, d_1, d_2, d_3, \dots) \in \mathbb{Z} \times \mathbb{Z} \times \cdots.\]

    This is injective; if \(\iota(x) = \iota(y)\) then the decimals of \(x\) and \(y\) are equal, so \(x = y\).

    1. Yes. Both \(\mathbb{Z}\) and \(\mathbb{Q}\) are countable and infinite.

    2. Yes. Both \(\mathbb{Z}\) and \(\mathbb{Z} \times \mathbb{Z}\) are countable and infinite.

    3. No. \(\mathbb{Q} \times \mathbb{Q}\) is countable, while \(\mathbb{R}\) is uncountable.

    1. Let \(A_1, A_2, A_3, \dots\) be countable. Then for each \(n\) we can list \(A_n\) as

    \[A_n = \{a_{n,1}, a_{n,2}, a_{n,3}, \dots\}.\]

    Now list the union by going along diagonals:

    \[a_{1,1};\ a_{1,2}, a_{2,1};\ a_{1,3}, a_{2,2}, a_{3,1};\ a_{1,4}, a_{2,3}, a_{3,2}, a_{4,1};\ \dots\]

    Every element of every \(A_n\) appears somewhere in this list, so \(\bigcup_{n=1}^{\infty} A_n\) can be listed in a sequence. Hence \(\bigcup_{n=1}^{\infty} A_n\) is countable.

    1. Let \(B_n\) be the set of all binary strings of length \(n\). Each \(B_n\) is finite (there are \(2^n\) such strings), hence countable. Every finite binary string has some length, so

    \[B = \{\text{all finite binary strings}\} = \bigcup_{n=0}^{\infty} B_n.\]

    By (a), a countable union of countable sets is countable, so \(B\) is countable.

    1. Suppose, for contradiction, that the set of irrationals \(\mathbb{R} \setminus \mathbb{Q}\) were countable. We already know \(\mathbb{Q}\) is countable. Then by part (a),

    \[\mathbb{R} = \mathbb{Q} \cup (\mathbb{R} \setminus \mathbb{Q})\]

    would be a union of two countable sets, hence countable. But \(\mathbb{R}\) is uncountable. This contradiction shows that \(\mathbb{R} \setminus \mathbb{Q}\) is uncountable.

Application — Computability of Functions

Not assessable for MATH1061. Taken from K. Rosen, Discrete Mathematics and its Applications, 7th edition, 1991, pages 170–177.

An important application of the topic of countability to computer science is examining the computability of functions. In computer science, a function is computable if there is a computer program in some programming language that can find the values of the function. If there is no such program, we say the function is uncomputable.

Application problems
    1. Show the set of all computer programs in any given computer language is countable.

    Hint: a computer program written in a programming language can be thought of as a string of symbols from a finite alphabet.

    1. Let \(T\) be the set of all functions from \(\mathbb{Z}^+\) to the set \(\{0,1,2,3,4,5,6,7,8,9\}\). Show that \(T\) is uncountable.

    Hint: start by finding an injection between the interval \((0,1)\) and a subset of \(T\).

    1. Hence show that there are uncomputable functions.
Application solutions
    1. Given a computer language with a finite alphabet \(A\), we let \(P\) be the set of all computer programs in the language. If \(P\) is finite, then it is countable and we are done, so assume \(P\) is infinite.

    Now we use the fact that we can think of each computer program as a series of strings made using the alphabet \(A\). Consider all strings of length \(m\) made from alphabet \(A\), and denote this set by \(A_m\). It is easy to show that there are \(|A|^m\) distinct strings of this type, and since \(A\) is finite, \(|A|^m\) is also finite. Now \(P\) is a subset of \(\bigcup_{i=1}^{\infty} A_m\). Since the union of a countable number of countable sets is also countable (see Question 3(a) from earlier in this section) it follows that \(\bigcup_{i=1}^{\infty} A_m\) is countable and hence \(P\) is countable.

    1. We use the hint and start by defining a function from \((0,1)\) to \(T\) as follows. Recall that every real number in the interval \((0,1)\) can be written in decimal notation as \(0.a_1 a_2 a_3 \dots a_n \dots\) where each \(a_i\) is an integer from 0 to 9. Also recall that \(T\) is the set of all functions from \(\mathbb{Z}^+\) to \(\{0,1,\dots,9\}\). Let \(f : (0,1) \to T\) be the function defined as

    \[f(0.a_1 a_2 a_3 \dots a_n \dots) = g,\]

    where \(g\) is the function in \(T\) such that \(g(n) = a_n\) for all \(n \in \mathbb{Z}^+\).

    So for example, if \(f(0.457013) = g\) then \(g\) is the function where \(g(1) = 4\), \(g(2) = 5\), \(g(3) = 7\), \(g(4) = 0\), \(g(5) = 1\), \(g(6) = 3\) and \(g(n) = 0\) for all \(n \geq 7\).

    Now we will prove that this function is one-to-one. Suppose that \(f(x_1) = f(x_2)\) for \(x_1, x_2 \in (0,1)\). Then for these functions to be equal, they map every integer to the same value. For this to happen, every decimal digit of \(x_1\) must be equal to the same decimal digit of \(x_2\) and so we must have \(x_1 = x_2\).

    Now denote the set of all outputs of \(f\) as \(S\). Hence \(S \subseteq T\). Now if we set the codomain of \(f\) to be \(S\), which is the range of \(f\), we have made this function surjective and hence a bijection. This means that the subset \(S \subseteq T\) is in bijection with the interval \((0,1)\). Since the interval \((0,1)\) is uncountable, we must have that \(S\) is also uncountable and since \(S \subseteq T\), \(T\) must also be uncountable.

    1. Part (a) showed us that the set of all computer programs in a given language is countable. This means that there are a countable number of computable functions. However part (b) showed us that there is an uncountable number of functions in the set \(T\), so we must have that some functions are uncomputable.

§29 — Relations on Sets

Epp 4th ed. pp. 442–449 · 5th ed. (metric) pp. 487–494. Lecture 23 (week 8).

Problems

  1. Define a relation \(P\) on \(\mathbb{Z}\) as follows: for every ordered pair \((m,n) \in \mathbb{Z} \times \mathbb{Z}\),

    \[m \mathrel{P} n \iff m \text{ and } n \text{ have a common prime factor.}\]

    1. Is \(15 \mathrel{P} 25\)?

    2. Is \(22 \mathrel{P} 27\)?

    3. Is \(0 \mathrel{P} 5\)?

    4. Is \(8 \mathrel{P} 8\)?

  2. Let \(X = \{a,b,c\}\). Define a relation \(S\) on the power set \(\mathcal{P}(X)\) as follows: for all sets \(A\) and \(B\) in \(\mathcal{P}(X)\),

    \[A \mathrel{S} B \iff A \text{ has the same number of elements as } B.\]

    1. Is \(\{a,b\} \mathrel{S} \{b,c\}\)?

    2. Is \(\{a\} \mathrel{S} \{a,b\}\)?

    3. Is \(\{c\} \mathrel{S} \{b\}\)?

  3. Let \(A = \{3,4,5\}\) and \(B = \{4,5,6\}\) be sets. Let \(R\) be the ‘divides’ relation. That is, for every ordered pair \((a,b)\) in \(A \times B\), we have \(a \mathrel{R} b\) if and only if \(a \mid b\). Draw an arrow diagram (i.e. the directed graph) of this relation.

Solutions

    1. Yes, since \(15 = 3 \cdot 5\) and \(25 = 5 \cdot 5\) share the prime factor 5.

    2. No, since \(22 = 2 \cdot 11\) and \(27 = 3^3\) share no common prime factor.

    3. Yes, since 0 is divisible by every prime; in particular \(5 \mid 0\) and \(5 \mid 5\), so 0 and 5 have a common prime factor.

    4. Yes, since \(8 = 2^3\) and 2 is a prime factor of 8 (so they share the prime factor 2).

    1. Yes, since \(|\{a,b\}| = 2 = |\{b,c\}|\).

    2. No, since \(|\{a\}| = 1 \neq 2 = |\{a,b\}|\).

    3. Yes, since \(|\{c\}| = 1 = |\{b\}|\).

  1. We have \(A \times B = \{(3,4), (3,5), (3,6), (4,4), (4,5), (4,6), (5,4), (5,5), (5,6)\}\). The pairs \((a,b)\) satisfying \(a \mid b\) are

    \[\{(3,6),\ (4,4),\ (5,5)\} \subseteq A \times B.\]

    So the directed graph has an arrow from 3 (in \(A\)) to 6 (in \(B\)), an arrow from 4 to 4, and an arrow from 5 to 5 — and no other arrows.

§30 — Reflexivity, Symmetry, Transitivity

Epp 4th ed. pp. 449–459 · 5th ed. (metric) pp. 495–505. Lecture 23 (week 8).

Problems

  1. Let \(X = \{1,2,3,4,5\}\) and consider the subset \(R \subseteq X \times X\) given by

    \[R = \{(1,1), (2,2), (3,3), (4,4), (5,5), (1,2), (2,3), (3,4), (4,5), (5,1)\}.\]

    1. Draw an arrow diagram for the binary relation \(R\).

    2. Draw an arrow diagram for the binary relation \(R^{-1}\).

    3. Draw an arrow diagram for the binary relation \(R \cup R^{-1}\).

    4. Is \(R\) symmetric? Is \(R^{-1}\) symmetric? Is \(R \cup R^{-1}\) symmetric?

    5. Is \(R\) transitive? Is \(R^{-1}\) transitive? Is \(R \cup R^{-1}\) transitive?

  2. Suppose \(R\) is a binary relation on the set \(\{1,2,3\}\) such that:

    • \((1,2) \in R\)
    • \((2,3) \in R\)
    • \(R\) is reflexive
    • \(R\) is transitive

    What is the minimum number of arrows required in an arrow diagram for a relation \(R\) that satisfies these criteria?

Solutions

    1. The arrow diagram for \(R\) has the five vertices \(1, 2, 3, 4, 5\) drawn around a circle, with a loop at every vertex (from the pairs \((1,1), \dots, (5,5)\)) together with the directed 5-cycle

    \[1 \to 2 \to 3 \to 4 \to 5 \to 1.\]

    1. The arrow diagram for \(R^{-1}\) is the same picture with every non-loop arrow reversed: a loop at every vertex, together with the directed 5-cycle

    \[1 \to 5 \to 4 \to 3 \to 2 \to 1,\]

    that is, \(R^{-1} = \{(1,1), (2,2), (3,3), (4,4), (5,5), (2,1), (3,2), (4,3), (5,4), (1,5)\}\).

    1. The arrow diagram for \(R \cup R^{-1}\) has a loop at every vertex and a double-headed arrow between each consecutive pair around the cycle — that is, both \((i, i+1)\) and \((i+1, i)\) for the cycle \(1,2,3,4,5,1\).

    2. \(R\) is not symmetric because \((1,2) \in R\) but \((2,1) \notin R\).

    \(R^{-1}\) is not symmetric because \((2,1) \in R^{-1}\) but \((1,2) \notin R^{-1}\).

    \(R \cup R^{-1}\) is symmetric. If \((a,b) \in R \cup R^{-1}\) then \((a,b) \in R\) or \((a,b) \in R^{-1}\), which means \((b,a) \in R^{-1}\) or \((b,a) \in R\), and thus, in either case, \((b,a) \in R \cup R^{-1}\).

    1. \(R\) is not transitive because \((1,2), (2,3) \in R\) but \((1,3) \notin R\).

    \(R^{-1}\) is not transitive because \((3,2), (2,1) \in R^{-1}\) but \((3,1) \notin R^{-1}\).

    \(R \cup R^{-1}\) is not transitive because \((1,2), (2,3) \in R \cup R^{-1}\) but \((1,3) \notin R \cup R^{-1}\).

  1. \(R\) is reflexive, so \((1,1), (2,2), (3,3) \in R\) — a loop on each point of the arrow diagram. We are also given \((1,2), (2,3) \in R\), and \(R\) is transitive, so \((1,3) \in R\). \(R\) is not required to be symmetric, so no further arrows are forced.

    \[R = \{(1,1), (2,2), (3,3), (1,2), (2,3), (1,3)\}\]

    So the minimum number of arrows required is 6.

§31 — Equivalence Relations

Epp 4th ed. pp. 459–478 · 5th ed. (metric) pp. 505–523. Lecture 24 (week 9).

Problems

  1. Define the binary relations \(\alpha\), \(\beta\) and \(\gamma\) on \(\mathbb{Z}\) as follows:

    \[\begin{aligned} x \mathrel{\alpha} y &\iff x + y \text{ is an even integer}, \\ x \mathrel{\beta} y &\iff x \text{ is a multiple of } y, \\ x \mathrel{\gamma} y &\iff 3 \text{ is a factor of } x^2 - y^2. \end{aligned}\]

    1. Fill in the table with ‘Yes’/‘No’ to indicate the properties satisfied by each relation.
    \(\alpha\) \(\beta\) \(\gamma\)
    Reflexive
    Symmetric
    Transitive
    Equivalence relation
    1. For each equivalence relation, state its equivalence classes.
  2. Let \(A = \{a,b,c,d,e\}\) be a set and \(\rho\) be a binary relation defined by

    \[\rho = \{(a,a), (a,b), (a,d), (b,b), (c,c), (d,d), (b,a), (d,a), (b,d), (d,b), (c,e), (e,c), (e,e)\}.\]

    For instance, \((a,b) \in \rho\) means \(a \mathrel{\rho} b\).

    1. Draw the directed graph representing \(\rho\).

    2. Is \(\rho\) reflexive? Is \(\rho\) symmetric? Is \(\rho\) transitive?

    3. Is \(\rho\) an equivalence relation on \(A\)? If so, give its equivalence classes.

    4. What is the smallest subset \(\tau\) of \(\rho\) which forms an equivalence relation on \(A\)?

    5. Give the equivalence classes for \(\tau\).

  3. Consider a positive integer \(n\) and any set of strings \(S\). Define a relation \(R_n\) on \(S\) by the condition \(s \mathrel{R_n} t\) if and only if \(s = t\) or both \(s\) and \(t\) have at least \(n\) characters and the first \(n\) characters are the same. Show that \(R_n\) is an equivalence relation.

Solutions

    \(\alpha\) \(\beta\) \(\gamma\)
    Reflexive Yes Yes Yes
    Symmetric Yes No Yes
    Transitive Yes Yes Yes
    Equivalence relation Yes No Yes
    1. The equivalence relations are \(\alpha\) and \(\gamma\).

    For \(\alpha\), notice \(x \mathrel{\alpha} y \iff x + y\) is even \(\iff x\) and \(y\) have the same parity. This means the equivalence classes are \(\mathbb{Z}^{\text{even}}\) and \(\mathbb{Z}^{\text{odd}}\).

    For \(\gamma\), notice \(x \mathrel{\gamma} y \iff 3 \mid (x^2 - y^2) \iff x^2 \equiv y^2 \pmod 3\). But modulo 3 the only possible square residues are

    \[0^2 \equiv 0, \qquad 1^2 \equiv 1, \qquad 2^2 \equiv 1 \pmod 3.\]

    Hence \(x^2 \equiv 0 \pmod 3\) exactly when \(3 \mid x\), and \(x^2 \equiv 1 \pmod 3\) exactly when \(3 \nmid x\). Therefore \(x \mathrel{\gamma} y\) iff either both \(x\) and \(y\) are multiples of 3, or neither is. So the equivalence classes are \(\{n \in \mathbb{Z} : 3 \mid n\}\) and \(\{n \in \mathbb{Z} : 3 \nmid n\}\). Note that another way to write the equivalence classes is \(\{3n \mid n \in \mathbb{Z}\}\) and \(\{3n+1,\ 3n+2 \mid n \in \mathbb{Z}\}\).

    1. A digraph for \(\rho\) (vertex set \(A = \{a,b,c,d,e\}\)) has a loop at every vertex, double-headed arrows between each pair of \(\{a,b,d\}\) — that is between \(a\) and \(b\), between \(a\) and \(d\), and between \(b\) and \(d\) — and a double-headed arrow between \(c\) and \(e\). There are no arrows between \(\{a,b,d\}\) and \(\{c,e\}\).

    2. \(\rho\) is reflexive: we have \((x,x) \in \rho\) for all \(x \in A\) (loops on all vertices).

    \(\rho\) is symmetric: whenever \((x,y) \in \rho\), we have \((y,x) \in \rho\) (no single-direction arrow between any two distinct vertices).

    \(\rho\) is transitive: within \(\{a,b,d\}\) all relations occur, within \(\{c,e\}\) all relations occur, and there are no relations between \(\{a,b,d\}\) and \(\{c,e\}\). That is, whenever \((x,y)\) and \((y,z)\) are in \(\rho\), then \((x,z)\) is also in \(\rho\) (and \(x = z\) is possible).

    1. Yes, \(\rho\) is an equivalence relation. Its equivalence classes are

    \[[a] = [b] = [d] = \{a,b,d\}, \qquad [c] = [e] = \{c,e\}.\]

    1. An equivalence relation \(\tau \subseteq \rho\) must be reflexive, so \((x,x) \in \tau\) for all \(x \in A\). But notice if \(\tau\) is the relation

    \[\tau = \{(a,a), (b,b), (c,c), (d,d), (e,e)\}\]

    then \(\tau\) is an equivalence relation. Therefore this is the smallest relation which is an equivalence relation.

    1. The equivalence classes for \(\tau\) are the singletons:

    \[[a] = \{a\},\ [b] = \{b\},\ [c] = \{c\},\ [d] = \{d\},\ [e] = \{e\}.\]

  1. Let \(S\) be any set of strings and fix \(n \in \mathbb{Z}^+\).

    \(R_n\) is reflexive: for any \(s \in S\), since \(s = s\), by definition \(s \mathrel{R_n} s\).

    \(R_n\) is symmetric: let \(s, t \in S\), and assume \(s \mathrel{R_n} t\). If \(s = t\) then \(t \mathrel{R_n} s\); otherwise both \(s\) and \(t\) must have at least \(n\) characters and, since \(s \mathrel{R_n} t\), they must have the same first \(n\) characters, so \(t \mathrel{R_n} s\).

    \(R_n\) is transitive: let \(s, t, u \in S\) and assume \(s \mathrel{R_n} t\) and \(t \mathrel{R_n} u\). Since \(s \mathrel{R_n} t\) we either have \(s = t\), or \(s\) and \(t\) both have at least \(n\) characters and have the same first \(n\) characters. Similarly, since \(t \mathrel{R_n} u\), either \(t = u\), or \(t\) and \(u\) both have at least \(n\) characters and have the same first \(n\) characters. Thus either \(s = t = u\), or \(s\) and \(u\) both have at least \(n\) characters and they both share the first \(n\) characters with \(t\) (and hence each other). Thus \(s \mathrel{R_n} u\) and so \(R_n\) is transitive.

    Since \(R_n\) is reflexive, symmetric and transitive it is an equivalence relation. \(\square\)

Application — Identifiers

Not assessable for MATH1061. Taken from K. Rosen, Discrete Mathematics and its Applications, 7th edition, 1991, pages 608–618.

In the C programming language, an identifier is the name of a variable, a function or another type of entity.

Each identifier is a nonempty string of characters where each character is a lowercase or uppercase English letter, a digit, or an underscore.

There is no limit on the number of characters used in the identifier; however, for some compilers for some versions of C, there is a limit on the number of characters checked when two names are compared.

For example, Standard C compilers consider two identifiers the same when they agree in their first 31 characters. We see that two identifiers will be considered the same if and only if they are related under the relation \(R_n\) when \(n = 31\) (defined in a previous question).

Application problems
  1. Using \(R_{31}\) on the set of all possible identifiers, determine the equivalence classes of each of the following identifiers:

    1. file_1

    2. Number_of_tropical_storms

    3. Number_of_named_tropical_storms

    4. Number_of_named_tropical_storms_in_the_Atlantic_in_2005

Application solutions
  1. Let \(I\) be the set of all possible identifiers (strings over letters, digits and underscores).

    1. Since file_1 has length \(6 < 31\), its equivalence class is just

    \[[\texttt{file\_1}]_{R_{31}} = \{\texttt{file\_1}\}.\]

    1. Since Number_of_tropical_storms has length \(25 < 31\), its equivalence class is

    \[[\texttt{Number\_of\_tropical\_storms}]_{R_{31}} = \{\texttt{Number\_of\_tropical\_storms}\}.\]

    1. The identifier Number_of_named_tropical_storms has length exactly 31, so two identifiers are \(R_{31}\)-equivalent to it iff they have at least 31 characters and their first 31 characters equal this string. Hence

    \[[\texttt{Number\_of\_named\_tropical\_storms}]_{R_{31}} = \left\{ t \in I : \begin{array}{l} |t| \geq 31 \text{ and the first 31 characters of} \\ t \text{ are } \texttt{Number\_of\_named\_tropical\_storms} \end{array} \right\}.\]

    1. The identifier Number_of_named_tropical_storms_in_the_Atlantic_in_2005 begins with the same first 31 characters Number_of_named_tropical_storms, so it has the same \(R_{31}\)-equivalence class as in (c):

    \[[\texttt{Number\_of\_named\_tropical\_storms\_in\_the\_Atlantic\_in\_2005}]_{R_{31}} = \left\{ t \in I : \begin{array}{l} |t| \geq 31 \text{ and the first 31 characters of} \\ t \text{ are } \texttt{Number\_of\_named\_tropical\_storms} \end{array} \right\}.\]

§32 — Partial Order Relations

Epp 4th ed. pp. 498–515 · 5th ed. (metric) pp. 546–563. Lecture 25 (week 9).

Problems

  1. Let \(\rho\) be a relation on \(\mathbb{Z}\) defined by \(x \mathrel{\rho} y\) if and only if \(x^3 - 1 < y^3\).

    1. Prove \(\rho\) is a partial order relation.

    2. Is \(\rho\) a total order?

  2. Let \(S\) be the set of binary strings of length 6. Define the weight of a binary string as the number of ones it contains. (For example, \(\operatorname{weight}(110110) = 4\).) Define a relation \(\rho\) on \(S\) by

    \[s \mathrel{\rho} t \text{ if and only if } \operatorname{weight}(s) \leq \operatorname{weight}(t).\]

    1. Prove that \(\rho\) is not an equivalence relation.

    2. Prove that \(\rho\) is not a partial order relation.

  3. Suppose \(R\) is a binary relation on the set \(\{1,2,3\}\) such that:

    • \((2,3) \in R\)
    • \((3,2) \in R\)
    • \((1,1) \notin R\)
    • \(R\) is transitive
    • \(R\) is not symmetric
    • \(R\) is not antisymmetric.

    Draw an arrow diagram for the relation \(R\).

Solutions

    1. To show that \(\rho\) is a partial order, we need to show that \(\rho\) is reflexive, antisymmetric and transitive.

    Reflexive: let \(x \in \mathbb{Z}\). Then \(x^3 - 1 < x^3\) and so \(x \mathrel{\rho} x\), so \(\rho\) is reflexive.

    Antisymmetric: suppose \(x \mathrel{\rho} y\) where \(x \neq y\). (We must show that then \(y \mathrel{\not\rho} x\).) Now \(x \mathrel{\rho} y\) means that \(x^3 - 1 < y^3\). And since \(x \neq y\) we must have \(x^3 < y^3\) and so \(x^3 \leqslant y^3 - 1\). Hence \(y^3 - 1 \not< x^3\) and so we do not have \(y \mathrel{\rho} x\). Therefore \(\rho\) is antisymmetric.

    Transitive: suppose that \(x \mathrel{\rho} y\) and \(y \mathrel{\rho} z\). Then we know that \(x^3 - 1 < y^3\) and \(y^3 - 1 < z^3\). So \(x^3 - 1 \leqslant y^3 - 1 < z^3\), that is, \(x^3 - 1 < z^3\) so that \(x \mathrel{\rho} z\), which means that \(\rho\) is transitive.

    Hence \(\rho\) is a partial order. \(\square\)

    1. To check whether \(\rho\) is a total order, let \(x\) and \(y\) be any two distinct integers. Then \(x^3 \neq y^3\) and so either \(x^3 < y^3\) or else \(y^3 < x^3\).

    If \(x^3 < y^3\) then \(x^3 - 1 < y^3\) and so \(x \mathrel{\rho} y\).

    If \(y^3 < x^3\) then \(y^3 - 1 < x^3\) and so \(y \mathrel{\rho} x\).

    So for any integers \(x \neq y\), we have either \(x \mathrel{\rho} y\) or else \(y \mathrel{\rho} x\). So any two integers are comparable under \(\rho\). Hence \(\rho\) is a total order.

    1. It is not an equivalence relation because it is not symmetric. For instance, let \(s = 000000\) and \(t = 111111\). Then \(\operatorname{weight}(s) = 0 \leq 6 = \operatorname{weight}(t)\), so \((s,t) \in \rho\), but \((t,s) \notin \rho\).

    2. It is not a partial order because it is not antisymmetric. For instance, let \(s = 100000\) and \(t = 010000\). Then \(\operatorname{weight}(s) = \operatorname{weight}(t) = 1\), so \(s \mathrel{\rho} t\) and \(t \mathrel{\rho} s\), but \(s \neq t\).

  1. One relation on \(\{1,2,3\}\) satisfying all conditions is

    \[R = \{(1,2), (1,3), (2,2), (3,3), (2,3), (3,2)\}.\]

    (It contains \((2,3)\) and \((3,2)\); it is transitive; it is not symmetric since \((1,2) \in R\) but \((2,1) \notin R\); it is not antisymmetric since \(2 \neq 3\) and \((2,3), (3,2) \in R\); and \((1,1) \notin R\).)

    An arrow diagram for this \(R\) has vertices \(1, 2, 3\) with: arrows \(1 \to 2\) and \(1 \to 3\); loops at 2 and at 3; and a double-headed arrow between 2 and 3. There is no loop at vertex 1.

§33 — Definitions and Examples of Groups

No relevant textbook pages. Lecture 26 (week 9).

Problems

  1. Show \(\mathbb{Z}_9 - \{0\} = \{1,2,3,\dots,7,8\}\) is not a group under multiplication modulo 9.

  2. Let \(G\) be the set of all odd integers. Define the binary operation \(\star\) on \(G\) by

    \[x \star y = x + y - 1.\]

    Show that \((G, \star)\) is an abelian group.

  3. Show \(\mathbb{Z}_9 - \{0,3,6\} = \{1,2,4,5,7,8\}\) is a group under multiplication modulo 9. Is the group abelian?

  4. Define the group \(H = \mathbb{Z}_5 - \{0\} = \{1,2,3,4\}\) under multiplication modulo 5. Write the Cayley table for \(H\).

  5. Define the group \(G = \mathbb{Z}_2 \times \mathbb{Z}_2\) under component-wise addition modulo 2. That is,

    \[(a,b) + (c,d) = (a + c \bmod 2,\ b + d \bmod 2).\]

    Write the Cayley table for \(G\).

Solutions

  1. It is not a group under multiplication mod 9 because, for example, it is not closed (\(3 \cdot 3 = 0 \notin \mathbb{Z}_9 - \{0\}\)).

  2. Closure: if \(x\) and \(y\) are odd, then \(x + y\) is even, so \(x + y - 1\) is odd. Hence \(x \star y \in G\).

    Associativity: for \(x, y, z \in G\),

    \[(x \star y) \star z = (x + y - 1) + z - 1 = x + y + z - 2 = x + (y + z - 1) - 1 = x \star (y \star z).\]

    Commutativity: for \(x, y \in G\),

    \[x \star y = x + y - 1 = y + x - 1 = y \star x.\]

    Identity: we seek \(e \in G\) such that \(x \star e = x = e \star x\) for all \(x\). Now the requirement \(x \star e = x + e - 1 = x\) implies that \(e = 1\). Note 1 lies in \(G\). Moreover, commutativity implies that \(e \star x = x\) as well. Thus the identity is 1.

    Inverses: given \(x \in G\), we want \(y \in G\) with \(x \star y = 1\). This means \(x + y - 1 = 1\), so \(y = 2 - x\). Since \(x\) is odd, \(2 - x\) is odd, hence \(2 - x \in G\). Thus the inverse of \(x\) is \(x^{-1} = 2 - x\).

    Therefore \((G, \star)\) is an abelian group. \(\square\)

  3. Closure: if \(a, b \in \{1,2,4,5,7,8\}\), then \(3 \nmid a\) and \(3 \nmid b\), hence \(3 \nmid ab\), so \(ab \not\equiv 0, 3, 6 \pmod 9\); thus \(\{1,2,4,5,7,8\}\) is closed under multiplication modulo 9.

    Associativity and commutativity: normal multiplication between integers is associative and commutative, so multiplication modulo 9 in the set \(\{1,2,4,5,7,8\}\) is also associative and commutative. This means the group is abelian.

    Identity: \(1 \in G\) is the identity because \(1 \cdot x \equiv x \pmod 9\).

    Inverse: every element has an inverse. We explicitly calculate the inverses:

    \[\begin{aligned} 1^{-1} &= 1 && \text{because } 1 \cdot 1 \equiv 1 \pmod 9, \\ 2^{-1} &= 5 && \text{because } 2 \cdot 5 = 10 \equiv 1 \pmod 9, \\ 4^{-1} &= 7 && \text{because } 4 \cdot 7 = 28 \equiv 1 \pmod 9, \\ 5^{-1} &= 2 && \text{because } 5 \cdot 2 = 10 \equiv 1 \pmod 9, \\ 7^{-1} &= 4 && \text{because } 7 \cdot 4 = 28 \equiv 1 \pmod 9, \\ 8^{-1} &= 8 && \text{because } 8 \cdot 8 = 64 \equiv 1 \pmod 9. \end{aligned}\]

  4. For \(H = \mathbb{Z}_5 \setminus \{0\} = \{1,2,3,4\}\) under multiplication mod 5, the Cayley table is:

    \(\cdot\) 1 2 3 4
    1 1 2 3 4
    2 2 4 1 3
    3 3 1 4 2
    4 4 3 2 1
  5. For \(G = \mathbb{Z}_2 \times \mathbb{Z}_2 = \{(0,0), (0,1), (1,0), (1,1)\}\) under componentwise addition mod 2, the Cayley table is:

    \(+\) \((0,0)\) \((0,1)\) \((1,0)\) \((1,1)\)
    \(\mathbf{(0,0)}\) \((0,0)\) \((0,1)\) \((1,0)\) \((1,1)\)
    \(\mathbf{(0,1)}\) \((0,1)\) \((0,0)\) \((1,1)\) \((1,0)\)
    \(\mathbf{(1,0)}\) \((1,0)\) \((1,1)\) \((0,0)\) \((0,1)\)
    \(\mathbf{(1,1)}\) \((1,1)\) \((1,0)\) \((0,1)\) \((0,0)\)

§34 — Elementary Properties of Groups

No relevant textbook pages. Lecture 27 (week 10).

Problems

  1. Let \(H = \{3n \mid n \in \mathbb{Z}\}\) be the set of multiples of 3. Prove \(H\) is a subgroup of \((\mathbb{Z}, +)\).

  2. Fix elements \(g\) and \(h\) in a group \(G\) where \(e\) is the group identity. Prove the following statements:

    1. If \(g^{-1} = h\) then \(h^{-1} = g\).

    2. \((g^{-1})^{-1} = g\).

    3. \((gh)^{-1} = h^{-1}g^{-1}\).

  3. Prove \(K = \{(0,0), (0,1), (2,0), (2,1)\}\) is a subgroup of \((\mathbb{Z}_4 \times \mathbb{Z}_2, +)\) where the group operation is componentwise addition:

    \[(a,b) + (c,d) = (a + c \pmod 4,\ b + d \pmod 2).\]

  4. Consider the group \((\mathbb{Z}_{11} - \{0\}, \cdot)\). Show that this group is cyclic.

    Hint: express every element of \(\mathbb{Z}_{11} - \{0\}\) as a power of 2.

Solutions

  1. First we note that \(H \subseteq \mathbb{Z}\), and \(H \neq \emptyset\).

    Closure: if \(x, y \in H\) then \(x = 3a\) and \(y = 3b\) for some \(a, b \in \mathbb{Z}\). Therefore \(x + y = 3a + 3b = 3(a+b) \in H\).

    Identity: the identity of \((\mathbb{Z},+)\) is 0, and since \(0 = 3 \cdot 0\), we have \(0 \in H\).

    Inverse: in \(\mathbb{Z}\), the inverse of \(x\) is \(-x\). In \(H\), if \(x \in H\) then \(x = 3a\) for some \(a \in \mathbb{Z}\). This means \(-x = -3a = 3 \cdot (-a)\), and so \(-x \in H\) also.

    Note that we do not need to check associativity for subgroups, since this property is automatically inherited from the parent group \((\mathbb{Z},+)\). \(\square\)

    1. Suppose \(g^{-1} = h\). Now \(gh = gg^{-1} = e\) and \(hg = g^{-1}g = e\). Thus \(h^{-1} = g\).

    2. By definition, \(gg^{-1} = e = g^{-1}g\), so \((g^{-1})^{-1} = g\).

    3. Notice we have

    \[(gh)(h^{-1}g^{-1}) = ghh^{-1}g^{-1} = e\]

    and

    \[(h^{-1}g^{-1})(gh) = h^{-1}g^{-1}gh = e.\]

    This means \((gh)^{-1} = h^{-1}g^{-1}\).

  2. We draw the Cayley table of \(K = \{(0,0), (0,1), (2,0), (2,1)\}\) under componentwise addition in \(\mathbb{Z}_4 \times \mathbb{Z}_2\):

    \(+\) \((0,0)\) \((0,1)\) \((2,0)\) \((2,1)\)
    \(\mathbf{(0,0)}\) \((0,0)\) \((0,1)\) \((2,0)\) \((2,1)\)
    \(\mathbf{(0,1)}\) \((0,1)\) \((0,0)\) \((2,1)\) \((2,0)\)
    \(\mathbf{(2,0)}\) \((2,0)\) \((2,1)\) \((0,0)\) \((0,1)\)
    \(\mathbf{(2,1)}\) \((2,1)\) \((2,0)\) \((0,1)\) \((0,0)\)

    Notice the operation is closed in \(K\) and every element of \(K\) has an inverse in \(K\). Therefore \(K\) is a subgroup. \(\square\)

  3. Note that 2 is a generator of \((\mathbb{Z}_{11} - \{0\}, \cdot)\), so it is cyclic:

    \[1 = 2^0,\ 2 = 2^1,\ 4 = 2^2,\ 8 = 2^3,\ 5 = 2^4,\ 10 = 2^5,\ 9 = 2^6,\ 7 = 2^7,\ 3 = 2^8,\ 6 = 2^9.\]

§35 — Group Isomorphisms

No relevant textbook pages. No lecture in the published schedule — the Week 10 row jumps from L27 (V34) to L28 (V36).

This content is optional and will not be assessed. The source document carries no practice problems or solutions for this section.

§36 — Definitions and Examples of Fields

No relevant textbook pages. Lecture 28 (week 10).

Problems

    1. Solve \(3x + 4 = 2\) in the field \((\mathbb{Z}_5, +, \cdot)\).

    2. Solve \(8x - 9 = 6\) in the field \((\mathbb{Z}_{11}, +, \cdot)\).

    3. Solve \(4(1-x) = 10\) in the field \((\mathbb{Z}_{13}, +, \cdot)\).

    4. Solve \(16(x-6) = 15\) in the field \((\mathbb{Z}_{19}, +, \cdot)\).

  1. Prove \(\mathbb{Z}_{12}\) with addition modulo 12 and multiplication modulo 12 is not a field.

  2. Define addition and multiplication on \(\mathbb{Q} \times \mathbb{Q}\) by

    \[(a,b) + (c,d) = (a+c,\ b+d)\]

    and

    \[(a,b) \cdot (c,d) = (ac + 3bd,\ ad + bc)\]

    for all \((a,b), (c,d) \in \mathbb{Q} \times \mathbb{Q}\). Prove that \((\mathbb{Q} \times \mathbb{Q}, +, \cdot)\) is a field.

Solutions

    \[\begin{aligned} 3x + 4 &= 2 \\ \to 3x &= -2 \end{aligned}\]

    Now, since \(3 \times 2 = 6 \equiv 1 \pmod 5\), 2 is the multiplicative inverse of 3, and so we multiply both sides by 2 and obtain \(x = -4\). That is, \(x = 1\).

    Check: \(3 \cdot 1 + 4 = 7 \equiv 2 \pmod 5\).

    \[\begin{aligned} 8x - 9 &= 6 \\ \to 8x &= 15 = 4 \end{aligned}\]

    Now, since \(8 \times 7 = 56 \equiv 1 \pmod{11}\), 7 is the multiplicative inverse of 8, and so we multiply both sides by 7 and obtain \(x = 28\). That is, \(x = 6\).

    Check: \(8 \cdot 6 - 9 = 39 \equiv 6 \pmod{11}\).

    1. Since \(4 \times 10 = 40 \equiv 1 \pmod{13}\), 10 is the multiplicative inverse of 4, and so we multiply both sides by 10 and obtain \(1 - x = 100 = 9\). That is, \(x = 1 - 9 = 14 - 9 = 5\).

    Check: \(4(1-5) = -16 \equiv 10 \pmod{13}\).

    1. Since \(16 \times 6 = 96 \equiv 1 \pmod{19}\), 6 is the multiplicative inverse of 16, and so we multiply both sides by 6 and obtain \(x - 6 = 90 = 14\). That is, \(x = 20 = 1\).

    Check: \(16(1-6) = -80 \equiv 15 \pmod{19}\).

  1. \(\mathbb{Z}_{12}\) is not a field because 2 has no multiplicative inverse modulo 12. Indeed \(2 \cdot k \equiv 1 \pmod{12}\) has no solution since \(2k\) is always even.

  2. We show that \((\mathbb{Q} \times \mathbb{Q}, +, \cdot)\) is a field in three steps.

    Step 1 — \((\mathbb{Q} \times \mathbb{Q}, +)\) is an abelian group.

    Closure: for all \((a,b), (c,d) \in \mathbb{Q} \times \mathbb{Q}\), we have \((a,b) + (c,d) = (a+c, b+d) \in \mathbb{Q} \times \mathbb{Q}\).

    Associative: for all \((a,b), (c,d), (e,f) \in \mathbb{Q} \times \mathbb{Q}\) we have \(\big((a,b) + (c,d)\big) + (e,f) = (a+c, b+d) + (e,f) = (a+c+e, b+d+f) = (a,b) + (c+e, d+f) = (a,b) + \big((c,d) + (e,f)\big)\).

    Identity: for all \((a,b) \in \mathbb{Q} \times \mathbb{Q}\) we have \((a,b) + (0,0) = (a+0, b+0) = (a,b) = (0+a, 0+b) = (0,0) + (a,b)\).

    Inverses: for all \((a,b) \in \mathbb{Q} \times \mathbb{Q}\) we have \((a,b) + (-a,-b) = (a-a, b-b) = (0,0)\) and \((-a,-b) + (a,b) = (-a+a, -b+b) = (0,0)\). So \((-a,-b)\) is the inverse of \((a,b)\).

    Commutativity: for all \((a,b), (c,d) \in \mathbb{Q} \times \mathbb{Q}\), we have \((a,b) + (c,d) = (a+c, b+d) = (c+a, d+b) = (c,d) + (a,b)\).

    Thus \((\mathbb{Q} \times \mathbb{Q}, +)\) is an abelian group.

    Step 2 — \(\big((\mathbb{Q} \times \mathbb{Q}) - \{(0,0)\},\ \cdot\big)\) is an abelian group.

    Closure: for all \((a,b), (c,d) \in (\mathbb{Q} \times \mathbb{Q}) - \{(0,0)\}\), we have \((a,b) \cdot (c,d) = (ac+3bd, ad+bc) \in \mathbb{Q} \times \mathbb{Q}\). So we only need to show that \((ac+3bd, ad+bc) \neq (0,0)\). For a contradiction, suppose \((ac+3bd, ad+bc) = (0,0)\). That is, \(ac + 3bd = 0\) and \(ad + bc = 0\). Multiplying the first equation by \(d\) and the second by \(c\) we obtain \(acd + 3bd^2 = 0\) and \(acd + bc^2 = 0\), from which it follows that \(3bd^2 = bc^2\).

    Now, if \(b = 0\), then \(ac = 0\), \(ad = 0\) and \(a \neq 0\), and so \(c = d = 0\); a contradiction. Thus \(b \neq 0\) and we have \(3d^2 = c^2\). This is a contradiction because if \(d = 0\), then \(c \neq 0\) and the equation does not hold, and if \(d \neq 0\), then we have \(3 = \frac{c^2}{d^2}\), which implies \(\sqrt{3} = \frac{c}{d}\) is rational. Thus \((ac+3bd, ad+bc) \neq (0,0)\) and we have closure.

    Associative: for all \((a,b), (c,d), (e,f) \in (\mathbb{Q} \times \mathbb{Q}) - \{(0,0)\}\) we have

    \[\begin{aligned} (a,b) \cdot \big((c,d) \cdot (e,f)\big) &= (a,b) \cdot (ce + 3df,\ cf + de) \\ &= \big(a(ce+3df) + 3b(cf+de),\ a(cf+de) + b(ce+3df)\big) \\ &= (ace + 3adf + 3bcf + 3bde,\ acf + ade + bce + 3bdf) \end{aligned}\]

    and

    \[\begin{aligned} \big((a,b) \cdot (c,d)\big) \cdot (e,f) &= (ac+3bd,\ ad+bc) \cdot (e,f) \\ &= \big((ac+3bd)e + 3(ad+bc)f,\ (ac+3bd)f + (ad+bc)e\big) \\ &= (ace + 3bde + 3adf + 3bcf,\ acf + 3bdf + ade + bce) \\ &= (ace + 3adf + 3bcf + 3bde,\ acf + ade + bce + 3bdf). \end{aligned}\]

    Identity: for all \((a,b) \in (\mathbb{Q} \times \mathbb{Q}) - \{(0,0)\}\) we have \((a,b) \cdot (1,0) = (a,b)\) and \((1,0) \cdot (a,b) = (a,b)\), so \((1,0)\) is a multiplicative identity.

    Commutativity: for all \((a,b), (c,d) \in \mathbb{Q} \times \mathbb{Q}\), we have

    \[(a,b) \cdot (c,d) = (ac+3bd,\ ad+bc) = (ca+3db,\ cb+da) = (c,d) \cdot (a,b).\]

    Inverses: we check that for all \((a,b) \in (\mathbb{Q} \times \mathbb{Q}) - \{(0,0)\}\),

    \[(a,b)^{-1} = \left(\frac{a}{a^2 - 3b^2},\ \frac{-b}{a^2 - 3b^2}\right).\]

    First, note that \(a^2 - 3b^2 \neq 0\), because \(a^2 - 3b^2 = 0\) implies either \((a,b) = (0,0)\) or \(\sqrt{3} = \frac{a}{b}\), both of which are contradictions (the second because \(\sqrt{3}\) is not rational). Thus \((a,b)^{-1} \in (\mathbb{Q} \times \mathbb{Q}) - \{(0,0)\}\), and notice

    \[\begin{aligned} (a,b) \cdot \left(\frac{a}{a^2-3b^2}, \frac{-b}{a^2-3b^2}\right) &= \left(a \cdot \frac{a}{a^2-3b^2} + 3b \cdot \frac{-b}{a^2-3b^2},\ a \cdot \frac{-b}{a^2-3b^2} + b \cdot \frac{a}{a^2-3b^2}\right) \\ &= \left(\frac{a^2 - 3b^2}{a^2 - 3b^2},\ \frac{-ab + ab}{a^2 - 3b^2}\right) \\ &= (1, 0). \end{aligned}\]

    So every nonzero element has an inverse. Thus \(\big((\mathbb{Q} \times \mathbb{Q}) - \{(0,0)\}, \cdot\big)\) is an abelian group.

    Step 3 — the distributive laws. For all \((a,b), (c,d), (e,f) \in \mathbb{Q} \times \mathbb{Q}\) we have

    \[\begin{aligned} (a,b) \cdot \big((c,d) + (e,f)\big) &= (a,b) \cdot (c+e,\ d+f) \\ &= \big(a(c+e) + 3b(d+f),\ a(d+f) + b(c+e)\big) \\ &= \big((ac+3bd) + (ae+3bf),\ (ad+bc) + (af+be)\big) \\ &= (ac+3bd,\ ad+bc) + (ae+3bf,\ af+be) \\ &= (a,b) \cdot (c,d) + (a,b) \cdot (e,f) \end{aligned}\]

    and (since we have already shown that multiplication and addition are commutative)

    \[\begin{aligned} \big((a,b) + (c,d)\big) \cdot (e,f) &= (e,f) \cdot \big((a,b) + (c,d)\big) \\ &= (e,f) \cdot (a,b) + (e,f) \cdot (c,d) && \text{(by the first distributive law)} \\ &= (a,b) \cdot (e,f) + (c,d) \cdot (e,f). \end{aligned}\]

    Thus \((\mathbb{Q} \times \mathbb{Q}, +, \cdot)\) is a field. \(\square\)

§37 — Introduction to Counting

Epp 4th ed. pp. 516–553 · 5th ed. (metric) pp. 564–604. Lecture 29 (week 11).

Problems

  1. The UQ drama club is auditioning for its annual production of Romeo and Juliet. Six people are auditioning for the role of Romeo, and eight other people are auditioning for the role of Juliet. In how many ways can the director cast the lead couple?

  2. There are eleven questions on the final exam, but you decide — based on your successful grade for your in-semester exam — that you will only do five of them. How many combinations of questions are there which you could complete?

    1. How many permutations of \(\{1, 2, \dots, 15\}\) start with an odd number?

    2. How many 5-element subsets of \(\{1, 2, \dots, 15\}\) contain exactly two numbers from \(\{1,2,3,4\}\)?

Solutions

  1. There are 6 choices for Romeo and 8 choices for Juliet, so \(6 \cdot 8 = 48\) ways.

  2. Since the order in which you do the 5 questions does not matter, there are \(\binom{11}{5} = 462\) combinations of questions.

    1. There are 8 odd numbers in the set \(\{1, \dots, 15\}\). Choose the first entry in 8 ways, then permute the remaining 14 numbers, so \(8 \cdot 14!\) permutations.

    2. We choose exactly 2 numbers from the set \(\{1,2,3,4\}\) and exactly 3 numbers from the remaining 11 numbers, so the number of subsets is

    \[\binom{4}{2}\binom{11}{3}.\]

Application — Runtime

Not assessable for MATH1061. Taken from S. Epp, Discrete Mathematics with Applications, 5th edition, pages 787–799.

In many algorithms, the runtime is well approximated by counting how many times the basic instructions are executed.

For programs with nested for-loops and no early exits, the total number of executions of the inner loop body equals the number of choices of the outer-loop index multiplied by the number of choices of the inner-loop index.

Concretely, if \(i\) runs through \(M\) values and, for each fixed \(i\), the variable \(j\) runs through \(N\) values, then the inner loop body is executed exactly \(MN\) times.

Application problems
  1. The following algorithms are written in pseudo-code and contain nested loops. For each algorithm, count the total number of iterations of the inner loop that the algorithm will run.

    for i := 1 to 30
        for j := 1 to 15
            [Statements in body of inner loop. None contain
             branching statements that lead outside the loop.]
        next j
    next i
    for i := 5 to 50
        for j := 10 to 20
            [Statements in body of inner loop. None contain
             branching statements that lead outside the loop.]
        next j
    next i
    1. (Fix integers \(a, b, c, d \in \mathbb{Z}^+\) with \(a \leq b\) and \(c \leq d\).)
    for i := a to b
        for j := c to d
            [Statements in body of inner loop. None contain
             branching statements that lead outside the loop.]
        next j
    next i
Application solutions
    1. \(i = 1, \dots, 30\) gives 30 iterations and \(j = 1, \dots, 15\) gives 15 iterations, so the inner loop runs \(30 \cdot 15 = 450\) times.

    2. \(i = 5, \dots, 50\) gives \(50 - 5 + 1 = 46\) iterations and \(j = 10, \dots, 20\) gives \(20 - 10 + 1 = 11\) iterations, so the inner loop runs \(46 \cdot 11 = 506\) times.

    3. \(i = a, \dots, b\) gives \(b - a + 1\) iterations and \(j = c, \dots, d\) gives \(d - c + 1\) iterations, so the inner loop runs \((b-a+1)(d-c+1)\) times.

§38 — Counting Selections

Epp 4th ed. pp. 516–553 · 5th ed. (metric) pp. 564–604. Lecture 30 (week 11).

Problems

    1. How many distinct arrangements can be made from the letters of the word INDOOROOPILLY?

    2. How many distinct arrangements can be made from the letters of the word INDOOROOPILLY that begin with R and end with P?

    3. How many distinct arrangements can be made from the letters of the word INDOOROOPILLY that begin with I and end with Y?

    4. How many distinct arrangements can be made from the letters of the word INDOOROOPILLY that have all four Os next to each other?

    1. How many distinct arrangements can be made of the letters of the word WOOLLOONGABBA?

    2. How many distinct arrangements are there of the letters of the word WOOLLOONGABBA such that exactly one of the following is true?

    • The two As are not next to each other.
    • The two Bs are not next to each other.
    • The two Ls are not next to each other.
    1. How many one-to-one (injective) functions are there from \(\{a,b,c\}\) to \(\{1,2,3,4,5\}\)?

    2. How many onto (surjective) functions are there from \(\{1,2,3,4,5\}\) to \(\{a,b,c\}\)?

Solutions

    1. There are 13 letters in INDOOROOPILLY, with four O’s, two I’s, two L’s, and the remaining letters appearing once each. Therefore the number of distinct arrangements is

    \[\frac{13!}{4!\,2!\,2!} = 64\,864\,800.\]

    1. Fix R in the first position and P in the last position. There are 11 letters remaining, with four O’s, two I’s, two L’s, and the remaining letters appearing once each. Therefore the number of distinct arrangements is

    \[\frac{11!}{4!\,2!\,2!} = 415\,800.\]

    1. Fix I in the first position and Y in the last position. There are 11 letters remaining, with four O’s, two L’s, and the remaining letters appearing once each. Therefore the number of distinct arrangements is

    \[\frac{11!}{4!\,2!} = 831\,600.\]

    1. Treat the four O’s as one ‘letter’. There are 10 ‘letters’ with two I’s, two L’s, and the remaining ‘letters’ appearing once each. Therefore the number of distinct arrangements is

    \[\frac{10!}{2!\,2!} = 907\,200.\]

    1. There are 13 letters in WOOLLOONGABBA, with four O’s, two L’s, two A’s, two B’s, and the remaining letters appearing once each. Therefore the number of distinct arrangements is

    \[\frac{13!}{4!\,2!\,2!\,2!} = 32\,432\,400.\]

    1. We want exactly one of the pairs AA, BB, LL to be non-adjacent. That is, there are three cases to consider:
    • Case 1: the two A’s are not adjacent but BB and LL are adjacent.
    • Case 2: the two B’s are not adjacent but AA and LL are adjacent.
    • Case 3: the two L’s are not adjacent but AA and BB are adjacent.

    Note that the number of ways Case 1 can occur is exactly the same as the number of ways Case 2 can occur (and similarly for Case 3).

    To count the number of ways Case 1 can occur, we first count arrangements with BB adjacent and LL adjacent. Treat BB as one ‘letter’ and LL as one ‘letter’. Then there are 11 ‘letters’ (W, O, O, LL, O, O, N, G, A, BB, A) to arrange, with four O’s, two A’s, and the remaining ‘letters’ appearing once each. Therefore the number of such arrangements is

    \[\frac{11!}{4!\,2!}.\]

    In order to determine the number of these for which the two A’s are not adjacent, we take this total and subtract those where AA is also adjacent. For this, treat AA, BB, and LL as ‘letters’. Then there are 10 ‘letters’ to arrange, with four O’s and the remaining ‘letters’ appearing once each. Therefore the number of arrangements with AA adjacent and BB adjacent and LL adjacent is

    \[\frac{10!}{4!}.\]

    Hence, the number of ways that Case 1 can occur is

    \[\frac{11!}{4!\,2!} - \frac{10!}{4!}.\]

    Case 2 and Case 3 have exactly the same number of occurrences as Case 1. Also note that the three cases are mutually exclusive (you cannot have both Case 1 and Case 2 occurring at the same time, for example).

    So the number of arrangements of the word WOOLLOONGABBA for which exactly one of the three cases occurs is

    \[3\left(\frac{11!}{4!\,2!} - \frac{10!}{4!}\right).\]

    1. An injective function \(f : \{a,b,c\} \to \{1,2,3,4,5\}\) is determined by choosing distinct images for \(a\), \(b\), \(c\). The number of choices is

    \[5 \cdot 4 \cdot 3 = {}^5P_3 = 60.\]

    1. The number of surjective functions \(f : \{1,2,3,4,5\} \to \{a,b,c\}\) is the total number of functions from \(\{1,2,3,4,5\}\) to \(\{a,b,c\}\) minus the number of functions from \(\{1,2,3,4,5\}\) to \(\{a,b,c\}\) which are not surjective. The former number is \(3^5\). A function which is not surjective must have image of size 1 or 2.
    • In the first case, there are 3 such functions (one for each of the three choices for the image).
    • In the latter case, first choose which 2 elements of \(\{a,b,c\}\) will be in the image. There are \(\binom{3}{2}\) such choices. For this fixed two-element image, the number of functions with this image is

    \[2^5 - 2,\]

    since there are \(2^5\) total functions from a 5-element set to a 2-element set, and we subtract the 2 constant ones (which fall into the first case, having image of size 1). Thus the number of functions from \(\{1,2,3,4,5\}\) to \(\{a,b,c\}\) with an image of size 2 is

    \[\binom{3}{2}(2^5 - 2) = 3 \cdot 30 = 90.\]

    This means there are 93 functions which are not surjective, so the number of surjective functions from \(\{1,2,3,4,5\}\) to \(\{a,b,c\}\) is

    \[3^5 - 93 = 150.\]

§39 — Introduction to Probability

Epp 4th ed. pp. 516–553 · 5th ed. (metric) pp. 564–604. Lecture 31 (week 11).

Problems

  1. To form a committee, five people are to be chosen from a group of ten people.

    1. How many different committees of five can be chosen?

    2. Amongst the ten people, we have Alice, Bob, Eve and Oscar. Alice and Bob are extremely talkative during meetings, and so at most one of these two can be chosen. Eve and Oscar are inseparable, and so a committee must either contain both of them or neither of them. With these constraints, how many different five-person committees can be formed?

    3. What is the probability that none of Alice, Bob, Eve and Oscar are on a committee, given that a random committee of five is chosen subject to the constraints in part (b)?

    1. How many positive two-digit integers are multiples of 3?

    2. What is the probability that a randomly chosen positive two-digit integer is a multiple of 3?

    3. What is the probability that a randomly chosen positive two-digit integer is a multiple of 4?

Solutions

    1. Number of different committees of five is

    \[\binom{10}{5} = \frac{10 \cdot 9 \cdot 8 \cdot 7 \cdot 6}{5 \cdot 4 \cdot 3 \cdot 2} = 252.\]

    1. Let \(A\), \(B\) be Alice and Bob, and let \(E\), \(O\) be Eve and Oscar. There are 6 other people. We split into cases.

    Case 1: the committee contains both \(E\) and \(O\). Then we choose 3 more people from the remaining 8 with the restriction that we cannot choose both \(A\) and \(B\). There are \(\binom{8}{3}\) ways to choose 3 more people from the remaining 8 to be on the committee; of these, there are \(\binom{6}{1}\) ways in which both \(A\) and \(B\) are chosen (such a committee must have \(E, O, A, B\) so there is only one more person to choose from the remaining 6 people). Thus the number of committees for this case is

    \[\binom{8}{3} - \binom{6}{1} = 56 - 6 = 50.\]

    Case 2: the committee contains neither \(E\) nor \(O\). Then we choose 5 people from the remaining 8, again not allowing both \(A\) and \(B\). There are \(\binom{8}{5}\) ways to choose 5 people from the remaining 8 people to be on the committee; of these, there are \(\binom{6}{3}\) ways in which both \(A\) and \(B\) are chosen. Thus the number of committees for this case is

    \[\binom{8}{5} - \binom{6}{3} = 56 - 20 = 36.\]

    Hence the total number is \(50 + 36 = 86\).

    1. To have none of Alice, Bob, Eve, or Oscar on the committee, we must choose all 5 members from the other 6 people, which can be done in \(\binom{6}{5} = 6\) ways. Therefore the probability is

    \[\frac{6}{86} = 0.069767\ldots\]

    1. The positive two-digit multiples of 3 occur as every third number from 12 to 99, so there are

    \[\frac{99 - 12}{3} + 1 = 30\]

    such numbers.

    1. There are 90 positive two-digit integers (9 choices for the first digit and 10 choices for the second digit). Thus, by part (a), the probability that a randomly selected positive two-digit number is a multiple of 3 is

    \[\frac{30}{90} = \frac{1}{3}.\]

    1. The positive two-digit multiples of 4 occur as every fourth number from 12 to 96, so there are

    \[\frac{96 - 12}{4} + 1 = 22\]

    such numbers. Hence the probability that a randomly selected positive two-digit number is a multiple of 4 is

    \[\frac{22}{90} = \frac{11}{45}.\]

§40 — Binomial Coefficients

Epp 4th ed. pp. 565–591 · 5th ed. (metric) pp. 617–641. Lecture 31 (week 11).

Problems

  1. What is the coefficient of \(x^4\) in \((2 + 3x)^7\)?

    1. What is the coefficient of \(x^{12}\) in \((3 - x^3)^6\)?

    2. What is the coefficient of \(x^8\) in \((3 - x^3)^6\)?

  2. What is the coefficient of \(x^2\) in the expansion of \((3-x)^6\)?

  3. What is the coefficient of \(x^5\) in the expansion of \((2x+1)^{20}\)?

  4. What is the coefficient of \(x^5\) in the expansion of \(\left(3x^2 - \dfrac{2}{x}\right)^7\)?

Solutions

  1. By the binomial theorem,

    \[(2 + 3x)^7 = \sum_{k=0}^{7} \binom{7}{k} 2^{7-k} (3x)^k.\]

    The term \(x^4\) appears only when \(k = 4\). This term is \(\binom{7}{4} 2^3 (3x)^4\). Therefore the coefficient of \(x^4\) is

    \[\binom{7}{4} 2^3 3^4 = 35 \times 8 \times 81 = 22\,680.\]

  2. By the binomial theorem,

    \[(3 - x^3)^6 = \sum_{k=0}^{6} \binom{6}{k} 3^{6-k} (-x^3)^k = \sum_{k=0}^{6} \binom{6}{k} 3^{6-k} (-1)^k x^{3k}.\]

    1. The term involving \(x^{12}\) is \(\binom{6}{4} 3^2 (-1)^4 x^{3 \cdot 4}\), so the coefficient of \(x^{12}\) is

    \[15 \times 9 \times (-1)^4 = 135.\]

    1. Every power of \(x\) that appears is a multiple of 3. Since 8 is not a multiple of 3, there is no \(x^8\) term, so the coefficient of \(x^8\) is 0.
  3. By the binomial theorem,

    \[(3-x)^6 = \sum_{i=0}^{6} \binom{6}{i} 3^{6-i} (-x)^i = \sum_{i=0}^{6} \binom{6}{i} 3^{6-i} (-1)^i x^i.\]

    The term \(x^2\) appears only when \(i = 2\), so the coefficient of \(x^2\) is

    \[\binom{6}{2} 3^{6-2} (-1)^2 = \binom{6}{2} 3^4 = 1215.\]

  4. By the binomial theorem,

    \[(2x+1)^{20} = \sum_{k=0}^{20} \binom{20}{k} (2x)^{20-k} (1)^k.\]

    The term \(x^5\) appears only when \(k = 15\), where the term is \(\binom{20}{15}(2x)^5\), so the coefficient of \(x^5\) is

    \[\binom{20}{5} 2^5 = 496\,128.\]

  5. By the binomial theorem,

    \[\left(3x^2 - \frac{2}{x}\right)^7 = \sum_{i=0}^{7} \binom{7}{i} (3x^2)^{7-i}\left(-\frac{2}{x}\right)^i = \sum_{i=0}^{7} \binom{7}{i} 3^{7-i} (-2)^i x^{14-3i}.\]

    The term \(x^5\) appears only when \(14 - 3i = 5\), i.e. \(i = 3\). Therefore the coefficient of \(x^5\) is

    \[\binom{7}{3} 3^4 (-2)^3 = -22\,680.\]

§41 — Inclusion Exclusion

Epp 4th ed. pp. 545–549 · 5th ed. (metric) pp. 595–599. Lecture 32 (week 12).

Problems

  1. Out of 100 people surveyed, 40 people like apples, 50 like bananas and 40 like cherries. 15 like apples and bananas, 16 like bananas and cherries, 17 like cherries and apples and 10 people like all three. How many people do not like any of the three fruits?

  2. How many integers between 1 and 600 inclusive are coprime with 15? That is, how many integers \(n\) satisfy \(1 \leq n \leq 600\) and \(\gcd(n, 15) = 1\)?

  3. Seven friends, named Alice, Bob, Charlie, Delia, Edward, Fred and George, stand in a line for a photograph. How many different arrangements are there of the seven friends standing in a line such that Alice is not next to Charlie and Bob is not next to Delia?

Solutions

  1. Let \(A\) be the set of people who like apples, \(B\) the set of people who like bananas, and \(C\) the set of people who like cherries. We are given

    \[|A| = 40,\ |B| = 50,\ |C| = 40,\quad |A \cap B| = 15,\ |B \cap C| = 16,\ |C \cap A| = 17,\quad |A \cap B \cap C| = 10.\]

    By inclusion–exclusion,

    \[\begin{aligned} |A \cup B \cup C| &= |A| + |B| + |C| - |A \cap B| - |B \cap C| - |C \cap A| + |A \cap B \cap C| \\ &= 40 + 50 + 40 - 15 - 17 - 16 + 10 \\ &= 92. \end{aligned}\]

    Hence the number of people who do not like any of the three fruits is

    \[100 - |A \cup B \cup C| = 100 - 92 = 8.\]

  2. Let \(U = \{1, 2, \dots, 600\}\). Define

    \[A = \{n \in U : 3 \mid n\}, \qquad B = \{n \in U : 5 \mid n\}.\]

    Then \(n\) is not coprime to 15 if and only if \(n \in A \cup B\). We have

    \[|A| = \frac{600}{3} = 200, \qquad |B| = \frac{600}{5} = 120, \qquad |A \cap B| = \frac{600}{15} = 40.\]

    By inclusion–exclusion,

    \[|A \cup B| = |A| + |B| - |A \cap B| = 200 + 120 - 40 = 280.\]

    Therefore the number of \(n\) with \(\gcd(n,15) = 1\) is

    \[|U| - |A \cup B| = 600 - 280 = 320.\]

  3. Let \(U\) be the set of all linear arrangements of the 7 friends, so \(|U| = 7!\). Define

    \[A = \{\text{arrangements in which Alice is next to Charlie}\},\] \[B = \{\text{arrangements in which Bob is next to Delia}\}.\]

    We want to count the number of arrangements in \(U\) which are neither in \(A\) nor in \(B\). The required number is \(|U \setminus (A \cup B)| = |U| - |A \cup B|\).

    To count \(|A|\), treat \(\{\)Alice, Charlie\(\}\) as a single block. There are \(6!\) ways to arrange the 6 “objects”, and for each of these there are 2 ways to order Alice and Charlie within their block (either Alice or Charlie to the left). Hence

    \[|A| = 2 \cdot 6!.\]

    Similarly, \(|B| = 2 \cdot 6!\).

    To count \(|A \cap B|\), treat both pairs \(\{\)Alice, Charlie\(\}\) and \(\{\)Bob, Delia\(\}\) as two single blocks. There are \(5!\) ways to arrange the 5 objects (two blocks plus the other three people), and for each of these there are \(2 \cdot 2\) internal orders of the two blocks. Hence

    \[|A \cap B| = 4 \cdot 5!.\]

    By inclusion–exclusion,

    \[|A \cup B| = |A| + |B| - |A \cap B| = 2 \cdot 6! + 2 \cdot 6! - 4 \cdot 5!.\]

    Therefore

    \[|U \setminus (A \cup B)| = 7! - \left(2 \cdot 6! + 2 \cdot 6! - 4 \cdot 5!\right) = 2640.\]

§42 — The Pigeonhole Principle

Epp 4th ed. pp. 554–565 · 5th ed. (metric) pp. 604–616. Lecture 33 (week 12).

Problems

    1. Explain why, in a group of 50 people, at least five people were born in the same month.

    2. How many people do you need to ensure that at least seven people were born in the same month?

  1. Show that, if we select six distinct integers from the set \(\{1, 2, \dots, 10\}\), two of these must add to give 11.

  2. A computer network consists of six computers. Each computer is directly connected to zero or more of the other computers. Show that there are at least two computers in the network that are directly connected to the same number of other computers.

  3. During a month with 30 days, a baseball team plays at least one game a day, but no more than 45 games. Show that there must be a period of some number of consecutive days during which the team must play exactly 14 games.

  4. Let \(S = \{1, 2, \dots, 100\}\). What is the minimum number of integers you must pick from \(S\) to be sure that at least one pair has a difference which is divisible by 3?

Solutions

    1. There are 12 months (pigeonholes) and 50 people (pigeons). If each month had at most 4 birthdays, then there would be at most \(12 \cdot 4 = 48\) people. Since \(50 > 48\), by the pigeonhole principle some month must contain at least 5 people.

    2. To guarantee at least 7 people are born in the same month, note that if each month had at most 6 birthdays, then there could be at most \(12 \cdot 6 = 72\) people. Therefore with 73 people, by the pigeonhole principle some month has at least 7 people. Moreover, 72 people is not enough (take 6 born in each month), so the minimum is 73.

  1. Partition \(A = \{1, 2, \dots, 10\}\) into the 5 pairs

    \[\{1,10\},\ \{2,9\},\ \{3,8\},\ \{4,7\},\ \{5,6\},\]

    each summing to 11. In order for a selection of distinct integers from the set \(A\) to not have two that add to give 11, at most one integer can be chosen from each of the subsets. Choosing 6 distinct integers from \(A\) forces us to select two numbers from the same pair (there are 5 pairs but 6 choices), by the pigeonhole principle. Those two numbers then add to 11.

  2. For each computer, let its degree be the number of other computers it is directly connected to. With 6 computers, each degree is an integer in \(\{0,1,2,3,4,5\}\). It is impossible to have both a computer of degree 0 and a computer of degree 5 (if one is connected to everyone, then none has degree 0). Hence the degrees must all lie in one of the two 5-element sets

    \[\{0,1,2,3,4\} \quad \text{or} \quad \{1,2,3,4,5\}.\]

    So, in either case, there are only 5 possible degree values available for 6 computers. By the pigeonhole principle, at least two computers have the same degree, i.e. are connected to the same number of other computers.

  3. Let \(g_i\) be the total number of games played in the first \(i\) days, for \(i = 1, 2, \dots, 30\). Then

    \[1 \leq g_1 < g_2 < \cdots < g_{30} \leq 45\]

    because at least one game is played each day. Consider also the 30 numbers

    \[g_1 + 14,\ g_2 + 14,\ \dots,\ g_{30} + 14,\]

    which satisfy \(15 \leq g_1 + 14 < \cdots < g_{30} + 14 \leq 59\). Thus we have 60 integers

    \[g_1, \dots, g_{30},\ g_1 + 14, \dots, g_{30} + 14\]

    all lying in the set \(\{1, 2, \dots, 59\}\) of size 59. By the pigeonhole principle, two of these 60 integers are equal. Since the \(g_i\) are strictly increasing, no equality can occur among the \(g_i\) themselves, and likewise no equality can occur among the \(g_i + 14\) themselves. Therefore we must have

    \[g_j = g_i + 14\]

    for some \(i, j \in \{1, \dots, 30\}\). Then the number of games played from day \(i+1\) through day \(j\) is

    \[g_j - g_i = 14,\]

    so there is a period of consecutive days during which exactly 14 games are played.

  4. First, we note that two integers have a difference that is divisible by 3 if and only if they are congruent modulo 3 (have the same remainder when divided by 3). Now, the set \(S\) can be partitioned into three sets based on their congruence class modulo 3. By the pigeonhole principle, if we select 4 numbers from \(S\), at least two of them must belong to the same congruence class. These two will have difference that is divisible by 3. Note that 4 is the minimum number of selections that guarantees this, since it is possible to select 3 numbers where no two of them have difference divisible by 3 (select one from each congruence class).

§43 — Introduction to Graph Theory

Epp 4th ed. pp. 625–660 · 5th ed. (metric) pp. 677–697. Lecture 33 (week 12).

Problems

  1. For each of the following, state whether or not there exists a simple graph with the stated number of vertices and degrees. Draw the graph, or explain why one does not exist.

    1. Five vertices with degrees 3, 3, 2, 2, 2.

    2. Seven vertices with degrees 4, 2, 3, 1, 1, 1, 1.

    3. Five vertices with degrees 5, 4, 3, 1, 1.

  2. A graph has 10 edges and 6 vertices. Five of the vertices have degree 3. What is the degree of the remaining vertex?

  3. Let \(n\) be a positive integer and let \(K_n\) be the complete graph with \(n\) vertices. How many edges does \(K_n\) have?

  4. A graph has 27 edges and every vertex has degree 2, 3 or 4. If there are 7 vertices that have degree 4 and 4 vertices that have degree 3, then how many vertices of degree 2 are there?

Solutions

    1. Yes, such a simple graph exists. One drawing has vertices \(a, b, c, d, e\) and the six edges

    \[ab,\quad ac,\quad ad,\quad bc,\quad be,\quad de,\]

    giving \(\deg(a) = 3\), \(\deg(b) = 3\), \(\deg(c) = 2\), \(\deg(d) = 2\), \(\deg(e) = 2\) as required.

    1. No such graph exists. If the degrees are 4, 2, 3, 1, 1, 1, 1, then the sum of degrees is

    \[4 + 2 + 3 + 1 + 1 + 1 + 1 = 13,\]

    which is odd. Remember (by the Handshake Theorem) that the sum of the degrees in any graph equals twice the number of edges, and so is always even. So such a graph cannot exist.

    1. No such graph exists. In a simple graph on 5 vertices, the largest possible degree is 4, but the sequence contains a vertex of degree 5, which is impossible. So there is no such simple graph.
  1. Since the graph has 10 edges, the sum of all the degrees is \(2 \cdot 10 = 20\). If five vertices have degree 3, the sum of their degrees is \(5 \cdot 3 = 15\), so the remaining vertex has degree

    \[20 - 15 = 5.\]

  2. In \(K_n\), an edge is determined by choosing 2 distinct vertices, so the number of edges is

    \[\binom{n}{2} = \frac{n(n-1)}{2}.\]

  3. Let \(x\) be the number of vertices of degree 2. Since the graph has 27 edges, the sum of degrees is \(2 \cdot 27 = 54\). Given that there are seven vertices of degree 4 and four vertices of degree 3, and all other vertices have degree 2, we have

    \[2x + 4 \cdot 7 + 3 \cdot 4 = 54.\]

    This means \(x = 7\).

§44 — Walks, Trails and Circuits

Epp 4th ed. pp. 625–660 · 5th ed. (metric) pp. 677–697. Lecture 34 (week 12).

Problems

  1. Let \(G\) and \(H\) be the graphs below.

    \(G\) has vertex set \(\{v_1, v_2, v_3, v_4, v_5\}\) and is drawn with a loop at \(v_2\), a loop at \(v_5\), and two parallel edges between \(v_1\) and \(v_2\). Its degree sequence is

    \[\deg(v_1) = 4,\quad \deg(v_2) = 6,\quad \deg(v_3) = 4,\quad \deg(v_4) = 2,\quad \deg(v_5) = 4,\]

    so \(G\) is connected with 10 edges.

    \(H\) has vertex set \(\{v_1, v_2, v_3, v_4, v_5\}\) and is a simple connected graph drawn as the 4-cycle \(v_1 v_2 v_3 v_4 v_1\) with a single pendant edge \(v_3 v_5\), giving degrees

    \[\deg(v_1) = 2,\quad \deg(v_2) = 2,\quad \deg(v_3) = 3,\quad \deg(v_4) = 2,\quad \deg(v_5) = 1.\]

    For each of the following statements, explain why it is true or false.

    1. \(G\) is a simple graph.

    2. The sum of the degrees of the vertices of \(G\) is 18.

    3. \(G\) contains an Euler circuit.

    4. \(H\) contains an Euler trail.

  2. Does the following graph contain an Euler circuit? (The graph shown is connected and has exactly two vertices of degree 3.)

  3. Let \(G\) be a simple, connected, Eulerian graph with eight vertices and let \(e\) be the number of edges in \(G\). Prove that \(8 \leq e \leq 24\).

Solutions

    1. This statement is false. The graph \(G\) has loops (at \(v_2\) and \(v_5\)) and it also has multiple edges between \(v_1\) and \(v_2\), so \(G\) is not simple.

    2. This statement is false. A loop contributes 2 to the degree, so we have

    \[\deg(v_1) = 4,\quad \deg(v_2) = 6,\quad \deg(v_3) = 4,\quad \deg(v_4) = 2,\quad \deg(v_5) = 4,\]

    and thus the sum of the degrees is \(4 + 6 + 4 + 2 + 4 = 20 \neq 18\).

    1. This statement is true. The graph \(G\) is connected and every vertex has even degree (as computed above), so \(G\) contains an Euler circuit.

    2. This statement is true. In \(H\), the degrees are

    \[\deg(v_1) = 2,\ \deg(v_2) = 2,\ \deg(v_4) = 2,\ \deg(v_3) = 3,\ \deg(v_5) = 1,\]

    so exactly two vertices have odd degree (namely \(v_3\) and \(v_5\)). Therefore \(H\) has an Euler trail (starting at one of \(v_3\), \(v_5\) and ending at the other).

  1. No. The graph is connected, but has two vertices of degree 3, so not all vertices have even degree. Hence there is no Euler circuit.

  2. Since \(G\) is connected and Eulerian, every vertex has even degree, and (because \(G\) is simple and connected with 8 vertices) each vertex has degree at least 2. Hence

    \[2e = \sum_{v} \deg(v) \geq 8 \cdot 2 = 16,\]

    so \(e \geq 8\). Moreover, in a simple graph on 8 vertices we have \(\deg(v) \leq 7\), and because \(G\) is Eulerian all degrees are even, so in fact \(\deg(v) \leq 6\) for every vertex. Therefore

    \[2e = \sum_{v} \deg(v) \leq 8 \cdot 6 = 48,\]

    so \(e \leq 24\). Hence \(8 \leq e \leq 24\). \(\square\)

§45 — Matrix Representations of Graphs

Epp 4th ed. pp. 661–675 · 5th ed. (metric) pp. 698–712. Lecture 35 (week 13).

Problems

  1. Let \(G\) be the graph with vertex set \(\{v_1, v_2, v_3, v_4\}\) and edge set \(\{a,b,c,d,e,f\}\), where the edges are

    edge endpoints
    \(a\) \(v_1 v_2\)
    \(b\) \(v_2 v_3\)
    \(c\) \(v_2 v_3\) (parallel to \(b\))
    \(d\) \(v_3 v_4\)
    \(e\) \(v_2 v_4\)
    \(f\) loop at \(v_4\)
    1. Write down an adjacency matrix for \(G\). Make sure you clearly label the rows and columns of your matrix.

    2. Write down an incidence matrix for \(G\). Make sure you clearly label the rows and columns of your matrix.

    3. Does \(G\) have an Euler circuit? If yes, write it out as a sequence of vertices and edges. If not, determine the fewest number of edges that would need to be added to \(G\) to create a new graph that does have an Euler circuit.

  2. Let \(G\) be the graph with the following adjacency matrix.

    \[\begin{pmatrix} 0 & 1 & 0 & 0 & 3 \\ 1 & 0 & 2 & 0 & 0 \\ 0 & 2 & 0 & 1 & 1 \\ 0 & 0 & 1 & 0 & 1 \\ 3 & 0 & 1 & 1 & 0 \end{pmatrix}\]

    1. Draw the graph \(G\).

    2. Explain whether or not \(G\) has an Euler circuit or an Euler trail.

  3. Let \(G\) be the graph with the following adjacency matrix.

    \[\begin{pmatrix} 0 & 1 & 0 & 1 & 0 & 0 \\ 1 & 0 & 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 0 & 1 & 1 \\ 1 & 1 & 0 & 0 & 0 & 0 \\ 0 & 0 & 1 & 0 & 0 & 1 \\ 0 & 0 & 1 & 0 & 1 & 0 \end{pmatrix}\]

    Explain whether or not \(G\) is Eulerian.

Solutions

    1. The labelled adjacency matrix for \(G\) is
    \(v_1\) \(v_2\) \(v_3\) \(v_4\)
    \(\mathbf{v_1}\) 0 1 0 0
    \(\mathbf{v_2}\) 1 0 2 1
    \(\mathbf{v_3}\) 0 2 0 1
    \(\mathbf{v_4}\) 0 1 1 1

    (where the entry 1 on the diagonal at \(v_4\) records the single loop at \(v_4\)).

    1. The labelled incidence matrix for \(G\) is
    \(a\) \(b\) \(c\) \(d\) \(e\) \(f\)
    \(\mathbf{v_1}\) 1 0 0 0 0 0
    \(\mathbf{v_2}\) 1 1 1 0 1 0
    \(\mathbf{v_3}\) 0 1 1 1 0 0
    \(\mathbf{v_4}\) 0 0 0 1 1 2

    (where the entry 2 in the bottom right indicates the loop that is incident twice with \(v_4\)).

    1. The degrees are

    \[\deg(v_1) = 1,\quad \deg(v_2) = 4,\quad \deg(v_3) = 3,\quad \deg(v_4) = 4,\]

    so \(G\) does not have an Euler circuit (since \(v_1\) and \(v_3\) have odd degree).

    The fewest edges to add to make all degrees even is 1; for example, adding an edge \(g\) between vertices \(v_1\) and \(v_3\) creates a new connected graph in which all of the vertices have even degree.

    An example of an Euler circuit in the new graph is

    \[v_1,\ a,\ v_2,\ b,\ v_3,\ d,\ v_4,\ f,\ v_4,\ e,\ v_2,\ c,\ v_3,\ g,\ v_1.\]

    1. Reading the matrix off (with vertices \(v_1, \dots, v_5\) in matrix order), \(G\) has the edges
    • \(v_1 v_2\) — a single edge,
    • \(v_1 v_5\) — three parallel edges,
    • \(v_2 v_3\) — two parallel edges,
    • \(v_3 v_4\) — a single edge,
    • \(v_3 v_5\) — a single edge,
    • \(v_4 v_5\) — a single edge.
    1. The degrees are

    \[\deg(v_1) = 4,\ \deg(v_2) = 3,\ \deg(v_3) = 4,\ \deg(v_4) = 2,\ \deg(v_5) = 5,\]

    so there is no Euler circuit (since \(v_2\) and \(v_5\) have odd degree), but there is an Euler trail (exactly two odd-degree vertices) from \(v_2\) to \(v_5\).

  1. Let the vertices be \(v_1, v_2, v_3, v_4, v_5, v_6\) in the order of the matrix. Reading the matrix off, \(G\) has edges

    \[v_1 v_2,\quad v_1 v_4,\quad v_2 v_4,\quad v_3 v_5,\quad v_3 v_6,\quad v_5 v_6,\]

    that is, two disjoint triangles: one on \(\{v_1, v_2, v_4\}\) and one on \(\{v_3, v_5, v_6\}\).

    The graph is not connected (e.g. there is no walk from vertex \(v_1\) to vertex \(v_3\)). Hence \(G\) is not Eulerian.

    (Errata note: the source document’s solution to 1(b) was corrected on 11 August 2026 — the note in brackets had previously copied the note from 1(a). The corrected note is the one given above.)

§46 — Trees

Epp 4th ed. pp. 683–701 · 5th ed. (metric) pp. 720–742. Lecture 36 (week 13).

Problems

  1. A tree has 11 vertices, of which exactly four vertices have degree 1 and exactly one vertex has degree 4. Determine the degrees of the remaining six vertices.

  2. Let \(T\) be a tree with \(p\) vertices, where \(p \geq 11\). Suppose \(T\) has precisely six vertices of degree 1 and precisely four vertices of degree 3. Determine the degree of each of the remaining \(p - 10\) vertices of \(T\).

  3. Let \(T\) be a tree on 12 vertices. Suppose \(T\) has exactly three vertices of degree 3 and each of the remaining vertices has degree 1 or 5. Using theorems that were given in lectures, determine the number of vertices of degree 5.

  4. Two trees with the same vertex set \(\{1, 2, \dots, n\}\) are distinct if their edge sets are not equal. Use mathematical induction to prove that for each integer \(n \geq 2\), the number of distinct trees with vertex set \(\{1, 2, \dots, n\}\) is at least \((n-1)!\).

Solutions

  1. A tree with 11 vertices has \(e = 10\) edges, so by the Handshake Theorem, the sum of degrees is \(2e = 20\). We are given that exactly four vertices have degree 1 and exactly one vertex has degree 4. Let \(v_1, v_2, \dots, v_6\) be the remaining vertices whose degrees are to be determined. We have

    \[\begin{aligned} \sum_{i=1}^{6} \deg(v_i) + 4 \cdot 1 + 1 \cdot 4 &= 20 \\ \sum_{i=1}^{6} \deg(v_i) &= 12. \end{aligned}\]

    None of these six vertices \(v_1, v_2, \dots, v_6\) can have degree 1 (since there are exactly four vertices of degree 1, already accounted for). Hence each has degree at least 2. Because six vertices of degree at least 2 have total degree 12, each must have degree exactly 2.

    Thus the remaining six vertices all have degree 2.

  2. \(T\) is a tree with \(p\) vertices and so must have \(p-1\) edges. Now by the Handshake Theorem we know

    \[2(p-1) = \sum_{i=1}^{p} \deg(v_i).\]

    Using the known degrees we can write this as

    \[\begin{aligned} 2(p-1) &= 6 \cdot 1 + 4 \cdot 3 + \sum_{i=1}^{p-10} \deg(v_i) \\ 2(p-1) &= 18 + \sum_{i=1}^{p-10} \deg(v_i) \\ 2p - 2 - 18 &= \sum_{i=1}^{p-10} \deg(v_i) \\ 2(p-10) &= \sum_{i=1}^{p-10} \deg(v_i). \end{aligned}\]

    Rewriting the final line as

    \[\frac{\sum_{i=1}^{p-10} \deg(v_i)}{p-10} = 2, \qquad \text{i.e.} \qquad \frac{\deg(v_1) + \deg(v_2) + \cdots + \deg(v_{p-10})}{p-10} = 2,\]

    we see that the average degree of the remaining \(p-10\) vertices is 2. However, we know that there are no more vertices of degree 1. Thus each of the remaining \(p-10\) vertices must have degree exactly 2.

  3. Let \(x\) be the number of vertices of degree 5 and \(y\) the number of vertices of degree 1. Then, because the tree has 12 vertices, we have

    \[x + y + 3 = 12.\]

    This simplifies to give \(x + y = 9\).

    Now, since \(T\) is a tree on 12 vertices, it must have \(e = 11\) edges. By the Handshake Theorem, the sum of degrees is \(2e = 22\), so

    \[3 \cdot 3 + 5x + 1 \cdot y = 22.\]

    This simplifies to give \(5x + y = 13\).

    Subtracting \(x + y = 9\) from \(5x + y = 13\) gives \(4x = 4\), hence \(x = 1\). Therefore there is exactly one vertex of degree 5.

  4. We prove by induction that the number of distinct trees with vertex set \(\{1, 2, \dots, n\}\) is at least \((n-1)!\).

    Basis: let \(n = 2\). Here \((n-1)! = (2-1)! = 1\). There is exactly one tree on \(\{1,2\}\), namely the single edge \(\{1,2\}\). So the base case is true.

    Inductive hypothesis. Suppose \(k \geq 2\) is an integer and there are at least \((k-1)!\) distinct trees on \(\{1, 2, \dots, k\}\).

    Now consider \(n = k+1\). We aim to show that there are at least \(k!\) distinct trees on vertex set \(\{1, 2, \dots, k, k+1\}\).

    Let \(T\) be a tree on vertex set \(\{1, 2, \dots, k\}\). For any choice of \(m \in \{1, 2, \dots, k\}\), we form a graph \(T_m\) on vertex set \(\{1, 2, \dots, k, k+1\}\) as follows: start with tree \(T\), add a new vertex \(k+1\), and the edge \(\{k+1, m\}\).

    Note that \(T_m\) is connected (since \(k+1\) is joined to \(m\), and \(T\) was connected). Also, \(T_m\) has no circuit because \(T\) is a tree and adding a new leaf does not create a circuit. Thus \(T_m\) is a tree. Moreover, different choices of \(m\) give rise to distinct trees \(T_m\), because their edge sets are different.

    By the inductive hypothesis, there are at least \((k-1)!\) choices for the tree \(T\). Since there are \(k\) choices for the vertex \(m\), there are at least \(k \cdot (k-1)! = k!\) trees \(T_m\) on vertex set \(\{1, 2, \dots, k+1\}\).

    By the principle of mathematical induction, this completes the proof. \(\square\)

See also

Proof Techniques — Direct, Counterexample, Contradiction, Contraposition

The four methods for establishing (or demolishing) a statement of the form \(\forall x \in D, P(x) \to Q(x)\), plus the definitions every proof in this course is expected to unfold.

Definitions to prove from

Proofs are written by demonstrating facts using the definitions — so these have to be at your fingertips. Each is a biconditional: it can be used in either direction.

Term Definition
\(n\) is even \(n = 2k\) for some integer \(k\)
\(n\) is odd \(n = 2k + 1\) for some integer \(k\) (equivalently \(n = 2k - 1\) for some integer \(k\))
\(n\) is prime \(n \in \mathbb{Z}\), \(n > 1\), and \(\forall r, s \in \mathbb{Z}^+\), if \(n = rs\) then (\(r = 1\) and \(s = n\)) or (\(r = n\) and \(s = 1\))
\(n\) is composite \(n \in \mathbb{Z}\), \(n > 1\), and \(\exists r, s \in \mathbb{Z}^+\) such that \(n = rs\) and \(1 < r < n\) and \(1 < s < n\)
\(r\) is rational \(\exists a, b \in \mathbb{Z}\) such that \(r = \frac{a}{b}\) and \(b \neq 0\)
\(d \mid n\) \(d \neq 0\) and \(n = dk\) for some integer \(k\)

Two facts used constantly, and worth naming explicitly when you lean on them:

  • Fact: the sum, difference, and product of integers is an integer.
  • Fact: every integer is either even or odd.

Parity

Almost every early proof in this course turns on whether something is even or odd, so the consequences of those two definitions are worth having ready rather than re-deriving:

Result
\(\text{even} + \text{even}\), \(\text{odd} + \text{odd}\) even
\(\text{even} + \text{odd}\) odd
\(\text{even} \times\) anything even
\(\text{odd} \times \text{odd}\) odd
consecutive integers opposite parity — so two integers two apart share their parity

Each row follows from the two definitions by a one-line substitution. Subtraction behaves like addition throughout, since \(m - n\) and \(m + n\) differ by \(2n\).

The “two apart” row is the whole content of the warm-up in 2026-07-27-welcome-and-intro-problems, where consecutive \(p < q < r\) forces \(p + r\) to be even and \(r - p\) to be exactly \(2\), while \(pr\) is free to be either.

Prime vs. not-prime. For \(n \in \mathbb{Z}^{>1}\), “\(n\) is not prime” means “\(n\) is composite”. For \(n \in \mathbb{Z}\) generally, “\(n\) is not prime” means “\(n\) is composite or \(n \leq 1\)” — the integers split into \(\mathbb{Z}^{\leq 1}\), \(\mathbb{Z}^{\text{prime}}\), and \(\mathbb{Z}^{\text{composite}}\), and \(1\) and everything below it is in none of the latter two. This is why \(n = 1\) is the standard counterexample to “every positive integer is prime or composite”.

Warning: don’t assume something that seems like an “obvious fact” until you have proven it, or can name the lecture or pre-work video it was proved in.

The four methods

All four target \(\forall x \in D, P(x) \to Q(x)\). Recall the truth table for \(p \to q\) — the only row that makes it false is \(p\) true, \(q\) false:

\(p\) \(q\) \(p \to q\)
T T T
T F F
F T T
F F T

Every method below is a different way of attacking that one row.

Figure 6: The skeletons of the four methods for a statement of the form ∀x ∈ D, P(x) → Q(x). Three of the four establish the statement and one demolishes it, which is what the red verdict lines record.

Direct proof (shows the statement is true)

Let \(x \in D\) and suppose \(P(x)\). \(\dots\) Hence \(Q(x)\). \(\square\)

Why suppose \(P(x)\) is true, when \(P(x) \to Q(x)\) holds anyway if \(P(x)\) is false? Because the only case that could make the statement false is \(P(x)\) true and \(Q(x)\) false. Ruling that out is the whole job — the \(P(x)\)-false rows are true for free.

Disproof by counterexample (shows the statement is false)

Let \(x = \cdots\). Then \(x \in D\) and \(P(x)\) is true. However, \(Q(x)\) is false. So \(x\) is a counterexample and the statement is false.

A counterexample to \(p \to q\) needs \(p\) true and \(q\) false — a value where the hypothesis fails proves nothing.

Proof by contradiction (shows the statement is true)

Suppose the negation of \(\forall x \in D, P(x) \to Q(x)\) is true. That is, suppose \(\exists x \in D\) such that \(P(x)\) is true but \(Q(x)\) is false. \(\dots\) Hence something that is obviously wrong, like \(P(x)\) is false, or \(Q(x)\) is true, or \(1 = 0\). This is a contradiction, and we conclude that \(\forall x \in D, P(x) \to Q(x)\) is true. \(\square\)

The set-up comes straight from quantified-statements: the negation of \(\forall x \in D, P(x) \to Q(x)\) is \(\exists x \in D\) such that \(P(x) \land \sim Q(x)\).

Useful negations to have ready (see logical-equivalence-laws):

\[\sim(p \lor q) \equiv \sim p \land \sim q \qquad \sim(p \land q) \equiv \sim p \lor \sim q \qquad \sim(p \to q) \equiv p \land \sim q\]

Proof by contraposition (shows the statement is true)

We prove the contrapositive of \(\forall x \in D, P(x) \to Q(x)\), namely that \(\forall x \in D, \sim Q(x) \to \sim P(x)\). Let \(x \in D\) and suppose \(\sim Q(x)\). \(\dots\) Hence, \(\sim P(x)\). \(\square\)

This is valid because \(p \to q \equiv \sim q \to \sim p\) (see conditional-statements).

Choosing a method

  • Contradiction is the most versatile, but it needs more set-up, so a direct proof is often easier when one is available.
  • Contraposition is the natural choice when \(\sim Q(x)\) gives you something concrete to work with and \(P(x)\) doesn’t — typically when the conclusion is a negative statement (“\(r\) is irrational”, “\(n\) is odd”) whose negation is a definition you can expand.
  • Contradiction vs. contraposition: a contraposition proof is a contradiction proof that never actually needs the assumption \(P(x)\) — you derive \(\sim P(x)\) from \(\sim Q(x)\) alone. If your “contradiction” proof only ever uses the \(\sim Q(x)\) half of the assumption, write it as a contraposition instead.
  • Counterexample vs. contradiction: counterexample is a disproof, contradiction is a proof. They are not variants of each other.

Worked comparisons on the same statements live in the lecture notes: 2026-08-13-proof-by-contradiction proves \(\forall n \in \mathbb{Z}\), if \(n\) is odd then \(3n+2\) is odd both directly and by contradiction, and 2026-08-13-proof-by-contraposition works through statements where one method gets stuck and another doesn’t.

Worked examples

Direct proof

\(\forall n \in \mathbb{Z}\), if \(n\) is odd then \(3n + 2\) is odd.

Let \(n \in \mathbb{Z}\) and suppose \(n\) is odd. Then \(n = 2k + 1\) for some \(k \in \mathbb{Z}\). Now \[3n + 2 = 3(2k+1) + 2 = 6k + 5 = 2(3k + 2) + 1.\] Since \(k \in \mathbb{Z}\), we have \(3k + 2 \in \mathbb{Z}\), and hence \(3n + 2\) is odd. \(\square\)

For any integer \(x\), if \(x + 6 = 4y\) for some integer \(y\), then \(\frac{x}{2}\) is an odd integer.

Let \(x \in \mathbb{Z}\) and suppose \(x + 6 = 4y\) for some \(y \in \mathbb{Z}\). Then \(x = 4y - 6 = 2(2y - 3)\), so \[\frac{x}{2} = \frac{2(2y-3)}{2} = 2y - 3 = 2(y - 1) - 1.\] Since \(y \in \mathbb{Z}\) we have \(2y - 3 \in \mathbb{Z}\), so \(\frac{x}{2} \in \mathbb{Z}\); and \(y - 1 \in \mathbb{Z}\), so by definition \(\frac{x}{2}\) is odd. \(\square\)

The sum of any pair of even integers is even.

Let \(m, n \in \mathbb{Z}\) be even. Then \(m = 2k\) and \(n = 2\ell\) for some \(k, \ell \in \mathbb{Z}\). Hence \(m + n = 2k + 2\ell = 2(k + \ell)\), which is even. \(\square\)

\(\forall n \in \mathbb{Z}^+\), if \(n \geq 4\) then \(2n^2 - 5n + 2\) is composite.

Suppose \(n \in \mathbb{Z}^+\) and \(n \geq 4\). Factor \(2n^2 - 5n + 2 = (2n - 1)(n - 2)\). Since \(n \geq 4\), we have \(2n - 1 \geq 7\) and \(n - 2 \geq 2\), so both factors are integers greater than \(1\). Hence the expression is composite. \(\square\)

Disproof by counterexample

  • For all integers \(m\) and \(n\), if \(2m + n\) is odd then \(m\) and \(n\) are both odd. Counterexample \(m = 4, n = 5\): \(2(4) + 5 = 13\) is odd, but \(m\) is even. (Not \(m=3, n=5\) — there the conclusion holds; not \(m=3,n=6\) or \(m=4,n=6\) — there the hypothesis fails.)
  • \(\forall n \in \mathbb{Z}^{\geq 2}\), \(n\) is composite or \(n + 1\) is composite. Take \(n = 2\): both \(2\) and \(3\) are prime.
  • For each integer \(n \geq 2\), the product of the first \(n\) primes minus \(1\) is prime. Take \(n = 4\): \(2 \cdot 3 \cdot 5 \cdot 7 - 1 = 209 = 11 \cdot 19\), composite.
  • \(\forall x, y \in \mathbb{R}\), if \(y^2 > x^2\) then \(y > x\). Take \(x = 1, y = -2\): \(4 > 1\) but \(-2 \not> 1\).
  • \(\forall x \in \mathbb{R}, \lfloor x^2 \rfloor = \lfloor x \rfloor^2\). Take \(x = \frac{3}{2}\): \(\lfloor \frac{9}{4} \rfloor = 2\) but \(\lfloor \frac{3}{2} \rfloor^2 = 1\). (See modular-arithmetic for the floor function.)

Disproof by proving the negation

When the statement to disprove is existential, a single counterexample isn’t available — you must prove the universal negation.

\(\exists n \in \mathbb{Z}^+\) such that \(n^2 + 5n + 6\) is prime. This is false; the negation is \(\forall n \in \mathbb{Z}^+\), \(n^2 + 5n + 6\) is composite.

Let \(n \in \mathbb{Z}^+\). Now \(n^2 + 5n + 6 = (n+2)(n+3)\). Since \(n \geq 1\) we have \(n + 2 > 1\) and \(n + 3 > 1\), so \(n^2 + 5n + 6\) is composite. \(\square\)

(Because \(n \geq 1\) forces \(n^2 + 5n + 6 \geq 2\), “not prime” here does mean “composite”.)

There is an even integer \(n\) such that \(5n - 4\) is prime. False; prove \(\forall n \in \mathbb{Z}^{\text{even}}\), \(5n - 4\) is not prime.

Suppose \(n\) is even, so \(n = 2k\) for some \(k \in \mathbb{Z}\). Then \(5n - 4 = 10k - 4 = 2(5k - 2)\), which is even. The only even prime is \(2\); if \(2(5k-2) = 2\) then \(k = \frac{3}{5} \notin \mathbb{Z}\). Therefore \(5n - 4\) is not prime. \(\square\)

Proof by contradiction

\(\forall m, n \in \mathbb{Z}\), if \(mn\) is even, then \(m\) is even or \(n\) is even.

Suppose \(\exists m, n \in \mathbb{Z}\) such that \(mn\) is even and both \(m\) and \(n\) are odd. Since \(m\) is odd, \(m = 2k+1\) for some \(k \in \mathbb{Z}\); since \(n\) is odd, \(n = 2\ell + 1\) for some \(\ell \in \mathbb{Z}\) — a new variable, since \(m\) and \(n\) need not be equal. Now \[mn = (2k+1)(2\ell+1) = 4k\ell + 2k + 2\ell + 1 = 2(2k\ell + k + \ell) + 1.\] Since \(k, \ell \in \mathbb{Z}\), \(2k\ell + k + \ell \in \mathbb{Z}\), so \(mn\) is odd. It is impossible for \(mn\) to be both even and odd, a contradiction. Therefore the original statement is true. \(\square\)

\(\forall n \in \mathbb{Z}\), if \(3n^3 - 2\) is odd, then \(n\) is odd.

Suppose the statement is false: \(\exists n \in \mathbb{Z}\) such that \(3n^3 - 2\) is odd and \(n\) is even. Since \(n\) is even, \(n = 2a\) for some \(a \in \mathbb{Z}\). Then \[3n^3 - 2 = 3(2a)^3 - 2 = 24a^3 - 2 = 2(12a^3 - 1).\] Since \(a \in \mathbb{Z}\), \(12a^3 - 1 \in \mathbb{Z}\), so \(3n^3 - 2\) is even — a contradiction. Therefore the original statement is true. \(\square\)

\(\forall m, n \in \mathbb{Z}\), if \(m + n\) is even, then \(m\) and \(n\) are both even or both odd.

Suppose the statement is false. Then \(\exists m, n \in \mathbb{Z}\) such that \(m + n\) is even but exactly one of \(m, n\) is even. Without loss of generality, assume \(m\) is even and \(n\) is odd — the other case (\(m\) odd, \(n\) even) follows because \(m + n = n + m\). Thus \(m = 2a\) for some \(a \in \mathbb{Z}\) and \(n = 2b + 1\) for some \(b \in \mathbb{Z}\). Now \(m + n = 2a + 2b + 1 = 2(a + b) + 1\), and \(a + b \in \mathbb{Z}\), so \(m + n\) is odd. This contradicts the assumption that \(m + n\) is even. Therefore the original statement is true. \(\square\)

For integers \(a\) and \(b\), if \(6a + 3b\) is odd, then \(b\) is odd.

Suppose there exist integers \(a, b\) with \(6a + 3b\) odd and \(b\) even. Then \(b = 2k\) for some \(k \in \mathbb{Z}\), so \(6a + 3b = 6a + 6k = 2(3a + 3k)\), which is even — contradicting that \(6a + 3b\) is odd. \(\square\)

For all positive integers \(x\) and \(y\), \(x^2 - y^2 \neq 1\).

Suppose \(x, y \in \mathbb{Z}^+\) with \(x^2 - y^2 = 1\). Factoring, \(1 = (x-y)(x+y)\). Since \(x, y\) are positive integers we must have \(x - y = 1\) and \(x + y = 1\), so \(2x = 2\) and \(x = 1\); then \(y = 0\), which is not a positive integer — a contradiction. \(\square\)

Proof by contraposition

\(\forall r \in \mathbb{R}\), if \(r^2\) is irrational, then \(r\) is irrational.

We prove the contrapositive: \(\forall r \in \mathbb{R}\), if \(r \in \mathbb{Q}\) then \(r^2 \in \mathbb{Q}\). Suppose \(r \in \mathbb{R}\) with \(r \in \mathbb{Q}\). Then \(r = \frac{a}{b}\) for some \(a, b \in \mathbb{Z}\) with \(b \neq 0\). Now \(r^2 = \frac{a^2}{b^2}\). Since \(a, b \in \mathbb{Z}\) we have \(a^2, b^2 \in \mathbb{Z}\), and \(b^2 \neq 0\) since \(b \neq 0\). So \(r^2 \in \mathbb{Q}\). \(\square\)

\(\forall m, n \in \mathbb{Z}\), if \(mn\) is odd, then \(m\) and \(n\) are both odd.

We prove the contrapositive: \(\forall m, n \in \mathbb{Z}\), if at least one of \(m, n\) is even then \(mn\) is even. Without loss of generality, assume \(m\) is even, so \(m = 2k\) for some \(k \in \mathbb{Z}\). Now \(mn = (2k)n = 2(kn)\), and \(kn \in \mathbb{Z}\), so \(mn\) is even. \(\square\)

For all integers \(a\) and \(b\), if \((ab)^2\) is odd then \(a\) is odd and \(b\) is odd.

Contrapositive: if \(a\) is even or \(b\) is even, then \((ab)^2\) is even. WLOG suppose \(a\) is even, so \(a = 2k\) for some \(k \in \mathbb{Z}\). Then \(ab = 2kb\) and \((ab)^2 = 4k^2b^2 = 2(2k^2b^2)\). Since \(k, b \in \mathbb{Z}\), \(2k^2b^2 \in \mathbb{Z}\), so \((ab)^2\) is even. \(\square\)

For all integers \(x\) and \(y\), if \(x^2(y^2 - 2y)\) is odd then \(x\) and \(y\) are odd.

Contrapositive: if \(x\) is even or \(y\) is even, then \(x^2(y^2 - 2y)\) is even. Here the two cases need separate treatment.

Case 1: \(x\) even. \(x = 2k\) for some \(k \in \mathbb{Z}\), so \(x^2(y^2-2y) = 4k^2(y^2-2y) = 2\big(2k^2(y^2-2y)\big)\), and \(2k^2(y^2-2y) \in \mathbb{Z}\).

Case 2: \(y\) even. \(y = 2\ell\) for some \(\ell \in \mathbb{Z}\), so \(x^2(y^2-2y) = x^2\big(4\ell^2 - 4\ell\big) = 2\big(x^2(2\ell^2 - 2\ell)\big)\), and \(x^2(2\ell^2-2\ell) \in \mathbb{Z}\).

In both cases \(x^2(y^2-2y)\) is even. \(\square\)

For any integer \(n\), \(n^2\) is odd if and only if \(n\) is odd. A biconditional needs both directions, and they want different methods:

(\(\Leftarrow\), direct) Suppose \(n\) is odd, so \(n = 2k+1\). Then \(n^2 = 4k^2 + 4k + 1 = 2(2k^2 + 2k) + 1\) is odd.

(\(\Rightarrow\), contraposition) The contrapositive is: if \(n\) is even then \(n^2\) is even. Suppose \(n = 2k\). Then \(n^2 = 4k^2 = 2(2k^2)\) is even. \(\square\)

This lemma — for all integers \(n\), if \(n^2\) is even then \(n\) is even — is used in the proof that \(\sqrt{2}\) is irrational (see rational-and-irrational-numbers).

Common proof errors

The course explicitly tests spotting these. The seven below are the list given with the practice problems’ §9 Q3 (see practice-problems, where the four proofs you are asked to find the errors in are worked):

  • Arguing from examples — a single example never establishes a universal statement. (“\(\forall x \in \mathbb{R}, \exists y \in \mathbb{R}\) s.t. \(xy = 1\). Proof: if \(x = 10\) take \(y = \frac{1}{10}\).” Not a proof — and the statement is false, since \(x = 0\) has no such \(y\).)
  • Using the same variable to mean two different things — “let \(m = 2k+1\) and \(n = 2k\)” only proves the case \(m = n + 1\). Use a fresh variable for each quantity.
  • Jumping to the conclusion — writing down the definition you need to satisfy and then asserting it holds, without exhibiting the witnesses.
  • Assuming what is to be proved — starting from “\(4a + 2b = 2\ell\)” and manipulating both sides assumes the conclusion. Derive the conclusion; don’t begin with it.
  • Confusion between what is known and what is to be shown.
  • Use of the word any instead of some.
  • Misuse of the word if.

One more, not on that list: forgetting to state your assumptions, or forgetting to say where a variable lives (“for some \(r, s \in \mathbb{Z}\)”). Introduce every existential witness with its domain.

See also

Sequences and Summations

Standard notation for sequences and summations, and the manipulations — change of variable, splitting off the last term — that every induction proof over a sum relies on. See mathematical-induction for the proofs those manipulations feed into.

Notation

A sequence is an ordered list of terms, written in any of these ways:

\[a_1, a_2, a_3, \dots \qquad (a_n)_{n \geq 0} \qquad \{a_n\}_{n=1}^{\infty}\]

These are notations for the same kind of object, not for the same sequence — note that the middle one starts at \(n = 0\) and the other two at \(n = 1\). Always read the bound.

The course uses all of these interchangeably and does not privilege one. A recursively defined sequence may be written \(\{b_n\}_{n \geq 0}\), and the braces may be dropped entirely for a bare \(a_0, a_1, a_2, \dots\). Read whichever form is in front of you; do not expect a house standard.

The subscript is the index; the bound on the index says where the sequence starts (and, if finite, where it stops). A sum over a range of the index is written

\[\sum_{n=1}^{m} a_n\]

Three ways to define a sequence

By an explicit formula that lets you calculate the value of any particular term directly.

\(a_1, a_2, a_3, \dots\) where \(a_n = 2^n - 1\), giving \(1, 3, 7, 15, 31, \dots\)

Recursively, by a rule that calculates a term from the value of one or more preceding terms.

\(\{b_n\}_{n \geq 0}\) where \(b_0 = 3\) and \(b_n = 2 \cdot b_{n-1}\) for \(n \geq 1\), giving \(3, 6, 12, 24, 48, \dots\)

The “or more” matters: the Fibonacci numbers are the standard example of a sequence whose rule reaches back two terms rather than one. A recursive definition needs both a recurrence relation and initial conditions — see recursive-definitions for what each part contributes and how to move between a recursive and an explicit definition.

In words.

For all positive integers \(n\), the sequence in which the \(n^{\text{th}}\) term is the sum of the digits of \(n^2\), giving \(1, 4, 9, 7, 7, 9, \dots\)

(Check: \(1^2 = 1\), \(2^2 = 4\), \(3^2 = 9\), \(4^2 = 16 \to 7\), \(5^2 = 25 \to 7\), \(6^2 = 36 \to 9\).)

A sequence cannot be defined by listing its first few terms

Unless you can list all of them. Initial terms never pin down a sequence — infinitely many sequences agree on any finite prefix.

What is the next term in the sequence with initial terms \(1, 4, 9, 16\)?

Definition Sequence
\(a_n = n^2\) \(1, 4, 9, 16, 25, 36, \dots\)
\(a_n = n^2 \bmod 20\) \(1, 4, 9, 16, 5, 16, 9, \dots\)

Both are legitimate, and they disagree from the fifth term on. The Online Encyclopedia of Integer Sequences (https://oeis.org) returns 397 sequences for a search on \(1, 4, 9, 16\).

Working with sums

Sums over the same index range add and scale termwise:

\[\sum_{i=m}^{n} a_i \pm \sum_{i=m}^{n} b_i = \sum_{i=m}^{n} (a_i \pm b_i) \qquad\qquad \sum_{i=m}^{n} c\,a_i = c \sum_{i=m}^{n} a_i\]

The second is just factoring out the common \(c\) once the sum is written out. The general statement is on the left below and the \(c = 2\) case on the right:

\[c\,a_m + c\,a_{m+1} + \cdots + c\,a_n = c\,(a_m + \cdots + a_n) \qquad\qquad 2a_m + 2a_{m+1} + \cdots + 2a_n = 2(a_m + \cdots + a_n)\]

Note the precondition on the first identity: the two index ranges must match before the terms can be combined. Making them match is what the change of variable below is for.

Change of variable

Substituting a new index shifts the limits and rewrites the general term. To apply \(j = i + 1\) (equivalently \(i = j - 1\)) to \(\displaystyle\sum_{i=0}^{n-1} (2i + 5)\), do three things:

  • lower limit: \(j = 0 + 1 = 1\)
  • upper limit: \(j = (n-1) + 1 = n\)
  • general term: \(2i + 5 = 2(j-1) + 5 = 2j + 3\)

\[\sum_{i=0}^{n-1} (2i+5) = \sum_{j=1}^{n} (2j+3)\]

The sum is unchanged — only its description moved. Figure 7 lays the two descriptions over the terms they describe, which is what makes that claim checkable rather than a rule to trust.

Figure 7: A change of variable relabels, it does not recompute. The boxed row is the list of numbers being added — \(5, 7, 9, \dots, 2n+3\) — and it is identical under both descriptions; only the index written above and below it changes. The three red checks are the three things the substitution has to fix, each attached to the part of the picture it governs.

Splitting off the last term

Writing a sum out and peeling the final term off is the manoeuvre used in almost every induction proof of a summation property, since it is what turns the \(P(k+1)\) side into “the \(P(k)\) side, plus one more term”:

\[\sum_{i=1}^{n+1} (3i+2) = \underbrace{(3 \cdot 1 + 2) + (3 \cdot 2 + 2) + \cdots + (3n + 2)}_{i \;=\; 1 \,\dots\, n} + \underbrace{(3(n+1) + 2)}_{i \;=\; n+1}\]

\[= \sum_{i=1}^{n} (3i+2) \; + \; (3(n+1)+2)\]

The same sum can also be split further back — peeling off the last two terms, or shifting the whole index — which is why a single sum has many equal-looking rewrites.

See also

Strong Induction and the Well-Ordering Principle

Strong mathematical induction, the well-ordering principle for the integers, and the sense in which these two and ordinary induction are all the same tool. See mathematical-induction for ordinary induction and the four-step proof template both methods share.

Strong mathematical induction

Let \(P(n)\) be a predicate written in terms of some integer \(n\). The difference from ordinary induction is entirely in what the inductive step is allowed to assume: every earlier case, not just the immediately preceding one — and, to pay for that, a basis step that establishes several cases rather than one.

Ordinary induction (MI) Strong induction (SMI)
Basis step \(P(2)\) \(P(2) \wedge P(3) \wedge P(4)\)
Inductive step \(\forall k \in \mathbb{Z}\) with \(k \geq 2\), if \(P(k)\) then \(P(k+1)\) \(\forall k \in \mathbb{Z}\) with \(k \geq 4\), if \(P(2) \wedge P(3) \wedge \cdots \wedge P(k)\) then \(P(k+1)\)
Conclusion \(P(n)\) for each integer \(n \geq 2\) \(P(n)\) for each integer \(n \geq 2\)

Both reach the same conclusion by the same kind of chain. For MI, \(P(5)\) follows from \(P(2)\) by three applications of the implication; for SMI with the basis above, \(P(6)\) follows in two.

Figure 8: Mathematical induction and strong mathematical induction as chains of inferences. Each column is an argument — premises stacked, conclusion prefixed by ∴, no rule line. MI climbs one rung at a time from a single base case; SMI starts from three base cases and each step may lean on everything established so far, which is why it lands on P(6) in two moves rather than four. Colour tracks one expression as it moves: green is what the basis step supplies, orange is a statement concluded one column earlier and carried forward. There are no arrows between columns — the repeated colour alone shows where a statement moves.

How many base cases?

As many as the inductive step reaches back for. This is the whole reason the basis step of an SMI proof is not a single case, and the answer to “why two values and not three?”

Count how far back the recurrence reaches, not how many terms it uses — those differ. \(a_n = a_{n-1} - a_{n-3}\) uses only two earlier terms, but the deepest is three positions back, so all three of \(P(0), P(1), P(2)\) must be established or the step could reach for an unproven case.

Figure 9: How many base cases a strong induction needs, on the two sequences worked below: b_n = b_{n-2} + b_{n-1} and a_n = a_{n-1} − a_{n-3}. Both use exactly two earlier terms, so counting terms gives the wrong answer; what settles it is the longest arrow. Reaching two positions back means the basis step must cover two cases; reaching three means three. The terms sit in a row, and each short curved arrow is labelled with what it does.

If the basis step establishes fewer cases than the inductive step consumes, the chain has a hole in it: the first few applications of the inductive step would reach for a case nobody ever proved.

Worked example — a two-term recurrence

Suppose \(b_1, b_2, b_3, \dots\) is defined by \(b_1 = 4\), \(b_2 = 12\), and \(b_n = b_{n-2} + b_{n-1}\) for each integer \(n \geq 3\). Prove that \(4 \mid b_n\) for each integer \(n \geq 1\).

Proof. Let \(P(n)\) be the predicate \(4 \mid b_n\).

Basis step. We know that \(b_1 = 4\) and \(b_2 = 12\). Since \(4 \mid 4\) and \(4 \mid 12\), \(P(1)\) is true and \(P(2)\) is true.

Inductive hypothesis. Suppose \(k \geq 2\) is an integer and that \(P(1), P(2), \dots, P(k)\) are true. Since \(P(k)\) is true, \(b_k = 4x\) for some \(x \in \mathbb{Z}\). Since \(P(k-1)\) is true, \(b_{k-1} = 4y\) for some \(y \in \mathbb{Z}\).

Inductive step. Consider \(n = k+1\):

\[b_{k+1} = b_{k-1} + b_k = 4y + 4x \qquad \text{(by the inductive hypothesis)}\] \[= 4(y+x)\]

Thus \(4 \mid b_{k+1}\), so \(P(k+1)\) is true.

Therefore, by PSMI, for each integer \(n \geq 1\), \(4 \mid b_n\). \(\square\)

Note where strong induction actually earns its keep: the step uses \(P(k)\) and \(P(k-1)\). Ordinary induction only hands you \(P(k)\), which is not enough to say anything about \(b_{k-1}\).

Worked example — every integer \(n > 1\) is a product of primes

Theorem. Every integer \(n > 1\) can be written as a product of primes.

Proof. Let \(P(n)\) be the predicate \(n\) can be written as a product of primes, for \(n \geq 2\).

Basis step. \(n = 2\) is prime and thus is a product of primes, so \(P(2)\) is true.

Inductive step. Suppose that for some integer \(k \geq 2\), every integer \(2, 3, \dots, k\) can be written as a product of primes — that is, \(P(2) \wedge P(3) \wedge \cdots \wedge P(k)\). We now need to prove that \(k+1\) is a product of primes. Since \(k + 1 > 1\), either \(k+1\) is prime or \(k+1\) is composite.

  • Case 1: \(k+1\) is prime. Then \(k+1\) is a product of primes.
  • Case 2: \(k+1\) is composite. Then \(k+1 = rs\) for some positive integers \(r \neq 1\) and \(s \neq 1\). This implies \(1 < r < k+1\) and \(1 < s < k+1\). So by the inductive hypothesis — specifically \(P(r)\) and \(P(s)\) — both \(r\) and \(s\) are products of primes. Let \(r = p_1p_2\dots p_a\) and \(s = q_1q_2\dots q_b\) where the \(p_i\) and \(q_j\) are primes; thus \(k+1 = rs = p_1p_2\dots p_a q_1q_2 \dots q_b\) is a product of primes.

In either case, \(k+1\) is a product of primes.

Therefore, by SMI, every integer \(n > 1\) can be written as a product of primes. \(\square\)

Why strong induction and not ordinary induction? Because \(r\) and \(s\) can be any integers from 2 to \(k\) — the factorisation gives no control over which — so the proof needs the inductive hypothesis available at all of those values, not just at \(k\).

The well-ordering principle for the integers

Let \(S\) be a non-empty set of integers, each of which is greater than some fixed integer. Then \(S\) has a least element.

Four pieces of terminology, all load-bearing:

  • non-empty — the empty set has no least element, trivially.
  • of integers — not of rationals or reals. This clause is not decoration: the positive rationals are non-empty and bounded below by 0, and still have no least element, because between any candidate and 0 there is always another one (halve it). See 2026-08-27-strong-induction-and-well-ordering.
  • greater than some fixed integer — \(S\) must be bounded below. Without this the principle fails: \(\mathbb{Z}^-\) is a non-empty set of integers with no least element.
  • a least element — an element of \(S\) that is \(\leq\) every element of \(S\). It lives inside the set, not outside it.

Example — integers congruent to \(1 \bmod 4\)

True or false? The set of all integers congruent to \(1 \bmod 4\) has a least element.

False. The set is \(\{\dots, -7, -3, 1, 5, 9, \dots\}\).

Suppose this set did have a least element, call it \(x\). Then, since \(x \equiv 1 \pmod 4\), we have \(x - 4 \equiv 1 \pmod 4\), so \(x-4\) is also in the set. But \(x - 4\) is smaller than \(x\) — a contradiction. \(\square\)

The set is non-empty, but it fails the “bounded below” condition, so the well-ordering principle simply does not apply to it.

MI, SMI and the WOP are equivalent

Anything provable by one is provable by the other two. Two of the translations:

SMI \(\to\) MI: reword the statement. A strong-induction proof of \(P(n)\) becomes an ordinary-induction proof of the accumulated predicate

\[Q(n) \;\equiv\; P(0) \wedge P(1) \wedge \cdots \wedge P(n)\]

Assuming \(Q(k)\) in ordinary induction gives you exactly what assuming \(P(0), \dots, P(k)\) gave you in strong induction. (Epp, Q27 p. 279.)

MI \(\to\) WOP: argue by contradiction. Suppose the set of integers \(\geq a\) for which \(P(n)\) is false is non-empty. It is bounded below by \(a\), so by the WOP it has a least element; derive a contradiction from that least element. (Epp, Q31 p. 279.)

The same theorem, three ways

Suppose \(a_0, a_1, a_2, \dots\) is defined by \(a_0 = 6\), \(a_1 = 9\), \(a_2 = 12\) and \(a_n = a_{n-1} - a_{n-3}\) for each integer \(n \geq 3\). Prove that for all integers \(n \geq 0\), \(3 \mid a_n\).

By SMI.

Let \(P(n)\) be the predicate \(3 \mid a_n\).

Basis step. \(a_0 = 6\), \(a_1 = 9\) and \(a_2 = 12\). Since \(3 \mid 6\), \(3 \mid 9\) and \(3 \mid 12\), \(P(0)\), \(P(1)\) and \(P(2)\) are true.

Inductive hypothesis. Suppose \(k \geq 2\) is an integer and that \(P(0), P(1), \dots, P(k)\) are true. Since \(P(k-2)\) is true, \(a_{k-2} = 3y\) for some \(y \in \mathbb{Z}\); since \(P(k)\) is true, \(a_k = 3x\) for some \(x \in \mathbb{Z}\).

Inductive step. Consider \(n = k+1\): \(a_{k+1} = a_k - a_{k-2} = 3x - 3y = 3(x-y)\). Thus \(3 \mid a_{k+1}\), so \(P(k+1)\) is true.

Therefore, by PSMI, for each integer \(n \geq 0\), \(3 \mid a_n\). \(\square\)

By MI, on the reworded predicate.

Let \(Q(n)\) be the predicate \(3 \mid a_j\) for all integers \(j\) where \(0 \leq j \leq n\) — that is, \(Q(n) \equiv P(0) \wedge P(1) \wedge \cdots \wedge P(n)\).

Basis step. Consider \(n = 2\). Since \(3 \mid a_0\), \(3 \mid a_1\) and \(3 \mid a_2\), \(Q(2)\) is true.

Inductive hypothesis. Suppose \(k \geq 2\) is an integer and that \(Q(k)\) is true. Since \(Q(k)\) is true and \(k \geq 2\), \(a_k = 3x\) and \(a_{k-2} = 3y\) for some \(x, y \in \mathbb{Z}\).

Inductive step. Consider \(n = k+1\): \(a_{k+1} = a_k - a_{k-2} = 3x - 3y = 3(x-y)\). Thus \(3 \mid a_{k+1}\). Since \(Q(k)\) already gives \(3 \mid a_i\) for every \(0 \leq i \leq k\), adding \(3 \mid a_{k+1}\) gives \(Q(k+1)\).

(That last step is the point of rewording the predicate: \(Q(k+1)\) is a conjunction over all indices up to \(k+1\), so it needs \(Q(k)\) and the new case — \(3 \mid a_{k+1}\) alone would not establish it.)

Therefore, by PMI, for each integer \(n \geq 0\), \(3 \mid a_n\). \(\square\)

The two proofs have the same algebraic core: the line \(a_{k+1} = a_k - a_{k-2} = 3x - 3y = 3(x-y)\) is identical in each. All the reworded predicate does is let a single assumption, \(Q(k)\), carry the several earlier facts the algebra needs.

Two details of the MI proof are easy to get wrong:

  • The basis is \(n = 2\), not \(n = 0\). \(Q(2)\) already asserts \(3 \mid a_0\), \(3 \mid a_1\) and \(3 \mid a_2\), so verifying \(Q(2)\) discharges all three of the SMI proof’s base cases in one statement. Three base cases have not disappeared; they have been packed into one predicate.
  • The step concludes \(Q(k+1)\), not \(P(k+1)\). An ordinary induction on \(Q\) has to deliver \(Q(k+1)\), and \(3 \mid a_{k+1}\) on its own does not — it needs \(Q(k)\) alongside it.

The conjunction in \(Q\) starts at \(P(0)\), not \(P(1)\). The predicate is \(3 \mid a_j\) for all integers \(j\) where \(0 \leq j \leq n\), so \(Q(2) \equiv P(0) \wedge P(1) \wedge P(2)\) — which is also what the basis step checks.

By the WOP, by contradiction.

Suppose the statement is false, so there exists an integer \(n \geq 0\) such that 3 does not divide \(a_n\). Let

\[A = \{\, n \in \mathbb{Z}^{\geq 0} \;:\; 3 \nmid a_n \,\}\]

By the WOP and our assumption, \(A\) has a least element — the smallest \(n\) for which \(3 \nmid a_n\). Call it \(k\). Thus \(3 \nmid a_k\), but \(3 \mid a_i\) for all \(0 \leq i \leq k-1\).

Note that \(k \geq 3\), since \(a_0\), \(a_1\) and \(a_2\) are all divisible by 3.

Thus \(a_{k-1} = 3x\) and \(a_{k-3} = 3y\) for some \(x, y \in \mathbb{Z}\), so

\[a_k = a_{k-1} - a_{k-3} = 3x - 3y = 3(x-y)\]

Thus \(3 \mid a_k\), which is a contradiction.

Therefore, for each integer \(n \geq 0\), \(3 \mid a_n\). \(\square\)

Reading the WOP proof next to the SMI one: “let \(k\) be the least counterexample” plays the role of the inductive hypothesis (everything below \(k\) works), and “\(3 \mid a_k\), contradiction” plays the role of the inductive step. Drawn on the index line, the whole proof is one picture:

Figure 10: The well-ordering proof of 3 | a_n as a picture of its own least counterexample. Everything to the left of k is green not because it was proved, but because k was chosen least — that single step is the only place the WOP is used, and it is exactly what the strong inductive hypothesis supplies in the SMI proof. The least element here is the smallest number that gives a contradiction. k ≥ 3 is what guarantees both arrows land inside the green stretch, since a_{k-3} has to be a term of the sequence at all.

Note which term the WOP proof reaches for. The SMI proof used \(a_k\) and \(a_{k-2}\) to build \(a_{k+1}\); the WOP proof is standing at \(k\) rather than at \(k+1\), so the same recurrence reads \(a_k = a_{k-1} - a_{k-3}\) and it reaches for \(a_{k-1}\) and \(a_{k-3}\). Same recurrence, index shifted by one — and that shift is why the “\(k \geq 3\)” line has to be there.

Spotting a WOP proof

A proof is using the well-ordering principle whenever it says “choose the smallest such \(n\)” about a set it has just argued is non-empty. That phrase is not free — it is the WOP being invoked, and it needs the set to be non-empty and bounded below.

For instance, the primes theorem above can be rewritten as a contradiction proof: suppose some integer \(n > 1\) is not a product of primes; by the WOP, choose the smallest such \(n\); then \(n\) is prime (done) or \(n = rs\) with \(1 < r, s < n\), and both \(r\) and \(s\) are products of primes precisely because \(n\) was chosen smallest — contradicting the choice of \(n\). Structurally it is the SMI proof with the induction turned inside out.

See also