• Aucun résultat trouvé

Concepts in Abstract Mathematics

N/A
N/A
Protected

Academic year: 2022

Partager "Concepts in Abstract Mathematics"

Copied!
7
0
0

Texte intégral

(1)

Concepts in Abstract Mathematics

Homework questions – Week 10

Jean-Baptiste Campesato March 29th, 2021 to April 2nd, 2021 Exercise 1.

Let𝐸be a set.

1. Prove that∀𝐴, 𝐵, 𝐶 ∈ 𝒫(𝐸), 𝐴 ∪ 𝐵 = 𝐵 ∩ 𝐶 ⟹ 𝐴 ⊂ 𝐵 ⊂ 𝐶.

2. Prove that∀𝐴, 𝐵 ∈ 𝒫(𝐸), 𝐴 ∩ 𝐵 = 𝐴 ∪ 𝐵 ⟹ 𝐴 = 𝐵.

Exercise 2.

Let𝑓 ∶ 𝐴 → 𝐵,𝑔 ∶ 𝐵 → 𝐶 andℎ ∶ 𝐶 → 𝐷be three functions.

Prove that𝑔 ∘ 𝑓 andℎ ∘ 𝑔are bijective if and only if𝑓,𝑔andℎare bijective.

Exercise 3.

Let𝑓 ∶ 𝐸 → 𝐹.

1. Prove that∀𝐴 ∈ 𝒫(𝐸), 𝐴 ⊂ 𝑓−1(𝑓 (𝐴)).

2. Prove that∀𝐵 ∈ 𝒫(𝐹 ), 𝑓 (𝑓−1(𝐵)) ⊂ 𝐵.

3. Can these inclusions be strict?

Exercise 4.

Let𝑓 ∶ 𝐸 → 𝐹.

1. Prove that∀𝐴, 𝐵 ∈ 𝒫(𝐹 ), 𝐴 ⊂ 𝐵 ⟹ 𝑓−1(𝐴) ⊂ 𝑓−1(𝐵).

Does the converse hold?

2. Prove that∀𝐴, 𝐵 ∈ 𝒫(𝐹 ), 𝑓−1(𝐴 ∩ 𝐵) = 𝑓−1(𝐴) ∩ 𝑓−1(𝐵).

3. Prove that∀𝐴, 𝐵 ∈ 𝒫(𝐹 ), 𝑓−1(𝐴 ∪ 𝐵) = 𝑓−1(𝐴) ∪ 𝑓−1(𝐵).

Exercise 5.

Let𝑓 ∶ 𝐸 → 𝐹.

1. Prove that∀𝐴, 𝐵 ∈ 𝒫(𝐸), 𝐴 ⊂ 𝐵 ⟹ 𝑓 (𝐴) ⊂ 𝑓 (𝐵).

Does the converse hold?

2. Prove that∀𝐴, 𝐵 ∈ 𝒫(𝐸), 𝑓 (𝐴 ∩ 𝐵) ⊂ 𝑓 (𝐴) ∩ 𝑓 (𝐵).

Can the inclusion be strict?

3. Prove that∀𝐴, 𝐵 ∈ 𝒫(𝐸), 𝑓 (𝐴 ∪ 𝐵) = 𝑓 (𝐴) ∪ 𝑓 (𝐵).

Exercise 6.

Let𝑓 ∶ 𝐸 → 𝐹. Prove that𝑓 is injective if and only if∀𝐴, 𝐵 ∈ 𝒫(𝐸), 𝑓 (𝐴 ∩ 𝐵) = 𝑓 (𝐴) ∩ 𝑓 (𝐵).

Exercise 7.

Let𝐸be a finite set. For𝐴, 𝐵 ∈ 𝒫(𝐸)we define the symmetric difference of𝐴and𝐵by𝐴Δ𝐵 = (𝐴∪𝐵)⧵(𝐴∩𝐵).

Prove that∀𝐴, 𝐵 ∈ 𝒫(𝐸), |𝐴Δ𝐵| = |𝐴| + |𝐵| − 2|𝐴 ∩ 𝐵|.

Exercise 8.

Let𝐸and𝐹 be two finite sets.

1. Prove that𝐹𝐸 (the set of functions𝐸 → 𝐹) is finite and express|𝐹𝐸|in terms of|𝐸|and|𝐹 |.

2. Prove that the set{𝑓 ∈ 𝐸𝐹 ∶ 𝑓 is injective}is finite and express its cardinal in terms of|𝐸|and|𝐹 |.

3. Prove that the set{𝑓 ∈ 𝐸𝐸 ∶ 𝑓 is bijective}is finite and express its cardinal in terms of|𝐸|.

The case of surjective functions is more tricky.

(2)

Let𝐸be a finite set and𝑘 ∈ {0, 1, … , |𝐸|}. What is the cardinal of{𝐴 ∈ 𝒫(𝐸) ∶ |𝐴| = 𝑘}?

Exercise 10.

Prove that a set𝐸is finite if and only if𝒫(𝐸)is finite.

In this case, give an expression of|𝒫(𝐸)|in terms of|𝐸|.

Exercise 11. The pigeonhole principle or Dirichlet’s drawer principle I had no enough time to cover this topic in lectures, so here it is :-).

1. Let𝐸and𝐹 be two finite sets. Prove that|𝐸| ≤ |𝐹 |if and only if there exists an injection𝑓 ∶ 𝐸 → 𝐹. 2. Let𝐸and𝐹 be two finite sets. Prove that if|𝐸| > |𝐹 |then there is no injective function𝐸 → 𝐹.

This statement is pigeonhole principle or Dirichlet’s drawer principle: if you have𝑛elements put in𝑘 < 𝑛boxes, then at least one box contains two elements.

3. During a post-covid party with𝑛 > 1participants, we may always find two people who shook hands to the same number of people.

4. Let𝑛 ∈ ℕ ⧵ {0}. Let𝑎1, 𝑎2, … , 𝑎𝑛 ∈ ℤ. Prove that there exists distinct𝑖1, … , 𝑖𝑟 ∈ {1, … , 𝑛}, 𝑟 ≥ 1,so that𝑛| ∑𝑟𝑘=1𝑎𝑖

𝑘.

5. Prove that among 13 distinct real numbers, there always exist two𝑥, 𝑦satisfying0 < 𝑥 − 𝑦

1 + 𝑥𝑦 < 2 − √3.

Hint: it looks like a trigonometric formula you know!

(3)

Sample solutions to Exercise 1.

1. Let𝐴, 𝐵, 𝐶 ∈ 𝒫(𝐸). Assume that𝐴 ∪ 𝐵 = 𝐵 ∩ 𝐶.

Let𝑥 ∈ 𝐴then𝑥 ∈ 𝐴 ∪ 𝐵 = 𝐵 ∩ 𝐶. Therefore𝑥 ∈ 𝐵. So𝐴 ⊂ 𝐵. Let𝑥 ∈ 𝐵then𝑥 ∈ 𝐴 ∪ 𝐵 = 𝐵 ∩ 𝐶. Therefore𝑥 ∈ 𝐶. So𝐵 ⊂ 𝐶. 2. Using the previous question.

Let𝐴, 𝐵 ∈ 𝒫(𝐸). Assume that𝐴 ∩ 𝐵 = 𝐴 ∪ 𝐵.

From the previous question we get that𝐴 ⊂ 𝐵 ⊂ 𝐴. Hence𝐴 = 𝐵.

Direct proof.

Let𝐴, 𝐵 ∈ 𝒫(𝐸). Assume that𝐴 ∩ 𝐵 = 𝐴 ∪ 𝐵.

Let𝑥 ∈ 𝐴then𝑥 ∈ 𝐴 ∪ 𝐵 = 𝐴 ∩ 𝐵. Thus𝑥 ∈ 𝐵. Therefore𝐴 ⊂ 𝐵.

Let𝑥 ∈ 𝐵then𝑥 ∈ 𝐴 ∪ 𝐵 = 𝐴 ∩ 𝐵. Thus𝑥 ∈ 𝐴. Therefore𝐵 ⊂ 𝐴.

Hence𝐴 = 𝐵.

Proof by contrapositive.

Let𝐴, 𝐵 ∈ 𝒫(𝐸). Assume that𝐴 ≠ 𝐵. Then

• either𝐴 ⧵ 𝐵 ≠ ∅and then there exists𝑥 ∈ 𝐸 such that𝑥 ∈ 𝐴and𝑥 ∉ 𝐵. Thus𝑥 ∈ 𝐴 ∪ 𝐵 but 𝑥 ∉ 𝐴 ∩ 𝐵. Therefore𝐴 ∩ 𝐵 ≠ 𝐴 ∪ 𝐵.

• or𝐵 ⧵𝐴 ≠ ∅and then there exists𝑥 ∈ 𝐸such that𝑥 ∈ 𝐵and𝑥 ∉ 𝐴. Thus𝑥 ∈ 𝐴∪𝐵but𝑥 ∉ 𝐴∩𝐵.

Therefore𝐴 ∩ 𝐵 ≠ 𝐴 ∪ 𝐵.

Sample solutions to Exercise 2.

⇐Assume that𝑓,𝑔andℎare bijective then𝑔 ∘ 𝑓 andℎ ∘ 𝑔are too.

⇒Assume that𝑔 ∘ 𝑓andℎ ∘ 𝑔are bijective.

Since𝑔 ∘ 𝑓 is surjective,𝑔is too. Sinceℎ ∘ 𝑔is injective,𝑔is too.

Hence𝑔is bijective, so it admits an inverse𝑔−1∶ 𝐶 → 𝐵.

Then𝑓 = 𝑔−1∘ (𝑔 ∘ 𝑓 )andℎ = (ℎ ∘ 𝑔) ∘ 𝑔−1are bijective as composition of bijective functions.

Sample solutions to Exercise 3.

1. Let𝐴 ∈ 𝒫(𝐸). Let𝑥 ∈ 𝐴. Then𝑓 (𝑥) ∈ 𝑓 (𝐴). Therefore𝑥 ∈ 𝑓−1(𝑓 (𝐴)).

We proved that𝐴 ⊂ 𝑓−1(𝑓 (𝐴)).

2. Let 𝐵 ⊂ 𝒫(𝐹 ). Let𝑦 ∈ 𝑓 (𝑓−1(𝐵)). Then there exists 𝑥 ∈ 𝑓−1(𝐵) such that 𝑦 = 𝑓 (𝑥). But since 𝑥 ∈ 𝑓−1(𝐵),𝑦 = 𝑓 (𝑥) ∈ 𝐵.

We proved that𝑓 (𝑓−1(𝐵)) ⊂ 𝐵.

3. Define

𝑓 ∶⎧⎪

⎨⎪

{1, 2} → {1, 2}

1 ↦ 1

2 ↦ 1

Then𝑓 (𝑓−1({1, 2})) = 𝑓 ({1, 2}) = {1} ⊊ {1, 2}.

And𝑓−1(𝑓 ({1})) = 𝑓−1({1}) = {1, 2} ⊋ {1}.

Sample solutions to Exercise 4.

1. Let𝐴, 𝐵 ∈ 𝒫(𝐹 )be such that𝐴 ⊂ 𝐵. Take𝑥 ∈ 𝑓−1(𝐴). Then𝑓 (𝑥) ∈ 𝐴 ⊂ 𝐵. Thus𝑥 ∈ 𝑓−1(𝐵).

The converse doesn’t hold. Indeed, define 𝑓 ∶{

{1} → {1, 2}

1 ↦ 1

then𝑓−1({2}) = ∅ ⊂ 𝑓−1({1}). But{2} ⊄ {1}.

(4)

2. Let𝐴, 𝐵 ∈ 𝒫(𝐹 ).

Since𝐴 ∩ 𝐵 ⊂ 𝐴and𝐴 ∩ 𝐵 ⊂ 𝐵, we get that𝑓−1(𝐴 ∩ 𝐵) ⊂ 𝑓−1(𝐴)and𝑓−1(𝐴 ∩ 𝐵) ⊂ 𝑓−1(𝐵). Thus 𝑓−1(𝐴 ∩ 𝐵) ⊂ 𝑓−1(𝐴) ∩ 𝑓−1(𝐵).

For the other inclusion, let𝑥 ∈ 𝑓−1(𝐴) ∩ 𝑓−1(𝐵). Then𝑓 (𝑥) ∈ 𝐴and𝑓 (𝑥) ∈ 𝐵. Thus𝑓 (𝑥) ∈ 𝐴 ∩ 𝐵so that𝑥 ∈ 𝑓−1(𝐴 ∩ 𝐵). Thus𝑓−1(𝐴) ∩ 𝑓−1(𝐵) ⊂ 𝑓−1(𝐴 ∩ 𝐵).

3. Let𝐴, 𝐵 ∈ 𝒫(𝐹 ).

Since𝐴, 𝐵 ⊂ 𝐴 ∪ 𝐵, we get𝑓−1(𝐴), 𝑓−1(𝐵) ⊂ 𝑓−1(𝐴 ∪ 𝐵). Thus𝑓−1(𝐴) ∪ 𝑓−1(𝐵) ⊂ 𝑓−1(𝐴 ∪ 𝐵).

For the other inclusion, let𝑥 ∈ 𝑓−1(𝐴 ∪ 𝐵). Then𝑓 (𝑥) ∈ 𝐴 ∪ 𝐵. Either𝑓 (𝑥) ∈ 𝐴and then𝑥 ∈ 𝑓−1(𝐴) ⊂ 𝑓−1(𝐴) ∪ 𝑓−1(𝐵)or𝑓 (𝑥) ∈ 𝐵and then𝑥 ∈ 𝑓−1(𝐵) ⊂ 𝑓−1(𝐴) ∪ 𝑓−1(𝐵). Thus𝑥 ∈ 𝑓−1(𝐴) ∪ 𝑓−1(𝐵). We proved that𝑓−1(𝐴 ∪ 𝐵) ⊂ 𝑓−1(𝐴) ∪ 𝑓−1(𝐵).

Sample solutions to Exercise 5.

1. Let𝐴, 𝐵 ∈ 𝒫(𝐸)be such that𝐴 ⊂ 𝐵. Let𝑦 ∈ 𝑓 (𝐴). Then𝑦 = 𝑓 (𝑥)for some𝑥 ∈ 𝐴. But𝑥 ∈ 𝐴 ⊂ 𝐵.

Thus𝑦 = 𝑓 (𝑥) ∈ 𝑓 (𝐵). We proved that𝑓 (𝐴) ⊂ 𝑓 (𝐵).

The converse doesn’t hold. Indeed define

𝑓 ∶⎧⎪

⎨⎪

{1, 2} → {1}

1 ↦ 1

2 ↦ 1

then𝑓 ({1}) = 𝑓 ({2}) = {1}but{1} ⊄ {2}.

2. Let𝐴, 𝐵 ∈ 𝒫(𝐸). Since𝐴 ∩ 𝐵 ⊂ 𝐴, 𝐵we get𝑓 (𝐴 ∩ 𝐵) ⊂ 𝑓 (𝐴), 𝑓 (𝐵). Thus𝑓 (𝐴 ∩ 𝐵) ⊂ 𝑓 (𝐴) ∩ 𝑓 (𝐵).

The inclusion can be strict using the same example as above.

3. Let𝐴, 𝐵 ∈ 𝒫(𝐸). Since𝐴, 𝐵 ⊂ 𝐴 ∪ 𝐵, we get𝑓 (𝐴), 𝑓 (𝐴) ⊂ 𝑓 (𝐴 ∪ 𝐵). Thus𝑓 (𝐴) ∪ 𝑓 (𝐵) ⊂ 𝑓 (𝐴 ∪ 𝐵).

For the other inclusion, let𝑦 ∈ 𝑓 (𝐴 ∪ 𝐵). Then𝑦 = 𝑓 (𝑥)for some𝑥 ∈ 𝐴 ∪ 𝐵. So either𝑥 ∈ 𝐴and then 𝑦 = 𝑓 (𝑥) ∈ 𝑓 (𝐴) ⊂ 𝑓 (𝐴) ∪ 𝑓 (𝐵), or𝑥 ∈ 𝐵 and then𝑦 = 𝑓 (𝑥) ∈ 𝑓 (𝐵) ⊂ 𝑓 (𝐴) ∪ 𝑓 (𝐵). In both cases 𝑦 ∈ 𝑓 (𝐴) ∪ 𝑓 (𝐵). So we proved that𝑓 (𝐴 ∪ 𝐵) ⊂ 𝑓 (𝐴) ∪ 𝑓 (𝐵).

Sample solutions to Exercise 6.

⇒Assume that𝑓 is injective. Let𝐴, 𝐵 ∈ 𝒫(𝐸).

We already know that𝑓 (𝐴 ∩ 𝐵) ⊂ 𝑓 (𝐴) ∩ 𝑓 (𝐵)holds (see the previous exercise).

Let’s prove that𝑓 (𝐴) ∩ 𝑓 (𝐵) ⊂ 𝑓 (𝐴 ∩ 𝐵).

Let𝑦 ∈ 𝑓 (𝐴) ∩ 𝑓 (𝐵). Then𝑦 = 𝑓 (𝑥1)for some𝑥1 ∈ 𝐴and𝑦 = 𝑓 (𝑥2)for some𝑥2 ∈ 𝐵.

Since𝑓 (𝑥1) = 𝑓 (𝑥2)and𝑓 is injective, we obtain that𝑥1= 𝑥2∈ 𝐴 ∩ 𝐵. Therefore𝑦 = 𝑓 (𝑥1) ∈ 𝑓 (𝐴 ∩ 𝐵).

We proved that𝑓 (𝐴) ∩ 𝑓 (𝐵) ⊂ 𝑓 (𝐴 ∩ 𝐵). Thus𝑓 (𝐴) ∩ 𝑓 (𝐵) = 𝑓 (𝐴 ∩ 𝐵).

⇐Assume that∀𝐴, 𝐵 ∈ 𝒫(𝐸), 𝑓 (𝐴 ∩ 𝐵) = 𝑓 (𝐴) ∩ 𝑓 (𝐵).

Let𝑥1, 𝑥2∈ 𝐸 be such that𝑓 (𝑥1) = 𝑓 (𝑥2). Set𝑦 ≔ 𝑓 (𝑥1) = 𝑓 (𝑥2).

Then𝑓 ({𝑥1} ∩ {𝑥2}) = 𝑓 ({𝑥1}) ∩ 𝑓 ({𝑥2}) = {𝑦} ∩ {𝑦} = {𝑦}.

Particularly{𝑥1} ∩ {𝑥2} ≠ ∅, thus𝑥1= 𝑥2. Sample solutions to Exercise 7.

|𝐴Δ𝐵| = |(𝐴 ∪ 𝐵) ⧵ (𝐴 ∩ 𝐵)|

= |𝐴 ∪ 𝐵| − |𝐴 ∩ 𝐵|

= |𝐴| + |𝐵| − |𝐴 ∩ 𝐵| − |𝐴 ∩ 𝐵|

= |𝐴| + |𝐵| − 2|𝐴 ∩ 𝐵|

(5)

Sample solutions to Exercise 8.

1. Let𝜑 ∶ {𝑘 ∈ ℕ ∶ 𝑘 < |𝐸|} → 𝐸.

Then𝜓 ∶ 𝐹𝐸 → 𝐹|𝐸|defined by𝜓(𝑓 ) = (𝑓 (𝜑(0)), … , 𝑓 (𝜑(|𝐸| − 1)))is a bijection (prove it).

Therefore|𝐹𝐸| = |𝐹|𝐸|| = |𝐹 ||𝐸|.

2. According to the last exercise, there exists an injective function𝐸 → 𝐹 if and only if|𝐸| ≤ |𝐹 |.

Next, since𝐸is finite, there exists a bijection𝜑 ∶ {𝑘 ∈ ℕ ∶ 𝑘 < |𝐸|} → 𝐸.

For𝑓 (𝜑(0))we have|𝐹 |possible choices. For𝑓 (𝜑(1))we have|𝐹 ⧵ {𝑓 (𝜑(0))}| = |𝐹 | − 1choices. For 𝑓 (𝜑(2))we have|𝐹 ⧵ {𝑓 (𝜑(0)), 𝑓 (𝜑(1))}| = |𝐹 | − 2choices. And so on.

Therefore,|{𝑓 ∈ 𝐸𝐹 ∶ 𝑓 is injective}| = |𝐹 |(|𝐹 | − 1) ⋯ (|𝐹 | − |𝐸| + 1) = (|𝐹 |−|𝐸|)!|𝐹 |! . Thus|{𝑓 ∈ 𝐸𝐹 ∶ 𝑓is injective}| =

{

0 if|𝐸| > |𝐹 |

|𝐹 |!

(|𝐹 |−|𝐸|)! if|𝐸| ≤ |𝐹 | .

3. It a special case of the above question when|𝐸| = |𝐹 |:|{𝑓 ∈ 𝐸𝐸 ∶ 𝑓 is bijective}| = |𝐸|!

(|𝐸|−|𝐸|)! = |𝐸|!.

Sample solutions to Exercise 9.

The number of subsets with cardinality𝑘included in a set of cardinality𝑛is denoted(𝑛𝑘)read ”𝑛choose 𝑘”.

We are going to prove that(𝑛𝑘) = (𝑛−𝑘)!𝑘!𝑛! .

Let𝐸a finite set. Set𝑛 = |𝐸|. Fix𝑘 ∈ {0, 1, … , 𝑛}.

An ordered list of𝑘distinct elements is the same as fixing an injection{0, 1, … , 𝑘 − 1} → 𝐸. So, using the previous question there are 𝑛!

(𝑛−𝑘)! such ordered lists.

Two ordered lists of 𝑘 elements give the same subset if and only if one is obtained from the other one permuting its elements, which is the same as constructing a bijection {0, 1, … , 𝑘 − 1} → {0, 1, … , 𝑘 − 1}.

From the previous question there are𝑘!such bijections.

Therefore(𝑛𝑘) =

𝑛!

(𝑛−𝑘)!

𝑘! = 𝑛!

(𝑛−𝑘)!𝑘!. Sample solutions to Exercise 10.

Method 1 (by induction):

Let’s prove by induction on𝑛 = |𝐸|that𝒫(𝐸)is finite and that|𝒫(𝐸)| = 2|𝐸|.

Base case at𝑛 = 0:if𝐸 = ∅then𝒫(𝐸) = {∅}is finite.

Induction step:assume that the statement holds for some𝑛 ∈ ℕ, i.e. if𝐸is a set with|𝐸| = 𝑛then𝒫(𝐸) is finite and|𝒫(𝐸)| = 2𝑛.

Let𝐸be a set such that|𝐸| = 𝑛 + 1. Since|𝐸| > 0, there exists𝑥 ∈ 𝐸.

By the induction hypothesis, since|𝐸 ⧵ {𝑥}| = 𝑛, we get that𝒫(𝐸 ⧵ {𝑥})is finite and|𝒫(𝐸 ⧵ {𝑥})| = 2𝑛. Note that𝒫(𝐸 ⧵ {𝑥}) = {𝐴 ∈ 𝒫(𝐸) ∶ 𝑥 ∉ 𝐴}and that

{𝐴 ∈ 𝒫(𝐸) ∶ 𝑥 ∉ 𝐴} → {𝐴 ∈ 𝒫(𝐸) ∶ 𝑥 ∈ 𝐴}

𝐴 ↦ 𝐴 ∪ {𝑥}

is a bijection.

Therefore𝒫(𝐸) = {𝐴 ∈ 𝒫(𝐸) ∶ 𝑥 ∉ 𝐴} ⊔ {𝐴 ∈ 𝒫(𝐸) ∶ 𝑥 ∈ 𝐴}is finite and

|𝒫(𝐸)| = |{𝐴 ∈ 𝒫(𝐸) ∶ 𝑥 ∉ 𝐴}| + |{𝐴 ∈ 𝒫(𝐸) ∶ 𝑥 ∈ 𝐴}| = 2𝑛+ 2𝑛= 2𝑛+1.

(6)

Method 2 (using the previous exercise):

Let𝐸be a finite set. We know that for𝑘 = 0, … , |𝐸|, the number of subsets with𝑘elements is(𝑘𝑛). Therefore the number of subsets included in𝐸is

|𝒫(𝐸)| =

|𝐸|

𝑘=0(𝑛 𝑘)=

|𝐸|

𝑘=0(𝑛

𝑘)1𝑘1|𝐸|−𝑘 = (1 + 1)|𝐸| = 2|𝐸|

Method 3 (which generalizes to infinite sets):

Let𝐸be a finite set. We define𝜓 ∶ 𝒫(𝐸) → {0, 1}𝐸 by𝜓(𝐴)(𝑥) = {

1 if𝑥 ∈ 𝐴 0 otherwise .

Then𝜓is a bijection thus𝒫(𝐸)is finite since{0, 1}𝐸is and moreover|𝒫(𝐸)| = |{0, 1}𝐸| = 2|𝐸|.

⇐Let𝐸be a set. Assume that𝒫(𝐸)is finite.

Note thatΦ ∶ 𝐸 → 𝒫(𝐸)defined byΦ(𝑥) = {𝑥}is injective. Therefore𝐸is finite too.

Sample solutions to Exercise 11.

1. ⇒Assume that|𝐸| ≤ |𝐹 |.

There exist bijections𝜑 ∶ {𝑘 ∈ ℕ ∶ 𝑘 < |𝐸|} → 𝐸and𝜓 ∶ {𝑘 ∈ ℕ ∶ 𝑘 < |𝐹 |} → 𝐹. Since|𝐸| ≤ |𝐹 |,𝑓 = 𝜓 ∘ 𝜑−1∶ 𝐸 → 𝐹 is well-defined and injective.

⇒Assume that there exists an injection𝑓 ∶ 𝐸 → 𝐹.

Then𝑓 induces a bijection𝑓 ∶ 𝐸 → 𝑓 (𝐸), so that|𝐸| = |𝑓 (𝐸)|.

And since𝑓 (𝐸) ⊂ 𝐹, we have|𝑓 (𝐸)| ≤ |𝐹 |.

2. It is a consequence of the previous question.

3. A participant shook either0, 1,… or𝑛 − 1hands. So we have𝑛”boxes”. Not that it is not possible to have at the same time the boxes0and𝑛 − 1non-empty. Therefore we have only𝑛 − 1boxes for𝑛 participants, so two participants must have shaken the same number of boxes.

Formally:

• First case: there is at least one participant who didn’t shake any hand. Then𝑓 ∶ {participants} → {0, 1, … , 𝑛 − 2}mapping each participant to the number of hands he shook is well-defined. Since

|{participants}| = 𝑛 > 𝑛 − 1 = |{0, 1, … , 𝑛 − 2}|, 𝑓 can’t be injective. Therefore at least two participants shooke the same number of hands.

• Second case: all participants shooke at least one hand. Then𝑓 ∶ {participants} → {1, … , 𝑛 − 1}

mapping each participant to the number of hands he shook is well-defined. Since|{participants}| = 𝑛 > 𝑛 − 1 = |{1, 2, … , 𝑛 − 1}|,𝑓 can’t be injective. Therefore at least two participants shooke the same number of hands.

4. For𝑟 = 1, 2, … , 𝑛, set𝑠𝑟 = ∑𝑟𝑘=1𝑎𝑘.

• First case: there exists𝑟such that𝑛|𝑠𝑟. Then we are done.

• Second case: otherwise, we have𝑛numbers𝑠1, … , 𝑠𝑛 whose remainders for the Euclidean divi- sion by𝑛are among1, … , 𝑛 − 1(i.e. 𝑛 − 1possible remainders). Hence at least two have the same remainders, let’s say𝑠𝑝and𝑠𝑞with𝑞 > 𝑝. Then𝑛|𝑠𝑞− 𝑠𝑝= ∑𝑞𝑘=𝑝+1𝑎𝑘.

5. First, note that 1+tantan𝑎−tan𝑎 𝑏

tan𝑏 =tan(𝑎 − 𝑏)and that

tan 𝜋

12 =tan(𝜋 3 − 𝜋

4 )= tan𝜋

3 −tan𝜋

4

1 +tan𝜋

3tan𝜋

4

= √3 − 1 1 + √3

= (√3 − 1)

2

2 = 2 − √3

(7)

Let𝑥1, … , 𝑥13 be 13 distinct real numbers. We set 𝛼𝑘 = arctan𝑥𝑘 ∈ (−𝜋2,𝜋

2). Note that the𝛼𝑘 are distinct since arctan is injective.

Note that(−𝜋2,𝜋

2) = 𝐼1⊔ 𝐼2⊔ 𝐼3⊔ ⋯ ⊔ 𝐼12where 𝐼1= (−𝜋

2, −𝜋 2 + 𝜋

12 ], 𝐼2 = (−𝜋 2 + 𝜋

12, −𝜋 2 + 2𝜋

12 ], … , 𝐼11 = (−𝜋

2 + 10 𝜋 12,𝜋

2 + 11𝜋

12 ], 𝐼12 = (−𝜋

2 + 11 𝜋 12,𝜋

2 ) We define𝑓 ∶ {𝛼1, … , 𝛼13} → {1, … , 12}by𝑓 (𝛼𝑘) = 𝑟where𝛼𝑘∈ 𝐼𝑟.

Since|{𝛼1, … , 𝛼13}| = 13 > 12 = |{1, … , 12}|, 𝑓 is not injective. So there exists 𝑎𝑙𝑝ℎ𝑎𝑘 < 𝛼𝑙 and 𝑟 = 1, … , 12such that𝛼𝑘, 𝛼𝑙∈ 𝐼𝑟. Then0 < 𝛼𝑙− 𝛼𝑘< 𝜋

12. Since tan is increasing on(−𝜋2,𝜋

2), we get that tan0 <tan(𝛼𝑙− 𝛼𝑘) <tan 𝜋

12 = 2 − √3.

Note that tan(𝛼𝑙− 𝛼𝑘) = tan𝛼𝑙−tan𝛼𝑘

1+tan𝛼𝑙tan𝛼𝑘 = 𝑥𝑙−𝑥𝑘

1+𝑥𝑙𝑥𝑘. Thus0 < 𝑥𝑙−𝑥𝑘

1+𝑥𝑙𝑥𝑘 <= 2 − √3as requested.

Références

Documents relatifs

This paper concentrates on a theoretical argument by Peter- son (2006) according to which the Precautionary Principle is incoherent with other desiderata of rational

Now, each of the

While the waiter left to write the sticky notes, one of the friends, who happens to be a mathematician, exclaims: ”I found a way so that we are 100% sure that one of us is

Prove that the following numbers are not rationals using the prime factorization

When two chameleons of different colors meet, they both change to the third color (for instance, if a green chameleon and a red chameleon meet, then they both become blue).. Cherge,

The pieces are allowed to move using the peg solitaire rules: a move consists of one piece jumping over another piece into an empty cell (either horizontally or vertically), the

As introduc- tory material consult [BG09] and [Das12b] for known results on the complexity of deep inference, [Jeˇr09] for the relationship between deep inference and the mono-

By weak monotone calculus we mean the fragment of Gentzen free of cuts between descendants of any structural rules..?. Weak