Sie sind auf Seite 1von 3

MAT246 Notes 3, Chapter 3

Divisibility revised:

Starting this section and for the continuation of course, in chapters 3-9 we need to apply the concept of
divisibility and factors to all the integers. So we must extend this definition to include all integers. Study
this section carefully and reflect on the idea that, whenever we extend one idea from one type of numbers
to include a larger group of numbers then we may have to deal with unexpected situations. We present a
conservative extension of divisibility because that is all need for chapter 3-7, however in Exercise 1 we will
discuss whatever could go wrong if we had not imposed the condition m > 0.

Definition: given integers m > 0 and n, we say m|n if there exists an integer k such that n = km.

Exercise 1:

a) Show that it is not possible to allow m to take value 0. (a formal proof please.)

b) However, n is allwoed to be any integer. Show that any number m divides 0.

c) Show that if a non-zero integer m divides an integer n, then the integer k that is guaranteed by the
definition, must be unique. (May be you don’t know how to show some particular number is unique; this is
your chance to really reflect on this kind of problem and to learn them because uniqueness arguments are going
to appear in the course very often.)
n
d) In this case, when m|n, because k is unique, then we can use the following Convention: m is the
m
unique integer k guaranteed by the definition. But but be careful: don’t think of n as division or as
fraction because we don’t know these two ideas yet. And more important: if we don’t know m|n then
n
we are not allowed to write the expression m as it is meaningless. Now explain why the expression n0 is
not allowed.

e) Explain what is wrong with the expression 00 ?

Exercise 2: Recall we proved a lemma in Notes 1, that for the original definition of divisibility, if m|n
then m ≤ n. Is this Lemma still true? How would you modify the statement of the old lemma to make it
into a lemma for the new definition of divisibility? Would we still be able to apply WOP in our arguments?

Essentials of Modular Arithmetics:

A) In section 3.1, the basic definition (3.1.1) of congruency relation modulo m > 1 is really all that is
needed to understand the proofs and computations of this section. Note, a and b can be any integers
(negative, positive or zero) but m must be greater than 1. Why?

Exercise 3: Check definition of congruency for m = 1, and explain why the assumption m > 1 is
needed.

Page 1 of 3
MAT246 Notes 3, Chapter 3
B) Theorem 3.1.2 suggests that the congruency relation is an ‘equivalence relation’ on the integers, (what
does that mean?), a property that is shared with equality. (in which ways?) This means that ‘congruency
relation’ works in similar ways as the equality relation works; see theorems 3.1.2, 3.1.5, 3.1.6. So when
it comes to do modular arithmetic (mod m), we can easily replace any number with anther number
the is congruent to it. For example since 278 ≡ 3 (mod 5) then 2754 ≡ 34 (mod 5). See all the
examples in section 3.2, and see exercises 1-7.

Important note: so far, before you took MAT246, you knew two very important relations between
numbers: ≤ as an order relation between numbers, and =, equality of two numbers. With these two
relations we performed all our Arithmetic and Algebra. From now on we are equipped, in addition to
the ≤ and =, with two new relations of divisibility, that is an order relation, like ≤, and Congruency,
that is an equivalence relation, like =. Please make sure to keep all your channels of intuition open to
see how the textbook uses these two new relation in conjunction with the old relations. This is how we
build an intuition about these two new relations.

C) One important property of any equivalence relation (and in particular of the congruency relation) is
that for any integer a and any m > 1 we have a ≡ a (mod m). This may seem trivial, and not
important in proofs and calculations, but read further: in questions 22 section 3.3, they asked if it is
possible at all to have a solution to x2 + y 2 = 4003. Assume there are numbers a and b that satisfy
a2 + b2 = 4003, and since the two sides are equal, then you can assume they are congruent in mod 4
... now continue to see if you can derive a contradiction. (Tutorials try to discuss this further.) Also
see exercise number 27 in 3.3.

Exercise 4: Show that the first implication below is correct and the second implication is incorrect:
for integers a and b, and m > 1,
a = b =⇒ a ≡ b (mod m)
a ≡ b (mod m) =⇒ a = b

D) The congruency relation is strongly relevant to the division algorithm. This close association makes
suggests another name for the Congruency relation: the same-remainder relation.

an outline of proof: (see also the argument given in the middle of page 24, theorems 3.1.3 and
3.1.4)

i) No two numbers in the range {0, 1, 2, . . . , m − 1} can be congruent to one another in mod m.
(Formally prove this please.) This means that within this range if two numbers are congruent, they
must be equal.
ii) Two numbers a and b that are congruent mod m must have the same remainder after division by
m; that is, if ra and rb are the corresponding remainders, then ra = rb .

Because of this property, modular arithmetic seems to help us with reducing computation involving
large numbers to computations involving small numbers in the range {0, 1, 2, . . . m − 1}. This means
we can replace any number with its remainder after division by m.

Page 2 of 3
MAT246 Notes 3, Chapter 3
E) theorem 3.1.4 states that a small set {0, 1, 2, . . . m − 1} is now a good (and for our purpose, an
accurate) representative of all other integers (see ex 22 and 27.) As such any result of arithmetic can
be examined in modular arithmetic using a properly selected modulus.
Sometimes they denote the set {0, 1, 2, . . . m − 1}, equipped with the summation and multiplication
of mod m by Zm . If a result is true in the usual arithmetic (that is in Z, then it must also be true in
and of the smaller ‘modular arithmetic’ Zm . And more importantly, if a result is false in Zm for
some m, then the result must be false for Z as well.

F) There is a big difference between modular arithmetic in Zm and Zp , where p is a prime number, and
when m is a composite number. Draw the multiplication table in Z5 and repeat in Z6 and see that in
case of 5 or and prime number, the modular arithmetic presents properties consistent with the usual
arithmetic (of integers.) But in the case of 6 or any composite number the modular arithmetic (mod
6) is not standard. We shall study this further in chapter 5.

G) One very important warning: item (b) above suggests that in any modular arithmetic mod m we
can replace any two numbers that are congruent mod m. However one may be tempted to perform the
following incorrect operation:

218 ≡ 23 (mod 5) because 18 ≡ 3 (mod 5)

What is wrong with this solution? after all we replaced 18 with 3 in mod 5. The problem is that the
number 18 in this problem is not really a number; it is rather a shorthand for 2 times 2 18 times. This
18 times should not be taken as a one of the numbers that are subject to the modular arithmetic. This
idea will be developed further in chapter 5.

H) Since the congruency relation is an equivalence relation, each number has an equivalence class in any
given mod. For example the equivalence class of 1 in mod 2 is the set of all odd numbers and the
equivalence class of 0 in mod 2 is the set of all even numbers. Try to answer the questions below:

1. if m = 1 then there is only one remainder, {0}. What is the equivalence class of 0 in mod 1?
2. Suppose someone asks you this natural question that appears in an IQ test: the relation between
Odd and Even numbers to mod 2 is like relation between what and what in mod 3. What would
you answer? In any case, instead of dividing the numbers between Odd and Even, in case of mod
2, we divide the numbers into three equivalent classes .... in mod 3.

Exercise 5: Let [a] denote the equivalence class of a number a in mod m, that is the set of all integers
that are congruent to a mod m.

a) Prove that if a ≡ b (mod m) then [a] = [b]. (Isn’t this just the transitivity property of the congruence
relation?)
b) Prove that if a and b are not congruent mod m then [a] ∩ [b] = ∅. This means that any two
equivalence classes are either the same or they are disjoint.
c) Prove that the set of all integers is partitioned by the equivalence classes of 0, 1, 2, . . . , m − 1.

Page 3 of 3

Das könnte Ihnen auch gefallen