In the previous note, we defined the congruence relation on the set of integers , and we proved that it is an equivalence relation. In this note, we will define , the set of congruence classes modulo .
Understand the Congruence Classes as formal symbols
We will define a map from to as follows:
where is the congruence class of modulo , we will drop the subscript and write when the modulus is clear from the context.
We would like to give a special name for elements in :
Notation
When we say congruence class modulo n, we mean an element in . When we refer to the elements in , we will call them congruence classes. A congruence class is denoted as for some integer .
As promised, we should have to be a set with elements, but from the map above, it seems like we are having infinitely many elements in . Well, even for , we may have ! Namely, you can start with two distinct integers and , but end up with the same congruence class .
The key point is that we are not considering the elements in as individual integers, but as equivalence classes, where we are identifying integers according to congruence relation.:w
Letβs begin by treating as formal symbols, and we impose the following identification:
Notation
Two elements are equal if and only if .
Now if we just treat as formal symbols, we can observe the following cyclic property:
- are all distinct as congruence classes.
- for any integer .
- for any integer .
So far, we have not defined the congruence classes in rigorously, we just introduced how to represent them as formal symbols, and how to identify them based on the congruence relation.
Understand the Congruence Classes as Equivalence Classes
One may still feel uncomfortable with the formal symbols and the identification we imposed. We will now propose a better, more concrete way to understand the congruence classes in .
Definition
Let be an integer. The congruence class is a subset of defined as:
Based on this definition, we can show this definition is consistent with the formal symbols:
Proposition
Based on the above definition of congruence classes. Let , then if and only if .
Proof of Proposition
We will show two directions:
- If , then .
- If , then .
The first direction is easy, if , then for some integer . Then we have:
For the second direction, if , then we know the following two sets are equal:
In particular, and , this implies for some integer , which means .
Understand the Quotient Map
Now, we have defined the congruence classes in as subsets of , and we have shown that this definition is consistent with the formal symbols we introduced earlier. We can now define the quotient map from to , and give some intuitive justification for the notation of congruence classes and the ambiguity in its representation.
Definition
The quotient map from to is defined as:
Intuitive Remark
The map is surjective, we can image this map is put all integers into boxes, and all integers in the same box are identified as the same congruence class. Now the problem is how should we label the boxes?
The notation is a self-reference to the box that contains . The map is trying to say βput into the box that contains , which we will call β. The ambiguity in the representation comes from the fact that there will be multiple (infinite) integers in the same box, and we can choose any representative of the box to label it.
One may dislike the ambiguity in the representation of congruence classes. Here is a way of avoiding it using division algorithm:
Proposition
We can express the quotient map as:
And there are distinct congruence classes in .
The proof is just applying the division algorithm to and . We will leave it as an exercise.