01 - Intro to the Course, Magma and Groups

Today we talked about:

magma

A magma is a set G together with a binary operation:
μ:G×G→G;μ(a,b)↦μ(a,b)

Some Examples

Take the set Z={1,2,3,4,…} which has some example binary operations:

  1. μ(a,b)=a+b
  2. μ(a,b)=ab
  3. μ(a,b)=min{a,b}
  4. μ(a,b)=max{a,b}
  5. μ(a,b)=a ("projection")
  6. μ(a,b)=ab
  7. μ(a,b)=2ab
  8. ...

Two Possible Properties

We may want to ask: Is the operation μ (for a,b,c∈G)

  1. associative: μ(a,μ(b,c))=μ(μ(a,b),c) (think of a(bc)=(ab)c from the integers, parenthesis don't matter, here the operations can be applied in whatever order as long as it's in-order))
  2. commutativity: μ(a,b)=μ(b,a) (ab=ba) (order doesn't matter)

Looking at the examples above:

  1. Both (commutative and associative)
  2. Both
  3. Both (equivalent to min{a,b,c})
  4. Both (same as (3))
  5. Not Commutative (since μ(2,1)=2≠1=μ(1,2)), but is Associative
  6. Not Commutative (since μ(1,2)=12≠21=μ(2,1)), and Not Associative (abc vs. (ab)c≡abc, just choose an example where bc≠bc)
  7. Commutative but not associative

It's actually not too uncommon to find operations that aren't commutative, but the associativity is an important property, since there's only a few operations that really aren't associative (ex: the cross product × in R3)

group

A group is a set G together with the binary operation μ (see Magma (Algebraic)) such that 3 properties hold:

  1. μ must be associative
  2. There must be some identity element e∈G such that for any other element a∈G then:
    μ(a,e)=a=μ(e,a)
  3. Every element a∈G has some inverse b∈G such that:
    μ(a,b)=μ(b,a)=e
Why not name the inverse?

Normally we denote a−1 as the inverse of a, but we don't name it yet since we haven't shown that the inverse is uniuqe.

There is some new notation that we'll use temporarily:

Write a⋆b for μ(a,b). Then the properties above become:

  1. a⋆(b⋆c)=(a⋆b)⋆c
  2. a⋆e=a=e⋆a
  3. a⋆b=e=b⋆a

Example

For instance, all the sets and their associated operations:

  1. Z,+
  2. Q,+
  3. R,+
  4. C,+
    Are all valid groups. That's because:
  5. They are all associative (easy to show)
  6. The identity is 0
  7. The inverse is −a for each a∈G since a+(−a)=0=(−a)+a

Another example is all the sets and their associated operations:

  1. Z,⋅
  2. Q,⋅
  3. R,⋅
  4. C,⋅
    None of these are valid groups. That's because:
  5. They are all associative
  6. The identity is 1 for all of them
  7. 0 doesn't have an inverse here!
    What we do here is just "throw away" the dud which is 0. If we only consider those sets with 0 removed then:
  8. (Q−{0},⋅) is not a group where each ab has its inverse ba
  9. The other groups follow a similar patter (other than (Z−{0},⋅) obviously)

Another example is the set of functions from R→R and the binary operation + is just adding their outputs. Namely (Fun(R,R),+) where:
(f+g)(x):=f(x)+g(x)

  1. Associativity comes from associativity of R,+
  2. The identity is the zero function o(x)=0
  3. The inverse of some f(x) is just −f(x) since f(x)+(−f(x))=o(x)=(−f(x))+f(x)

Another example is doing compositions, namely (Fun(R,R),∘) where for all f,g in this set:
(f∘g)(x)=f(g(x))
Here:

  1. Associativity is definitely here
  2. Identity is just I(x)=x
  3. The inverse is the problem! There isn't always an inverse. Namely f has an inverse iff ∃g where (f∘g)=I=(g∘f).