Let n be a positive integer, and let σ : {1, . . . , n} → {1, . . . , n} be a one-to-one and onto map. Then σ is called a permutation on n elements. The set of all permutations on n elements is denoted by Sn. The set Sn is important in understanding the determinant. Suppose τ is a permutation on n elements. If there exist i, j ∈ {1, . . . , n} with i $\ne$ j such that τ (i) = j, τ (j) = i and τ (k) = k for all k ∈ {1, . . . , n} \ {i, j}, we say that τ is a transposition.
(a) Let σ, τ ∈ Sn. Define στ := σ ◦ τ . That is, στ is τ composed with σ. Prove that στ ∈ Sn.
(b) Let σ1, σ2, σ3 ∈ Sn. Prove that (σ1σ2)σ3 = σ1(σ2σ3).
(c) Let I : {1, . . . , n} → {1, . . . , n} denote the identity permutation. That is, I(i) := i for all i ∈ {1, . . . , n}. Show that, for all σ ∈ Sn, Iσ = σI = σ.
(d) Given σ ∈ Sn, show that there exists a unique element σ −1 ∈ Sn such that σσ−1 = σ−1σ = I.
My Answers:
(a)
στ(i) = σ(τ(i)) = σ(j) = m for some m $\in$ {1, . . . , n}
στ(j) = σ(τ(j)) = σ(i) = g for some g $\in$ {1, . . . , n} \ {m}
στ(k) = σ(τ(k)) = σ(k) = b for some b $\in$ {1, . . . , n} \ {m. g} $\forall$ k$\in${1, . . .,n} \ {i, j}
since σ is one-to-one and onto, therefore στ is a permutation of n elements so στ$\in$Sn
(b)
since σ1, σ2, σ3 ∈ Sn, σ1, σ2, σ3: {1, . . . , n} $\to$ {1, . . . , n}. For any k $\in${1, . . . , n}, then
((σ1σ2)σ3)(k) = (σ1σ2)(σ3(k)) = σ1(σ2(σ3(k)))
and
(σ1(σ2σ3))(k) = σ1((σ2σ3)(k)) = σ1(σ2(σ3(k)))
since ((σ1σ2)σ3)(k) = (σ1(σ2σ3))(k) $\forall$ k $\in${1, . . . , n}, we have (σ1σ2)σ3 = σ1(σ2σ3).
(c) σ : {1, . . . , n} → {1, . . . , n}, let σ(k)=m for some k, m $\in$ {1, . . . , n}, then
Iσ(k) = I(σ(k)) = I(m) = m and σI(k) = σ(I(k)) = σ(k) = m $\forall$ k$\in${1, . . . , n}
(d)
Since σ is one-to-one and onto, it is an invertible mapping, thus σ-1 exists and is unique by the properties of the inverse. suppose σ(i)=j, then σ-1(j)=(i), thus
σσ-1(j)=σ(σ-1(j))=σ(i)=j thus σσ-1=I
σ-1σ(i) = σ-1(σ(i)) = σ-1(j) = i thus σ-1σ = I
How correct/incorrect am I? Thanks!
These all seem fine to me except for part a; I don't quite understand your argument for that one. You are trying to show that $\sigma\tau$ is a one-to-one and onto map from $\{1,\ldots,n\}\to\{1,\ldots, n\}$.
To show the map is one-to-one, you might suppose that $\sigma(\tau(i))=\sigma(\tau(j))$. Then since $\sigma$ is one-to-one, you have $\tau(i)=\tau(j)$, and similarly, since $\tau$ is one-to-one, it must be that $i=j$. Therefore $\sigma\tau$ is one-to-one.
To show the map is onto, let $k\in\{1,\ldots,n\}$. Then since $\sigma$ is onto, there is some $j$ such that $\sigma(j)=k$. Similarly, since $\tau$ is onto, there is some $i$ such that $\tau(i)=j$. Then $\sigma(\tau(i))=\sigma(j)=k$, so $\sigma\tau$ is onto.