Every injection yields a bijection onto its image
Contents
Statement
[ilmath]\newcommand{\f}{\overline{f}}[/ilmath]Let [ilmath]X[/ilmath] and [ilmath]Y[/ilmath] be sets and suppose [ilmath]f:X\rightarrow Y[/ilmath] is any injective map between them. Then we claim that there is a map:
- [ilmath]\overline{f}:X\rightarrow f(X)[/ilmath] given by [ilmath]\overline{f}:x\mapsto f(x)[/ilmath]- where [ilmath]f(X)[/ilmath] denotes the image of [ilmath]X[/ilmath] under [ilmath]f[/ilmath][Note 1]
that is a bijection between [ilmath]X[/ilmath] and [ilmath]f(X)[/ilmath] (that is to say, we claim that [ilmath]\f[/ilmath] is a bijection, a bijection is any map that is both injective and surjective)
Proof
We must show that [ilmath]\f:X\rightarrow f(X)[/ilmath] is a bijection (that means [ilmath]\f[/ilmath] is both injective and surjective), even though injectiveness is (even more than surjectiveness) trivial, we do it anyway. That's why this page is marked as first-year friendly
- Injectivity of [ilmath]\f:X\rightarrow f(X)[/ilmath] given by [ilmath]\f:x\mapsto f(x)[/ilmath]
- We will take injectivity to mean: [ilmath]\forall x_1,x_2\in X[\f(x_1)=\f(x_2)\implies x_1=x_2][/ilmath] (rather than other equivalent conditions documented on the injection page)
- Let [ilmath]x_1,x_2\in X[/ilmath] be given
- Suppose [ilmath]\f(x_1)\ne\f(x_2)[/ilmath]
- By the nature of logical implication we do not care if the RHS ([ilmath]x_1=x_2[/ilmath]) is true or false.
- By the principle of excluded middle [ilmath]x_1=x_2[/ilmath] must be true xor false[Note 2], so we're done in either case.
- By the nature of logical implication we do not care if the RHS ([ilmath]x_1=x_2[/ilmath]) is true or false.
- Suppose [ilmath]\f(x_1)=\f(x_2)[/ilmath]
- In order for the implication to be true, we require that this means [ilmath]x_1=x_2[/ilmath]
- Notice that [ilmath]\f(x_1)=f(x_1)[/ilmath] and [ilmath]\f(x_2)=f(x_2)[/ilmath] (by definition of [ilmath]\f[/ilmath])
- As [ilmath]f:X\rightarrow Y[/ilmath] was injective itself, this means:
- [ilmath]\forall x_1,x_2\in X[\f(x_1)=\f(x_2)\implies x_1=x_2][/ilmath]
- By hypothesis, [ilmath]\f(x_1)=\f(x_2)[/ilmath] so [ilmath]f(x_1)=f(x_2)[/ilmath]
- This means that [ilmath]x_1=x_2[/ilmath] by injectivity of [ilmath]f[/ilmath]
- So [ilmath]\f(x_1)=\f(x_2)\implies f(x_1)=f(x_2)\implies x_1=x_2[/ilmath] (then by transitivity of implies [ilmath]\f(x_1)=\f(x_2)\implies x_1=x_2[/ilmath])
- We know [ilmath]\f(x_1)=\f(x_2)[/ilmath] to be true, therefore, by definition of implies and by knowing [ilmath]\f(x_1)=\f(x_2)\implies x_1=x_2[/ilmath] we must have:
- [ilmath]x_1=x_2[/ilmath]
- We know [ilmath]\f(x_1)=\f(x_2)[/ilmath] to be true, therefore, by definition of implies and by knowing [ilmath]\f(x_1)=\f(x_2)\implies x_1=x_2[/ilmath] we must have:
- This completes the first part of the proof
- Suppose [ilmath]\f(x_1)\ne\f(x_2)[/ilmath]
- Surjectivity of [ilmath]\f:X\rightarrow f(X)[/ilmath] given by [ilmath]\f:x\mapsto f(x)[/ilmath]
- To be surjective we require: [ilmath]\forall y\in f(X)\exists x\in X[\f(x)=y][/ilmath]
- Let [ilmath]y\in f(X)[/ilmath] be given
- By definition of [ilmath]f(X)[/ilmath] (that is [ilmath]f(X):=\{y\in Y\ \vert\ \exists x\in X[f(x)=y]\}[/ilmath] remember) we see:
- [ilmath]y\in f(X)\iff\exists x\in X[f(x)=y][/ilmath]
- Choose [ilmath]x\in X[/ilmath] such that [ilmath]f(x)=y[/ilmath] (which we can do thanks to the above)
- Now [ilmath]\f(x)=f(x)[/ilmath] by definition of [ilmath]\f[/ilmath] and
- [ilmath]f(x)=y[/ilmath] by our choice of [ilmath]x[/ilmath]
- So [ilmath]\f(x)=f(x)=y[/ilmath] or just [ilmath]\f(x)=y[/ilmath]
- Now [ilmath]\f(x)=f(x)[/ilmath] by definition of [ilmath]\f[/ilmath] and
- We have shown there exists an [ilmath]x\in X[/ilmath] such that [ilmath]\f(x)=y[/ilmath] for our given [ilmath]y[/ilmath]
- By definition of [ilmath]f(X)[/ilmath] (that is [ilmath]f(X):=\{y\in Y\ \vert\ \exists x\in X[f(x)=y]\}[/ilmath] remember) we see:
- Since [ilmath]y\in f(X)[/ilmath] was arbitrary we have shown this for all [ilmath]y[/ilmath]
- That completes the surjectivity part of the proof
TODO: Check, I should have left this blank and marked it as low-hanging fruit....
Thus we have shown that [ilmath]\f:X\rightarrow f(X)[/ilmath] given by [ilmath]\f:x\mapsto f(x)[/ilmath] is a bijection
See also
Notes
- ↑ Formally:
- for [ilmath]A\in\mathcal{P}(X)[/ilmath] we define [ilmath]f(A):=\{y\in Y\ \vert\ \exists a\in A[f(a)=y]\}[/ilmath]
- [ilmath]f(X):=\{y\in Y\ \vert \exists x\in X[f(x)=y] \}[/ilmath] - the set of all things in [ilmath]Y[/ilmath] that are mapped to by [ilmath]f[/ilmath] for some [ilmath]x\in X[/ilmath]
- ↑ Xor means "exclusive or", [ilmath]A\text{ XOR }B[/ilmath] means [ilmath]A[/ilmath] or [ilmath]B[/ilmath] is true and [ilmath]A\text{ AND }B[/ilmath] is false. That is to say we have [ilmath]A[/ilmath] or [ilmath]B[/ilmath] but not both
References
|