F bijective <=> f has an inverse

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
7 replies · 3K views
member 587159

Homework Statement



Proof that: f has an inverse ##\iff## f is a bijection

Homework Equations

/definitions[/B]

A) ##f: X \rightarrow Y##

If there is a function ##g: Y \rightarrow X## for which ##f \circ g = f(g(x)) = i_Y## and ##g \circ f = g(f(x)) = i_X##, then ##g## is the inverse function of ##f##.

B) A function ##f## is a bijection if
##\forall y \in Y, \exists! x \in X: y = f(x)##
or ##f## is surjective and injective
or there is a function ##g: Y \rightarrow X## for which ##f \circ g = f(g(x)) = i_Y## and ##g \circ f = g(f(x)) = i_X##

The Attempt at a Solution



##\Rightarrow##

##f##

##f: X \rightarrow Y## has an inverse, thus ##f \circ g = f(g(x)) = i_Y## and ##g \circ f = g(f(x)) = i_X##. We must show that f is bijective, so we will show it is both injective and surjective.

1) Take ##x_1,x_2 \in X##

##f(x_1) = f(x_2)##
##\Rightarrow g(f(x_1)) = g(f(x_2))##
##\Rightarrow i_X(x_1) = i_X(x_2)##
##\Rightarrow x_1 = x_2##

Thus, f is injective.

2) Take ##y \in Y##

Then, ##y = i_Y(y) = f(g(y))##.

Since g is a function ##Y \rightarrow X##, we have ##\forall y \in Y, \exists g(y) \in X: y = f(g(y))## thus f is surjective.

##\Leftarrow##

##f## is a bijection, thus we have that there is a function ##g: Y \rightarrow X## for which ##f \circ g = f(g(x)) = i_Y## and ##g \circ f = g(f(x)) = i_X##. And this is exactly the definition of the inverse function.

So, QED. Or not? Now comes the actual question. Is the definition my textbook provides for bijection wrong? I think this, because what I actually have to proof, is in the definition. I do understand that it is a useful criterium to see whether a function is a bijection, but should it be in the very definition of bijection? Thank you.
 
Physics news on Phys.org
No, the second part of the proof is incorrect. You have assumed the definition of bijective is equivalent to the definition of having an inverse, before proving it. Assume ##f## is a bijection, and use the definition that it is both surjective and injective. Then use surjectivity and injectivity to show some ##g## exists with the properties of the inverse.
 
  • Like
Likes   Reactions: member 587159
Lucas SV said:
No, the second part of the proof is incorrect. You have assumed the definition of bijective is equivalent to the definition of having an inverse, before proving it. Assume ##f## is a bijection, and use the definition that it is both surjective and injective. Then use surjectivity and injectivity to show some ##g## exists with the properties of the inverse.

So now I need to show that

##f## is a bijection ##\Rightarrow f## has in inverse.

Proof: ##f## is a bijection. Thus it is both injective and surjective. This means, that:
##f(x_1) = f(x_2) \Rightarrow x_1 = x_2## and ##\forall y \in Y, \exists x \in X: y = f(x)##

Now, we need to proof that there exists a function ##g: Y \rightarrow X## for which ##f \circ g = f(g(x)) = i_Y## and ##g \circ f = g(f(x)) = i_X##.
So, let us first define ##g: Y \rightarrow X##:

##g(y) = x \iff f(x) = y##. (we can write this because f is a bijection, this is why the surjectivity is nessecary)

Take ##x_1,x_2 \in X##: Then we can write:

##g(f(x_1)) = x_2 \Rightarrow f(x_2) = f(x_1) \Rightarrow x_2 = x_1##. Thus, ##g(f(x)) = x = i_X##
Analogue, we obtain: ##f(g(x)) = x = i_Y##. Thus, ##g## is the inverse function of ##f## by definition.

I'm not to sure about the way I defined ##g## though.
 
Math_QED said:
So, let us first define ##g: Y \rightarrow X##:

##g(y) = x \iff f(x) = y##. (we can write this because f is a bijection, this is why the surjectivity is nessecary)

That's not so much wrong as not quite right! You can patch it up logically as follows:

We define ##g: Y \rightarrow X##:

For ##y \in Y##, we define ##g(y) = x## where ##f(x) = y##.

##g## is well-defined as ##\forall y \in Y, \exists## unique ##x## such that ##f(x) = y\ ## (as ##f## is a bijection).

In general, what you have done is not wrong but often you have done more than is necessary, For example:

If ##f## has an inverse, then ##f## is surjective:

For ##y \in Y##, ##f(f^{-1}(y)) = y##.

Hence ##f## is surjective.

That's actually all you need.
 
  • Like
Likes   Reactions: member 587159
PeroK said:
That's not so much wrong as not quite right! You can patch it up logically as follows:

We define ##g: Y \rightarrow X##:

For ##y \in Y##, we define ##g(y) = x## where ##f(x) = y##.

##g## is well-defined as ##\forall y \in Y, \exists## unique ##x## such that ##f(x) = y\ ## (as ##f## is a bijection).

Thanks a lot for this.

In general, what you have done is not wrong but often you have done more than is necessary, For example:

If ##f## has an inverse, then ##f## is surjective:

For ##y \in Y##, ##f(f^{-1}(y)) = y##.

Hence ##f## is surjective.

That's actually all you need.

What have I done more here that was unnessecary? Can you point out other places where I did too much?
 
Math_QED said:
2) Take ##y \in Y##

Then, ##y = i_Y(y) = f(g(y))##.

**Since g is a function ##Y \rightarrow X##, we have ##\forall y \in Y, \exists g(y) \in X: y = f(g(y))##**

thus f is surjective.**

You could have missed out the bit I've marked **. Your logic stands up without spelling it out.

I think it's a natural process when you learn this to say things like "as ##g## is a function". It's not wrong and - perhaps it's even a good thing at this stage. But, it's useful to know that you can start to drop that. And, if you have more difficult problems, you can soon get lost in all the detail and lose the wood in the trees.
 
  • Like
Likes   Reactions: member 587159
PeroK said:
You could have missed out the bit I've marked **. Your logic stands up without spelling it out.

I think it's a natural process when you learn this to say things like "as ##g## is a function". It's not wrong and - perhaps it's even a good thing at this stage. But, it's useful to know that you can start to drop that. And, if you have more difficult problems, you can soon get lost in all the detail and lose the wood in the trees.

Is that the only thing that was unnessecary? I believe you are right. Those are things that may be obvious to you and other people who have a lot of experience in mathematics, but all the concepts are rather new for me and it only becomes obvious to me when I write it out in that way. Probably later..

I hope the proof is correct now. Thanks a lot for your help. You have helped me out a lot of times.
 
Math_QED said:
Is that the only thing that was unnessecary? I believe you are right. Those are things that may be obvious to you and other people who have a lot of experience in mathematics, but all the concepts are rather new for me and it only becomes obvious to me when I write it out in that way. Probably later..

I hope the proof is correct now. Thanks a lot for your help. You have helped me out a lot of times.

The proof is well done. And, the more confident you become, the more you will want to drop some of the details yourself.
 
  • Like
Likes   Reactions: member 587159