Proving a piece-wise defined function is bijective.

pondzo
Messages
168
Reaction score
0

Homework Statement


##f:\mathbb{Z} \to \mathbb{N}##
## f(x) = \begin{cases} 2x+2 & \text{if } x \geq 0 \\ -2x-1 & \text{if } x < 0 \end{cases} ##
Prove that f is a bijection from ##\mathbb{Z}## to ##\mathbb{N}##

Homework Equations

The Attempt at a Solution


Proving that it is an injection if quite simple. Proving it is a surjection is where i get confused.

Let ##y\in\mathbb{N} \text{ and } x \in \mathbb{Z}##

If ##x \geq 0 \text{ then } y=2x+2 ~~\Rightarrow~~ x=\frac{y-2}{2} ##
if ##y\geq 2 \text{ then } f(\frac{y-2}{2})=2(\frac{y-2}{2})+2=y##
if ##y<2 \text{ then } f(\frac{y-2}{2})=-2(\frac{y-2}{2})-1=1-y##

If ##x<0 \text{ then } y=-2x-1 ~~\Rightarrow~~ x=\frac{-(y+1)}{2}##
if ## y \geq -1 \text{ then } f(\frac{-(y+1)}{2})=2(\frac{-(y+1)}{2})+2=1-y##
if ## y<-1 \text{ then } f(\frac{-(y+1)}{2}) = -2(\frac{-(y+1)}{2})-1=y##

I'm pretty sure there should only be 2 or at most 4 cases that i need to investigate but i have 6. So I am clearly doing something wrong... Could someone help me out please?
 
Last edited:
Physics news on Phys.org
For the surjection, why not consider the cases where y is even and odd.
 
PeroK said:
For the surjection, why not consider the cases where y is even and odd.
Ahhh thank you PeroK! I'm certain i was thinking that on the bus home, and then i lost the thought and couldn't find it again. Thank you for reminding me!

Just an aside; Is my method even valid? i don't know what to make of the 1-y results.. does this matter?
 
pondzo said:
Ahhh thank you PeroK! I'm certain i was thinking that on the bus home, and then i lost the thought and couldn't find it again. Thank you for reminding me!

Just an aside; Is my method even valid? i don't know what to make of the 1-y results.. does this matter?

If I could follow your method, I'd tell you.

Normally, if you're trying to show something is a surjection onto a set Y, you would start with:

Let ##y \in Y##

Then find the relevant ##x \in X##
 
So if I split the problem into whether y is odd or even will I end up dealing with four cases?

1. y is odd and y is less than or equal to 1
2. y is odd and y is greater than 1
3. y is even and y is greater than or equal to 2
4. y is even and y is less than 2

This is how I am thinking about it; The first two cases arise because y being odd leads to ##x=\frac{1-y}{2}## and if y <= 1 then x would be positive so i use 2x+2 but if y>1 then x is negative so i would use -2x-1. The same argument goes for the last two.

I feel somewhere in that argument i am contradicting myself and in general don't have a whole lot of confidence in the way i have done things. Could you show me how the "y is odd" argument would look?
 
You don't need four cases, only two. To get you started:

Let ##y \in \mathbb{N}## be even. Then ##y = 2n## for some ##n \in \mathbb{N}##

Can you take it from there?
 
How does this look?

Let ##y \in \mathbb{N}## be even. Then ##y=2n## for some ##n \in \mathbb{N}##
##y=2x+2 ~~\Rightarrow~~ x=\frac{y-2}{2}=\frac{2n-2}{2}=n-1##

Let ##y \in \mathbb{N}## be odd. Then ##y=2n+1## for some ##n \in \mathbb{N}##
##y=-2x-1 ~~\Rightarrow~~ x=\frac{-(y+1)}{2}=\frac{-(2n+2)}{2}=-n-1##
 
Thanks for all your help PeroK.
 
A couple of things.

First, note that ##\mathbb{N} = {1, 2, 3 ...}##. So, you need to check your definition of an odd natural number for n = 1.

Second, you're writing what I would call rough notes. These do not form a logical argument.

Third, you haven't finished the proof. You've stopped half-way through.

To take the even case a step further:

Let ##y \in \mathbb{N}## be even. Then ##y = 2n## for some ##n \in \mathbb{N}##

(Now, do your rough notes off to one side, then come back with:

Let ##x = n - 1##

##x \in \mathbb{Z}## and ##f(x) = 2n = y##

That is a formal proof (of the even case). Do you see the difference?
 
  • #10
I really appreciate the tips you are giving me! I'm just starting pure maths in uni and i need to become familiar with the procedures and formalism.

So is this a little better?

Let ##y \in \mathbb{N}## be even. Then ##y=2n## for some ##n \in \mathbb{N}##
And let ##x=n-1~~\text{with}~~x\in\mathbb{Z}##
##f(x)=2(n-1)+2=2n=y##

Let ##y \in \mathbb{N}## be odd. Then ##y=2n+1## for some ##n \in \mathbb{N}##
And let ##x=-n-1~~\text{with}~~x\in\mathbb{Z}##
##f(x)=-2(-n-1)-1=2n+1=y##

This shows that for every element of ##\mathbb{N}## there is a corresponding element of ##\mathbb{Z}## which f maps to it. Therefore f is a surjection of N onto Z.

Is the conclusion sufficient or must it be altered? Also is it a good idea to make the "rough notes" part of the proof or at least visible?
 
  • #11
And regarding my definition of an odd number i would change it to;

Let ##y \in \mathbb{N}## be odd. Then ##y=2n-1## for some ##n \in \mathbb{N}##
And let ##x=-n~~\text{with}~~x\in\mathbb{Z}##
##f(x)=-2(-n)-1=2n-1=y##
 
  • #12
Yes, much better.

A minor detail (I missed this too.) It's better to mention whether x is < 0 or not, so you know you're using the correct formula for f.

Let ##x = -n##

##x \in \mathbb{Z}, x < 0 \ \ and \ \ f(x) = -2(-n) -1 = 2n - 1 = y##

Re notes and proofs in general. You should ask your lecturers what they expect in an exam. What I do is work out the notes to one side then decide how much (if any) needs to be mentioned in the proof for clarity and justification.

A brief recap at the end of exactly what you've shown is often worthwhile.
 
  • #13
If it is necessary to state the parity of x then does this mean there will be more cases I need to consider (i.e. if x<0 or x##\geq 0##)?

Also I could tighten up the conclusion by writing instead: "this shows that##~~\forall~ y~\in\mathbb{N}\exists x~\in~\mathbb{Z}:f(x)=y## which if anyone else typeset would look more aesthetically pleasing to the eye...
 
  • #14
pondzo said:
If it is necessary to state the parity of x then does this mean there will be more cases I need to consider (i.e. if x<0 or x##\geq 0##)?

No. But the sign of x tells you which formula to use. So, this justifies why you used the formula in each case.
 
  • #15
PeroK said:
No. But the sign of x tells you which formula to use. So, this justifies why you used the formula in each case.
Ahh yes of course, I forgot that ##n\in\mathbb{N}## and not ##n\in\mathbb{Z}##.
Thanks again for all your help PeroK.
 
Back
Top