Does there exist a surjection from the integers to the naturals?

  • #1
docnet
799
487
Homework Statement
a) Does there exist a surjection from the integers to the natural numbers?

b) Does there exist a surjection from the real numbers to the natural numbers?
Relevant Equations
##\mathbb{R},\quad \mathbb{N},\quad\mathbb{Z}##
a)
Yes.
One surjection from ##\mathbb{Z}## to ##\mathbb{N}## is the double cover of ##\mathbb{N}## induced by ##f:\mathbb{Z}\longmapsto\mathbb{N}## with
$$f(z)=\begin{cases}
-z & ,\forall z<0\\
z+1 & ,\forall 0\leq z
\end{cases}$$
b)
Yes.
One surjection from ##\mathbb{R}## to ##\mathbb{N}## is the the projection ##p:\mathbb{R}\longmapsto\mathbb{N}## with ##f(r)=r## for all ##r\in\mathbb{N}\subset\mathbb{R}##.
 
Last edited:
Physics news on Phys.org
  • #2
Does ##\mathbb{N}## not include 0 for you?

(b) feels incomplete. What is the relationship between p and f, and what are you supposed to to with real numbers that are not natural numbers?
 
  • #3
Office_Shredder said:
Does ##\mathbb{N}## not include 0 for you?
If ##\mathbb{N}## includes ##0##, a surjection from ##\mathbb{Z}## to ##\mathbb{N}## is given by
$$f(z)=\begin{cases}
-z & ,\forall z<0\\
z & ,\forall 0\leq z
\end{cases}$$
Office_Shredder said:
(b) feels incomplete. What is the relationship between p and f, and what are you supposed to to with real numbers that are not natural numbers?
I agree with you. They were all meant to be ##p##'s, but I hit ##f## by force of habit. The "complete" projection function is ##p:\mathbb{R}\longmapsto\mathbb{N}## with
$$p(r)=\begin{cases}r& ,\forall r\in\mathbb{N}\\
\text{undefined}& ,\forall r\notin\mathbb{N}
\end{cases}$$
 
  • #4
p(r) can't be undefined, by definition a function is always defined. But notice it doesn't matter much what value you pick - you already covered the surjection part, so you could make it e.g.. always 0 for the rest of the space

docnet said:
If ##\mathbb{N}## includes ##0##, a surjection from ##\mathbb{Z}## to ##\mathbb{N}## is given by
$$f(z)=\begin{cases}
-z & ,\forall z<0\\
z & ,\forall 0\leq z
\end{cases}$$

I agree that works. But which one does your class/textbook use? I was just surprised because they usually include 0, but if it didn't then your first answer is fine.
 
  • Like
Likes docnet
  • #5
Office_Shredder said:
p(r) can't be undefined, by definition a function is always defined. But notice it doesn't matter much what value you pick - you already covered the surjection part, so you could make it e.g.. always 0 for the rest of the space
Thank you. I understand, and I can't believe I missed that.
Then let's make it ##p:\mathbb{R}\longmapsto\mathbb{N}## with
$$p(r)=\begin{cases}r& ,\forall r\in\mathbb{N}\\
0 & ,\forall r\notin\mathbb{N}
\end{cases}$$
Office_Shredder said:
I agree that works. But which one does your class/textbook use? I was just surprised because they usually include 0, but if it didn't then your first answer is fine.
Those questions are taken from a midterm review sheet that I stumbled upon. The class is named "Introduction to higher maths" and it is a proof-based course I have not taken, so I don't know what the professor wants to do.
 
  • #6
In general if ##A \subseteq B## are non-empty sets and ##a \in A##, then ##f:B \rightarrow A## defined by:
$$f(b)=\begin{cases}b& \ \ (b \in A)\\
a & \ \ (b \notin A)
\end{cases}$$ is a surjection.
 
  • Informative
Likes docnet
  • #8
Office_Shredder said:
p(r) can't be undefined, by definition a function is always defined. But notice it doesn't matter much what value you pick - you already covered the surjection part, so you could make it e.g.. always 0 for the rest of the space
Question: Does this imply that any function that has an infinity, like ##f(x)=\frac{1}{x}## does at ##x=0##, and like ##f(x)=\tan(x)## does at ##x=\frac{\pi}{2}+\pi n## is not a proper function over all of ##\mathbb{R}##? (really, curious.)
 
  • #9
docnet said:
Question: Does this imply that any function that has an infinity, like ##f(x)=\frac{1}{x}## does at ##x=0##, and like ##f(x)=\tan(x)## does at ##x=\frac{\pi}{2}+\pi n## is not a proper function over all of ##\mathbb{R}##? (really, curious.)
Yes. The domain of the function ##\frac 1 x## is ##\mathbb R - \{0\}##.

Note that technically this function is continuous at every point in its domain. That said, it's common to talk about a discontinuity at ##x = 0##. What that really means is that we cannot extend the function ##\frac 1 x## to be continuous on ##\mathbb R##.

We can compare this to the function ##\frac{\sin x}{x}##, which likewise is not defined at ##x = 0##, but which can be extended to a continuous function on ##\mathbb R## by defining the function to be ##1## at ##x = 0##. This is called a removable discontinuity. Although, again, it's not technically a discontinuity in the function, but a disconnectedness of the domain that is being removed.
 
  • Informative
Likes docnet
Back
Top