4
$\begingroup$

Are there short, common, and intuitive notations for intervals of real and natural numbers that explicitly use $\mathbb R$ and $\mathbb N$ as basis of their notation?

The main purpose of the notation is denote and distinguish between discrete and continuous intervals.

Currently, based on the following notation examples

$$ \mathbb{R}^{>a}=\{x\in\mathbb{R} \mid x>a\} \\ \mathbb{R}^{\neq a}=\{x\in\mathbb{R} \mid x\neq a\} \\ \mathbb{N}^{+}=\{x\in\mathbb{N} \mid x>0\} $$

I would like to define intervals of real numbers and natural positive numbers -- and explicitly distinguish between real and natural numbers -- as follows.

$$ \mathbb{R}^{[0,1]}=\{x\in\mathbb{R} \mid 0\leq x \leq 1\} \\ \mathbb{N}^{[1,w]}=\{x\in\mathbb{N} \mid 1\leq x \leq w\} $$

I prefer this notation over $[0,1]$ and $\{1,2,...,w\}$, since it is short, emphasizes the distinction between $\mathbb R$ and $\mathbb N$, and allows me to use the same interval notation for both types of numbers. I also regard it as more explicit than, e.g., $[0,1]$, which may be misinterpreted as $\{0,1\}$ by the occasional non-mathaffine reader, especially since $0$ and $1$ represent the possible values of a bit.

Currently, I would define this notation at the beginning of my thesis and do not use it for other purposes, such as "the set of all functions from the set $[0,1]$ into the set $R$" [1].

I also use the notation for defining two-dimensional ranges as follows.

$$ \mathbb{N}^{[1,w]}\times \mathbb{N}^{[1,h]} $$

Are there better, i.e., short, well-known, and intuitive interval notations for the described purposes?

Edit (considering other alternatives):

Goal. I want to denote, i.e., name the following sets to repeatedly refer to them in my thesis (see new naming ideas in preceding braces).

  • The set natural numbers $\{1,2,...,w\}$ -- ($\mathbb N _w$)
  • The set natural numbers $\{1,2,...,h\}$ -- ($\mathbb N _h$)
  • The range of real numbers $[0,w]$ -- ($\mathbb R _w$)
  • The range of real numbers $[0,h]$ -- ($\mathbb R _h$)
  • The range of real numbers $[0,1]$ -- (just use $[0,1]$, as $\mathbb R _{1}$ would be too confusing)
  • The set of two-tuples $(x,y)$ with $x\in [0,w], y\in[0,h]$ -- ($\mathbb R _w \times \mathbb R _h$)
  • The set of two-tuples $(x,y)$ with $x\in \mathbb N \cap [1,w], y\in \mathbb N \cap [1,h]$ -- ($\mathbb N _w \times \mathbb N _h$)

Moreover, I also just found $[1.. w]$ to denote integer intervals, which seems much more convenient than $\mathbb N \cap [1,w]$, i.e., is easier to read, as it does not require you to process the set operation.

Options. As a result, my notation options are the following (presented as example text, to allow for evaluation of readability)

  1. This option uses $\mathbb N \cap [1,w]$ for integers, $[0,w]$ for real numbers, and eventually $\mathbb N \cap [1,w] \times \mathbb N \cap [1,n]$ for 2D integer intervals.
  2. This option uses $[1..w]$ for integers, $[0,w]$ for real numbers, and eventually $[1..w] \times [1..n]$ for 2D integer intervals.
  3. This option uses $\mathbb N _w$ for integers, $\mathbb R _w$ for real numbers, and eventually $\mathbb N _w \times \mathbb N _h$ for 2D integer intervals.
  4. This option uses $ N _w$ for integers, $ R _w$ for real numbers, and eventually $ N _w \times N _h$ for 2D integer intervals.

Evaluation

  • Option 1 is hardly readable (does not easily convey the message).
  • Options 2 to 4 are OK.
  • Options 3 and 4 are a little more readable (but need to introduced once).
  • Option 4 moreover avoids redefining existing interpretations of $\mathbb N _?$ and $\mathbb R _?$.

Question: Did I overlook something, i.e., what are the potential problems of Options 2, 3, and 4?

$\endgroup$
  • $\begingroup$ But they are typographically inconvenient... $\endgroup$ – Megadeth Sep 11 '15 at 8:58
  • $\begingroup$ Try $\mathbb{N} \cap [1, w]$. $\endgroup$ – Megadeth Sep 11 '15 at 8:58
  • 3
    $\begingroup$ $\Bbb R^{[0,1]}$ already has a meaning; it's the set of functions from $[0,1]$ to $\Bbb R$. (I think that $[a,b]_A$ should be a notation for $[a,b]\cap A$, but I doubt that would ever be widely accepted.) $\endgroup$ – Akiva Weinberger Sep 11 '15 at 11:33
1
$\begingroup$

My currently preferred notations are listed below. However I have not understood your general problem yet, which might deserve a new notation.

$$ \mathbb{R}^{>a}=\{x\in\mathbb{R} \mid x>a\} $$

$(a, \infty)$

$$ \mathbb{R}^{\neq a}=\{x\in\mathbb{R} \mid x\neq a\} $$

$\mathbb{R} \setminus \{ a \}$

$$ \mathbb{N}^{+}=\{x\in\mathbb{N} \mid x>0\} $$

If $\mathbb{N}$ starts at $1$: $\mathbb{N}$ in contrast to $\mathbb{N}_0 = \mathbb{N} \cup \{ 0 \} $

If $\mathbb{N}$ starts at $0$: $\mathbb{N} \setminus \{ 0 \}$

$$ \mathbb{R}^{[0,1]}=\{x\in\mathbb{R} \mid 0\leq x \leq 1\} $$

$[0,1]$ or maybe $\mathbb{R} \cap [0,1] = [0,1] \cap \mathbb{R}$. If I see nothing mentioned I usually assume $\mathbb{R}$ as base set, like with the first case.

$$ \mathbb{N}^{[1,w]}=\{x\in\mathbb{N} \mid 1\leq x \leq w\} $$

$\mathbb{N} \cap [1,w] = [1,w] \cap \mathbb{N}$

$\endgroup$
1
$\begingroup$

I see sometimes $[\![1,n]\!]$ instead of $\{1,\dots,n\}.$

$\endgroup$
  • $\begingroup$ OK, but like $[1..n]$ this may not be known to the untrained reader and may be easily misread as a real numbered interval $[1,n]$. $\endgroup$ – Juve Jan 31 '16 at 12:03
  • $\begingroup$ In France it is often used in probability books : I agree that it could be an "official" notation only in this particular context. $\endgroup$ – Balloon Jan 31 '16 at 12:29
0
$\begingroup$

In my document, I now use the following notations and conventions.

  1. I use $[1,w]$ for real numbered intervals and $\{1,2,...,w\}$ for integer intervals.

  2. If I reuse a specific interval several times, I introduce and name it once, e.g., as $\mathbb{N}_w=\{1,2,...,w\}$, and then refer to it using its name, e.g., $\mathbb{N}_w$.

  3. Thereby, throughout the document, $\mathbb{N}_n$ always denotes an integer interval from $1$ to $n$, while $\mathbb{R}_n$ always denotes a real numbered interval between and including $0$ and $n$.

  4. However, I do not use this notation unless having introduced each specific $\mathbb{N}_?$ and $\mathbb{R}_?$ explicitly.

Using short names for my common intervals, I can now easily and consistently denote multidimensional ranges, e.g., $\mathbb{R}_w \times \mathbb{R}_h$, or express membership, e.g., $x \in \mathbb{N}_w, y \in \mathbb{N}_h$.

I believe that using a subscript index for $\mathbb{R}$ and $\mathbb{N}$ is safe enough and creates the least confusion with other common notations. I now prefer the longer notation $\{1,2,...,n\}$ over $[1..n]$ as it is more common, and over $\mathbb{N}\cap[1,n]$ as it does not require the reader to process the expression, i.e., as it is more explicit.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.