2

This post did not quite seem to answer my question. If we have $\quad a+bi\in\mathbb{C}\quad$ can ordering be done the way of Pythagorean triples $\quad A^2+B^2=C^2?\quad$

There are infinite triples for each of $\quad (C-B),\space (C-A),\space\text{and}\space (B\pm A)\quad$ but these may be further ordered as sets of sets

$ (C-B)=(2n-1)^2\in\big\{1,3,5,\cdots\big\} $

$ (C=A)=2k^2\in\big\{2,8,18,\cdots\big\} $

\begin{align*} &(B\pm A)=P:\\ &P=p_1,\cdots ,p_n),\space p_k\equiv \pm 1 \text{ (mod }8) \in\big\{1,7,17, \cdots\big\} \end{align*}

For each of $(A\text{ or } B\space\text{ or } \space C),\space$ there are $2^{n-1}\space$ primitive triples where $\space n\space $ is the number of unique prime factors of $\space(A\text{ or } B\space\text{ or } \space C)\space $ and these may be of vastly different sizes. Perhaps these may be ordered by the pairs of natural numbers $\space(n,k)\space$ that generate them in this formula

\begin{align*} A=(2n-1)^2+ & 2(2n-1)k \tag{a} \\ B= \qquad\qquad\quad & 2(2n-1)k+ ]] 2k^2 \tag{b} \\ C=(2n-1)^2+ & 2(2n-1)k+ 2k^2 \tag{c} \end{align*}

where $\space n\space $ is a set number and $\space k\space$ is the ordinal triple within each set. For example, $C=65\space$ represents triples $\space f(4,1)=(63,16,65)\quad f(5,4)=(33,56,65).\space$ where $\space(63,16,65)\space$ is the first member of $Set_4$ and $\space(33,56,65)\space$ is the fourth member of $Set_5.$

Can $\space (4+1i),\space (5+4i)\space$ be treated as natural ordering of these complex numbers or are their other considerations such as the size of the resulting squares or products to consider?

Thomas Andrews
  • 186,215
poetasis
  • 6,795
  • 3
    There is a lot of notation and language in here you haven’t explained. What do you mean by $A\lor B\lor C?$ Is that the least common multiple? What do you mean by “ordered?” If $A,B,C$ are integers, are you seeking $a,b$ integers? – Thomas Andrews Sep 19 '21 at 23:02
  • 1
    @Thomas Andrews By $\space"\lor"\space$ I meant "or". $\quad$ I will fix it. – poetasis Sep 19 '21 at 23:04
  • 2
    I don't know if this is relevant, but you should be aware that the complex numbers cannot be ordered in a way that usefully respects their additive and multiplicative structures See this article on ordered rings for the relevant definitions. – Rob Arthan Sep 19 '21 at 23:20
  • 1
    Ordering for what purpose? You can define an ordering any way you want as long as it satisfies basic rule $a\lt b$ and $b\lt c$ implies $a\lt c$. – herb steinberg Sep 19 '21 at 23:23
  • 1
    But $A\lor B\lor C$ is not a number, so it doesn’t have any prime factors. – Thomas Andrews Sep 19 '21 at 23:38
  • 1
    @herb steinberg I am not sure of what useful purpose could be served but it does seem that $9+5i$ is larger than the addends $4+i$ and $5+4i$. On the other hand, $1+100i$ does seem larger than $3+ 4i$ so, perhaps there is no magnitude ordering other that the $pairs$ I mentioned for complex numbers with the same sum-of-parts. – poetasis Sep 19 '21 at 23:38
  • 1
    @Thomas Andrews I refer to factors of A or factors of B or factors of C. – poetasis Sep 19 '21 at 23:41
  • 1
    Why exactly does (33,56,65) come after (64,16,65) in your opinion? / if you have are suggesting of family of strategies, can you give AN opinion on why (64, 16, 65) could be seen as preceding (33 , 56, 65). Your entire strategy hinges on saying (5,4) is after (4,1) because (33,56,65) comes after (64,16,65) but it is not clear to me why this ordering is reasonable. – Sidharth Ghoshal Sep 20 '21 at 00:04
  • 1
    If the answer that you give to my earlier question actually yields a partial order then sure any pair of natural numbers can ordered by it including the complex integers. – Sidharth Ghoshal Sep 20 '21 at 00:04
  • 1
    @frogeyedpeas.Since (33,56,65) and (63,16,63) have the same $C$-value, it makes sense to order them by $\space n,k\space$ which happens to match the order of the $B$-value as in

    \begin{equation} f(18,4)=(1073,264,1105)\ f(20,9)=(943,576,1105)\ f(21,12)=(817,744,1105)\
    f(23,23)=(47,1104,1105)\ \end{equation}

    – poetasis Sep 20 '21 at 00:16
  • 1
    What does ordering by $(n,k)$? mean does it mean order by $n$ and then if the two $n$-values are equal order by $k$? Does it mean order by the sum $n+k$? Like what is your exact (n,k) ordering strategy. – Sidharth Ghoshal Sep 20 '21 at 00:18
  • 1
    If we think of the Real numbers that $x < y$ means $x$ is to the left of $y$ on the number line, the we can think on the Complex numbers that $w < z$ means either that $w$ is to the left of $z$ on the complex plane, or that $w$ and $z$ are in the same horizontal position but $w$ is below $z$ on the plane. That's fine. No problem. The problem is we can't do any math on that. We can't say if $a < b\implies ac < bc$ for positive $c$. That just can't be true on the complex numbers. – fleablood Sep 20 '21 at 00:36

1 Answers1

4

There are an infinite number of ways to order the complex numbers as a set.

There is absolutely no way to order the complex numbers as a field.

I didn't bother to read your explanation of the ordering of pythogorean triples but it seems similar to "lexigraphic ordering".

For two complex number as numbers $w,z$ we can define the lexigraphic order as:

  1. If $Re(w) < Re(z)$ then $w \prec z$
  2. If $Re(w) = Re(z)$ but $Im(w) < Im(z)$ then $w\prec z$.
  3. Otherwise $w \not \prec z$.

(In other words: If $w = a+bi; z = c+di; a,b,c,d \in \mathbb R$ to compare $w$ to $z$ compare $a$ to $c$. If $a = c$ then compare $b$ to $d$.)

That satisfies everything we need to have an ordered set. To be an order on a set we need:

  1. Given complex $w$ and $z$ then exactly one of the following is true: Either $w \prec z$ or $z\prec w$ or $w = z$.

  2. if $w\prec z$ and $z\prec \alpha$ then it always follows that $w \prec \alpha$.

It's easy to see that those are both always true. So $\prec$ is an order on $\mathbb C$ as a set.

......

But it does not satisfy an order on a field.

To be an order on a field we need to be able to do math on these numbers and have the resulting inequalities be preserved. We need:

  1. If $w \prec z$ then for every $a$ we have $w + a \prec z+a$.

This is actually true.

and

  1. If $w \prec z$ and $0 \prec a$ then $aw \prec az$.

This fails. This fails big time!

Let $w = 4 - 6i$ and $z = 5-3i$ and $a = i$.

Then $4 - 6i \prec 5-3i$ and $0\prec i$ but $(4-6i)i = 6 + 4i \not\prec 3 + 5i = (5-3i)i$.

.....

Now having an order as a field is impossible.

Why. Consider any $w$ in a field. We either have $w > 0$ or $w = 0$ or $w< 0$.

If $w > 0$ then $w^2 = w\times w > 0$. If $w=0$ then $w^2 = 0$. And if $w < 0$ then $0 =w +(-w) < 0 + (-w) =-w$ and so $w^2 = (-w)\cdot (-w) > 0$.

So we have $w^2 \ge 0$ and we never have $w^2 < 0$.

But in the complex we have $1^2 = 1$ so we would have to have $1^2 = 1 > 0$. Then we'd have to have $0 < 1$ so $-1 =0+(-1) < 1 +(-1) = 0$ so $-1 < 0$.

But we also have And we have $i^2 = -1$ so we would HAVE to have $i^2 = -1 > 0$. So we have BOTH $-1 < 0$ AND $-1 > 0$. That's utterly impossible.

So the complex numbers can not be ordered as a field.

fleablood
  • 130,341
  • Your lexigraphic ordering does describe my suggested ordering by pairs and you have answered some things I never though of asking. Thanks. – poetasis Sep 20 '21 at 00:27
  • 2
    You have to be careful what you mean by order. If all you mean is that you can compare any two and one is smaller (for what ever reason you want to define it) and the other is larger then there are gazillions of ways to do that. But if you want it to have the algebraic properties we care about ... we can't. – fleablood Sep 20 '21 at 00:51