Author Topic: Surreal Numbers, and my slightly nervous connection with them.  (Read 20629 times)

0 Members and 1 Guest are viewing this topic.

Offline Quantum Burrito

  • K.C. Munchkin Maestro
  • ****
  • Posts: 948
  • Up, Down, Strange, Charm, Top, Bottom, Burrito
    • View Profile
Re: Surreal Numbers, and my slightly nervous connection with them.
« Reply #30 on: January 14, 2008, 03:13:42 pm »
Actually, I might as well tell you the formula for multiplication, so you can check your answers:

ab = {ALb+aBL−ALBL, ARb+aBR−ARBR | ALb+aBR−ALBR, ARb+aBL−ARBL}

(where AL represents the left set of a, etc.)
Quote from: The Will
In a movie, one can always pull back and condemn the character or the artist ... But in playing a game ... we can be encouraged to examine our own values by seeing how we behave within virtual space.
tl;dr. But with all those shiny symbols and numbers it's probably awsome.

Offline stuck

  • Robotron Automaton
  • *****
  • Posts: 1876
  • Who is John Galt?
    • View Profile
Re: Surreal Numbers, and my slightly nervous connection with them.
« Reply #31 on: January 14, 2008, 03:23:24 pm »
Well, yeah, if you do anything but science and maths this is rather trivial. But Quantum Mechanics has been taking a lot from pure maths, specifically linear algebra and abstract algebra, and knowledge of quantum mechanics without that is extremely shallow.

I suppose we can show 1/3 like so. *This is a preliminary idea, so it is probably not very rigorous, though I think I've done a good job* Similar to the way you defined ε, we define a series greater than n, NG, such that its tail tends toward n. Note, however, that n is not in the series, it just gets infinitesimally  close. Then we construct another series, NL, less than n such that its tail becomes infinitesimally close. Again, n is not in NL.

Using these two series, we construct the surreal number {NL| NG}. Since n is between NL and NG, the surreal number must be n.

Still, there must be a "cleaner" way to define the N series. Since we know all numbers of the form 1/2k*a can be generated by QB's previous construction of ε, there are two numbers of this set that must be infinitesimally above and below n, meaning the surreal number is n.

I'll work on that multiplication.


Offline Quantum Burrito

  • K.C. Munchkin Maestro
  • ****
  • Posts: 948
  • Up, Down, Strange, Charm, Top, Bottom, Burrito
    • View Profile
Re: Surreal Numbers, and my slightly nervous connection with them.
« Reply #32 on: January 15, 2008, 12:24:45 pm »
I'l like to drop in something else I have found (well, reconstructed from drips and drabs on the internet):

Surcomplex Numbers

The Surcomplex numbers are a proper superset (i.e., they contain) the Complex numbers. Also, 'surcomplex' is an awesome word.
Anyway, as you may or may not know, the complexes are formally defined as an ordered pair of reals, (r,i).

These ordered pairs are defined in set theory as {{r},{r,i}} (as are all ordered pairs). However, the arithmetic operations on them must be defined.

Addition:
(a,b)+(c,d)=(a+c,b+d)

Multiplication:
(a,b)*(c,d)=(ac-bd,bc+ad)

The inverses can be derived from these equations. As can exponentiation, etc.

Now, although the Surreals on their own cannot construct the square root of -1 (try it from the definition of multiplication I gave, you'll end up needing a set that can multiply with the empty set without becoming the empty set, which is impossible), neither can the reals. Bit using the same ordered pair methodology, but with Surreal numbers, a set of Surcomplex numbers can be defined:

(0,1)*(0,1)=(0*0-1*1,1*0+0*1)=(-1,0)

Any ordered pair with 0 as the right projection is a Surreal number, so (0,1) times itself in the Surcomplexes is -1. Just as in the Complexes, and so we shall label this number i.

Best thing about this, is we now have complex infinites and infinitesimals.

(ε,0)*(0,ω)=(ε*0-0*ω,0*0+εω)=(0,1)=i
(ε,4ω)*(1,2)=(ε-8ω,4ω+2ε)=(-8ω+ε,4ω+2ε)=(-8ω+ε)+(4ω+2ε)i
(0,ω)*(ω,0)=(0*ω-ω*0,ω*ω+0*0)=(0,ω2)=ω2i

And, Seeing as the complexes remain unchanged as a subset of the Surcomplexes, eπi+1=0. So, now all we need to do is find an identity that includes ω and ε as well!
Quote from: The Will
In a movie, one can always pull back and condemn the character or the artist ... But in playing a game ... we can be encouraged to examine our own values by seeing how we behave within virtual space.
tl;dr. But with all those shiny symbols and numbers it's probably awsome.