Faffing about with complex numbers again
Integers have these things called complex conjugates. This is something I have just been learning about. Or not. You’ll soon see my mathematical shortcomings. This thing is, I think, a way to express them as complex numbers. (I think that was the takeaway of the videos I just watched).
I’d better quote the wiki because that feels like a rubbish explanation:
In mathematics, the complex conjugate of a complex number is the number with an equal real part and an imaginary part equal in magnitude but opposite in sign.
Complex conjugate, Wikipedia
The first five look like this according to Google’s AI
IntergerComplex conjugate1122-0i33-i4455
I’m only actually interested in the first three because we could do something fun with those.
I previously showed my broken work as I wondered about the Collatz Conjecture over the complex plane. Mostly I wanted to know if complex numbers could be odd or even.
It’s possible to define even and odd Gaussian integers* in such a way that many of the familiar properties of the plain old integers are preserved, but it’s a little counterintuitive at first. The definition is that a+bi is even when a^2+b^2 is even (this is the norm of a+bi) and that a+bi is odd if it’s not even.
An answer from Mastodon that I shared here
So this got the old thinker ticking again. What if, I wondered, instead of 3x+1 or x/2, we used the complex numbers and went (3-i)x+1 and x/(2-0i) would that yield something interesting?
The Collatz Conjecture is that if you play a game where if a number is odd you times by three and add one and if even you divide by two that eventually you always get to one which loops to 4, 2, and back to 1. This is the only loop.
Test number 1 – powers of two
I asked Wolfram Alpha what 64/(2-0i) was. It is 32. Apply again and get 16. This seems to work all the way down.
Test number 2 – arbitrary odd numbers
I chose 7 – (3-i)7+1. The answer is 22-7i. That’s interesting.
Then I stopped and asked myself is 7 as a complex whatnot still odd? I asked what the complex conjugate of 7 is and, apparently, it is 7. So, I assume that’s the same as 7+0i but at this point, I clearly have no idea what I am talking about.
I’ve never let that stop me so let’s assume I am correct and have not utterly failed to understand some stuff.
7+0i is even if 7^2+0^2 is even. 49 is not even so, no 7 is not even on the complex doo-dah.
I’m not sure if this is good or bad but we have the Collatz step from 7 to 22 but with an extra -7i as baggage.
Test 3 – just run some numbers
I’m going to pick some smallish real integers and do this more complicated thing to them and see how long the real part stays correct for the simple version of Collatz.
Remember, I’m going (3-i)x+1 for complex odds and x/(2-0i) for evens.
7 -> 22-7i That’s odd for a complex number (the sum of the powers is 533)
22 should have gone to 11. Thus we have already deviated. Our new game leads us to 60-43i
And… I stopped.
First Conclusion
This game does not seem to play well over the complex plane. By “play well”, I mean do fun things.
Just out of idle curiosity, I asked what 22-7i over 2-0i was 11-3.5i. 11(3-i)+1 is 34-11i -> 17-5.5i…
Second conclusion
Well, I just got all excited over nothing. I truly hoped there was a nice pattern to find adding complex numbers to the Collatz thingy. Maybe there is but it might require writing some bad python to make a chart or something.
Ideas?
#CollatzConjecture #complexNumbers