>>3202

I have spent the last weeks on the binary angle without really getting anywhere. Will post some thoughts and why I think the algebra is going in circles, starting with some crumbs

Couldn't find the direct quote, but think it went something like "odd or even is the first branch of the tree"

>At most O(log t), where log t is the natural logarithm of the length of c in bits

>Because a search exponential in the length of c is turned into a calculation which is the same complexity as finding a square root < O(log l) where l is the length of c in bits

Now, the only way I can think of is that length of c in bits is relevant to the complexity of the problem is if they are a direct part of the calculation. The hint about odd and even makes sense since the LSB of a number is either 0 or 1. So the second bit (0 or 2) would be the second branch, and the third bit (0 or 4) the third… etc.

I have been trying all kinds of crap and still no closer to how this would map to a tree..

Per the latest hints, the binary patterns seem even more relevant. I generated the csv in binary and there are definitely some crazy shit going on but haven't quite figured it out (story of these threads..). I feel the difference between two squares is almost perfectly represented in base 2!

Now I haven't posted in a while due to lack of results but am here with you! We got this thing, don't go backing out now! And thank you all for keeping it going, I read every single post here

Basically I think the path we're trying to follow depends on making decisions on where to go next in the tree based on each cell we can jump to.

Also updating trip after the algo change, for when we need it in the future!

Remember the Ponies <3

>>3207

Suggest using fixed length {:b08} to {:b16} format strings etc. just so things line up - easier to find patterns that way

>>3208

And obviously I meant {:08b} :)

>>3210

Nothing comes to mind relating to primes that would seem useful, without looking at them all. Binary is pretty easy!

I have some multithreaded bruteforce code to generate arbitrary formulas and functions ready to go, and access to enough cores to run it on, but am completely stuck at creating useful verification code that isn't exponential. If we could just jump from one record to the solution it would be easy, but there could be anywhere between 100 to 100000 jumps even for small numbers.

My thinking is just creating test cases where we already have the solution and using the known results to score each function. 100% hits for even small numbers is probably gonna be right on. Ideas on verification algos (ie. try formula, generate every possible record for the result using the previous variables, and generate the rest). This easily becomes a few thousand tries per formula, which is fine, but quickly becomes insane when you add jumps. Assuming an upper bound without any data to back it up will just cause false negatives and make the results useless.

>>3211

Checked! Thanks :)

>>3212

By arbitrary formulas I mean every permutation of something like this (and more) def init(self, lit_min = -3, lit_max = 3): self.var = list("endxabc") self.lit = [i for i in range(lit_min, 0)] + [i for i in range(1, lit_max + 1)] self.op = [ast.Add(), ast.Sub(), ast.Mult(), ast.Div(), ast.Mod(), ast.Pow()] self.func = ["cos", "sin", "sqrt", "tan", "hyp", "gcd"]

>>3217

Welcome to my world! :)

There are patterns but I don't understand them, hence not posting here. Meant as inspiration iin case it triggers ideas, but could very well be the wrong tree (literally, binary numbers are a fractal tree on their own. if you want more on that just ask)!

With monospace fonts I prefer replacing 0 with space and 1 with X or | for readability, after you get over how binary works. But Chris seems to think its relevant, and if you can explain how the number or bits of a number maps to the O notation I would love to hear it. Binary search is close but not directly related to bits. Why be so explicit if its not part of our quantum computer?

>>3218

>>3217

Fractal tree (can't be arsed to convert to horizontal, so use your scroll wheel)>>for i in range(1, 50): print("{:08b}".format(i)[::-1].replace('0', ' ').replace('1', '|'))... | | || | | | || ||| | | | | | || | || | || ||| |||| | | | | | || | | | | | | || | ||| | || | || | || || || ||| | ||| |||| ||||| | | | | | || | | | | | | || | ||| | | | | | | | | | || | | || | | || | ||| | |||| | || | || >>>

>>3221

In my opinion fractals are the closest thing we have to explaining the universe with, well.. words and concepts.. Makes sense they apply to numbers too. Haven't seen any better ideas, anywhere. Then again we're in it ("the universe"), trying to understand so its not like its gonna work beyond our what our thougts tell it to anyway, like dividing by zero, except with magic :)

Will check out the links

>>3223

Glad you liked it. I still can't believe this is not common knowledge! Try telling people there are fractals in binary numbers and see what happens. Yet, one line of python here is enough :)

>>3226

>>3227

These are interesting but get back to me when we can do something with them. I've tried!