Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon.

Pages: 1-

Calculation of pi

Name: Anonymous 2020-07-12 7:25

Continuation of /sci thread https://boards.4channel.org/sci/thread/11876020/thoughts

Do you think it's a good approach /prog?

Name: Anonymous 2020-07-12 7:37

ah, wanna see something interesting?

k/m = 0.414213562373
m/k = 2.414213562373

I was using 0.5

Name: Anonymous 2020-07-12 7:45

Its a trollpost for retards that believe that 0.999...=1 and that infinitesimals don't exist(spoiler: they are).
They discard tiny infinitesimal segments at the edges.

Name: Anonymous 2020-07-12 7:48

It's 2-sqrt(2) / sqrt(2)

Name: Anonymous 2020-07-12 7:50

Lets call it a subset of the Inverse fractional dubs theory

Name: Anonymous 2020-07-12 7:57

>>3
OP pic was borrowed from a trollish pi=4 thread, don't mind it too much

Name: Anonymous 2020-07-12 8:32

It's in the same group as the golden ratio (-also an Inverse fractional dub)

They might be constructable too?

Name: Anonymous 2020-07-12 8:45

0.302775637.. is an artificial one

Name: Anonymous 2020-07-12 8:52

HOME BUTTON will take you to the top of the page. Do not need a floating blue arrow button that obscures the webpage content. You are welcome.

Name: Anonymous 2020-07-12 8:54

Sort of continued fraction related maybe

x = 1/x +- some integer i

Name: Anonymous 2020-07-12 8:58

Convergent fraction seems like a good description, but alas

Name: Anonymous 2020-07-12 9:34

So anyway, how to factor in the .4142 to the (1,-1):(1,0) - slopedist(x)=sqrt(1+x^2)

Name: Anonymous 2020-07-12 11:00

I have once wrote one in Symta. It worked the same way you trace fractals - by recursively descending to measure arc size. So in a sense a circle is a fractal too. Just a very round one.

Name: Anonymous 2020-07-12 12:08

>>6
Both cases involve infinitesimal segments hidden by "close fit" scaling, which preserve specific non-pi perimeter,like (1-0.999...) scaled by N propositional to (1-1/10^N of 0.999...).

Name: Anonymous 2020-07-12 12:36

>>13
It might have to branch to count, or probably at least change scale and direction

Name: Anonymous 2020-07-12 12:42

22/7

Name: Anonymous 2020-07-12 12:49

>>16
This is what 16-year-olds think ``based'' is.

Name: Anonymous 2020-07-12 12:52

Not quite
>0.5
(1 - 0.4142) / 0.5
1:1.1716

<0.5
(0.4142) / 0.5
1:0.8284

Name: Anonymous 2020-07-12 13:01

*second order deriv?

These coordinates are pretty loose so it might work

Name: Anonymous 2020-07-12 17:09

>>15
it rotate the vector by a/2 on each iteration, then re-normalized it and rotate again. No sin/cos are needed for 1/2 rotation starting from PI/2 angle

Name: Anonymous 2020-07-12 17:11

>>20
BTW, one can also calculate sin and cos using the same algorithm.

And the "Pi" will be different if the function metric picked is not euclidean.

Name: Anonymous 2020-07-12 17:11

>>21
In fact, I think for actual fractals PI can be infinite.

Name: Anonymous 2020-07-15 3:09

There are many ways to calculate π.

This is adequate for up to a few thousand digits (but really slows down in that range). https://files.catbox.moe/hwojk9.c

Name: Anonymous 2020-07-15 15:46

Name: Anonymous 2020-07-16 4:52

>>24
Bisecting a chord also bisects the circle's arc, just re-norm the vector with each split. Nice!

Name: Anonymous 2020-07-17 12:28

x=0
rep: x = 1/(x+2)
sqrt(2) = x+1

Don't change these.
Name: Email:
Entire Thread Thread List