Π day
|
03-27-2022, 01:44 PM
(This post was last modified: 03-30-2022 01:20 AM by Albert Chan.)
Post: #52
|
|||
|
|||
RE: Π day
If arugment for asin is big, taylor series take a long time to converge.
Ratio of coefficient = (2k+1)^2 / ((2k+2)*(2k+3)), approach 1 when k is huge. For asin(1), this implied dropped term is almost same size as the last kept term. We can use the "half-angle" formula for asin() (05-31-2021 07:30 PM)Albert Chan Wrote: \(\displaystyle\arcsin(x) = 2\arcsin\left( {x \over \sqrt{2\sqrt{1-x^2}+2}} \right)\) Code: function asinq(x) -- = asin(sqrt(x)) lua> asinq(1/4), pi/6 0.5235987755982989 0.5235987755982988 lua> asinq(2/4), pi/4 0.7853981633974483 0.7853981633974483 lua> asinq(3/4), pi/3 1.0471975511965979 1.0471975511965976 lua> asinq(4/4), pi/2 1.5707963267948966 1.5707963267948966 Another way is to use Carlson Elliptic Integrals: RC(1-x,1) = asin(√x)/(√x) Code: function RC(x,y, verbose) -- RC(1-x,1) = asin(sqrt(x))/sqrt(x) This is a simplified version of RF, RC(x,y) = RF(x,y,y) see https://www.hpmuseum.org/forum/thread-17...#pid148498 It has no problem getting asin(1)/1 = pi/2, 1 sqrt per iteration lua> RC(1-1, 1, true) 1.7320508075688772 1.5764775210064272 1.5711174700143078 1.5708159330462599 1.5707975451380392 1.570796402832061 1.5707963315455151 1.5707963270917837 1.5707963268134517 1.5707963267960565 1.5707963267949692 1.5707963267949012 1.570796326794897 1.5707963267948968 1.5707963267948968 Another example, from area of inscribed and circumsribed square, for pi lua> A, B = 2, 4 lua> r = (B-A)/B -- = 1/2 lua> c = RC(1-r, 1, true) -- = asin(sqrt(1/2)) / sqrt(1/2) = pi/4 * sqrt(2) 1.1147379454918027 1.1109478170877694 1.1107345982528842 1.1107215960382895 1.1107207883059862 1.110720737898786 1.1107207347495225 1.110720734552712 1.1107207345404118 1.1107207345396428 1.1107207345395949 1.1107207345395917 1.1107207345395915 lua> A * c/sqrt(1-r) 3.141592653589793 lua> A * asinq(r)/sqrt(r-r*r) 3.141592653589793 |
|||
« Next Oldest | Next Newest »
|
Messages In This Thread |
RE: Π day - Dave Britten - 03-14-2022, 11:44 AM
RE: Π day - Gerson W. Barbosa - 03-14-2022, 12:25 PM
RE: Π day - Dave Britten - 03-14-2022, 06:15 PM
RE: Π day - Gerson W. Barbosa - 03-14-2022, 11:53 AM
RE: Π day - Gerson W. Barbosa - 03-14-2022, 06:06 PM
RE: Π day - Gerson W. Barbosa - 03-14-2022, 10:30 PM
RE: Π day - Gerson W. Barbosa - 03-14-2022, 08:29 PM
RE: π day - Thomas Klemm - 03-14-2022, 09:17 PM
RE: Π day - Eddie W. Shore - 03-15-2022, 01:09 AM
RE: π day - Thomas Klemm - 03-15-2022, 07:55 PM
RE: Π day - Thomas Klemm - 03-17-2022, 03:40 AM
RE: Π day - Thomas Klemm - 03-17-2022, 03:54 AM
RE: Π day - Gerson W. Barbosa - 03-17-2022, 11:39 AM
RE: Π day - Thomas Klemm - 03-17-2022, 12:29 PM
RE: Π day - Gerson W. Barbosa - 03-17-2022, 02:10 PM
RE: Π day - Ángel Martin - 03-18-2022, 09:07 AM
RE: Π day - Frido Bohn - 03-19-2022, 09:45 AM
RE: Π day - Ángel Martin - 03-19-2022, 11:17 AM
RE: Π day - Frido Bohn - 03-19-2022, 01:01 PM
RE: Π day - Frido Bohn - 03-19-2022, 03:13 PM
RE: Π day - Steve Simpkin - 03-18-2022, 04:31 AM
RE: Π day - MeindertKuipers - 03-18-2022, 10:48 AM
RE: Π day - Ángel Martin - 03-18-2022, 11:04 AM
RE: Π day - Ángel Martin - 03-19-2022, 11:18 AM
RE: Π day - Ángel Martin - 03-20-2022, 07:39 AM
RE: Π day - Frido Bohn - 03-20-2022, 07:28 PM
RE: π day - Thomas Klemm - 03-21-2022, 07:24 AM
RE: Π day - Frido Bohn - 03-21-2022, 04:03 PM
RE: Π day - Albert Chan - 03-21-2022, 10:45 PM
RE: Π day - Gerson W. Barbosa - 03-24-2022, 01:36 AM
RE: Π day - Albert Chan - 03-26-2022, 03:59 PM
RE: Π day - Gerson W. Barbosa - 03-26-2022, 05:37 PM
RE: Π day - Thomas Klemm - 03-21-2022, 05:27 PM
RE: π day - Thomas Klemm - 03-21-2022, 05:54 PM
RE: π day - Thomas Klemm - 03-21-2022, 06:33 PM
RE: Π day - Albert Chan - 03-26-2022, 11:24 PM
RE: Π day - Albert Chan - 03-27-2022 01:44 PM
RE: Π day - Albert Chan - 03-27-2022, 04:00 PM
|
User(s) browsing this thread: 1 Guest(s)