Pi and accuracy?

**oyad** · May 18th, 2003, 01:25 PM

please, how can i get an accurate value or function for Pi, VB allows me 2 use 3.14159265358979, but the error increases so much when i work with large figures?

again, given x, i want to implement inverse tangent of x

**BillyYank42** · May 18th, 2003, 06:26 PM

3.14159265358979323846264338327950288419716939937510582097494459230781640628620899862803482534211706 7982148086513282306647093844609550582231725359408128481117450284102701938521105559644622948954930381 9644288109756659334461284756482337867831652712019091456485669234603486104543266482133936072602491412 7372458700660631558817488152092096282925409171536436789259036001133053054882046652138414695194151160 9433057270365759591953092186117381932611793105118548074462379962749567351885752724891227938183011949 12.

Chris

**BillyYank42** · May 18th, 2003, 06:35 PM

Also, the tangent inverse of X could be also the following

tan^-1(x)=(i/2)log(i+z/i-z)

Chris

**siyan** · May 18th, 2003, 09:32 PM

I thought it was ArcTan(1)*4? (Radians)

But in any case VB won't allow you to use such a simple calculation to get an accurate result (i don't think...) since its accuracy is limited anyways.

Look on mathworld.wolfram.com for some Pi calculating algorithms.

**NoteMe** · May 19th, 2003, 03:20 PM

Here you got 50 000 000 digits...I am not allowed to post more then 10 000 so you have to down load them by your self...

**NoteMe** · May 19th, 2003, 03:20 PM

Here you got 50 000 000 digits...I am not allowed to post more then 10 000 so you have to down load them by your self...

http://www.cecm.sfu.ca/projects/ISC/data/pi.html

**oyad** · May 19th, 2003, 03:32 PM

Originally posted by BillyYank42
3.14159265358979323846264338327950288419716939937510582097494459230781640628620899862803482534211706 7982148086513282306647093844609550582231725359408128481117450284102701938521105559644622948954930381 9644288109756659334461284756482337867831652712019091456485669234603486104543266482133936072602491412 7372458700660631558817488152092096282925409171536436789259036001133053054882046652138414695194151160 9433057270365759591953092186117381932611793105118548074462379962749567351885752724891227938183011949 12.

Chris

please, how will these chains of numbers enter my code, i declare the variable as variant or what?.

**oyad** · May 19th, 2003, 03:38 PM

Originally posted by BillyYank42
Also, the tangent inverse of X could be also the following

tan^-1(x)=(i/2)log(i+z/i-z)

Chris

Hey, Bill , i can't make anything out of these formula (or is it an equation). where is the i and z coming from?
please, i need a code-like version that i can implement!

**oyad** · May 19th, 2003, 03:47 PM

Originally posted by NoteMe
Here you got 50 000 000 digits...I am not allowed to post more then 10 000 so you have to down load them by your self...

yea, i saw it, can you put all those in your codes?, if so, how do you do the magic?
(how old is the Pi, dates back to 5000 years ago with the Egyptians!)

**NoteMe** · May 19th, 2003, 03:56 PM

You know there are no magic formula for pi....all the formulas is just formulas that are giving close digits, and not the acurat number...

**NotLKH** · May 19th, 2003, 03:58 PM

Oyad,
exactly why do you think you need pi beyond the precision of a double? Even if you did, and you used variants, surely you wouldn't need Pi to more than 32 decimal places?

Pi to the 32nd decimal place will calculate the circumference of the universe to within an accuracy of the width of a hydrogen atom. So it would seem that anything beyond 32 decimals is a bit of an overkill.

-Lou

**jim mcnamara** · May 19th, 2003, 04:35 PM

Plus, humans cannot measure things to a precision greater than 7 significant digits, even in Physics.

This doesn't mean they can't measure at 10²⁵
it means that if they do post a measurement then only the first seven numbers in the number string mean anything.

**oyad** · May 22nd, 2003, 02:20 PM

you guys are all right, 'cos i was getting wrong results, i felt the Pi i was using was not accurate enough. i re-examined my codes and every is now okay.
but i think tan(x*(Pi/180)) should return an error?

**Fragolata** · May 23rd, 2003, 07:01 AM

Originally posted by NotLKH

Pi to the 32nd decimal place will calculate the circumference of the universe to within an accuracy of the width of a hydrogen atom. So it would seem that anything beyond 32 decimals is a bit of an overkill.

-Lou

yeah, what are you trying to do?

**oyad** · May 23rd, 2003, 02:28 PM

Originally posted by Fragolata
yeah, what are you trying to do?

i am trying to generate results for the following functions tan, sin,cos and the inverse also

**siyan** · May 23rd, 2003, 04:18 PM

Originally posted by NotLKH
Pi to the 32nd decimal place will calculate the circumference of the universe to within an accuracy of the width of a hydrogen atom. So it would seem that anything beyond 32 decimals is a bit of an overkill.

How was that figured out? I would presume that you'd have to know something about the size of the universe....

**RAEsquivelC** · Jul 12th, 2003, 10:45 PM

All BASIC-derived languages give SIN(X), COS(X), TAN(X) AND ATN(X). To get the ASIN(X) and ACOS(X), you can use the formulas:

ASIN(X) = ATN(X/SQR(1-X^2))
ACOS(X) = ATN(SQR(1-X^2)/X)

Also, remember that the answers are in radians. To obtain degrees, multiply by 180/3.14159

**NotLKH** · Jul 12th, 2003, 11:37 PM

Originally posted by siyan
How was that figured out? I would presume that you'd have to know something about the size of the universe....

I believe it was in referance to an estimated 10-15 billion light year radius of the visible universe, and I'd like to attribute that to Isaac A., but it might have been Arthur C.

-Lou

Thread: Pi and accuracy?

Thread Tools

Display

Pi and accuracy?

da chain

code?

too long ?

you're all right

clue?

Posting Permissions