How Can I Store and Print Specific Decimal Digits of Pi Using MATLAB?

  • Thread starter Thread starter phil.st
  • Start date Start date
  • Tags Tags
    Pi
Click For Summary
The discussion focuses on a user seeking to extract and print specific decimal digits of pi, specifically from the 9901st to the 10000th, using the Brent-Salamin algorithm implemented in MATLAB. The user has successfully generated the first 10,000 digits but is struggling to isolate the desired range. Key insights include the limitation of MATLAB's default double-precision floating-point representation, which only allows for about 16 significant figures, making it inadequate for calculations beyond this range. The conversation suggests exploring MATLAB's capabilities for handling larger vectors to potentially resolve the issue of storing and printing the specific digits.
phil.st
Messages
6
Reaction score
0
Hi! I'm using MATLAB and I want to store or print some particular decimal digits of pi, from 9901 to 10000. I'm using the algorithm below (Brent-Salamin algorithm) to print the first 10000 digits but I can't find out a way to save and print only the decimal digits from 9901 to 10000. What function can I use? Any ideas?

Code:
 function P = agm_pi(d)
% AGM_PI  Arithmetic-geometric mean for pi.
% Brent-Salamin algorithm.
% agm_pi(d) produces d decimal digits.
% See Cleve's Corner, "Computing Pi",
% http://www.mathworks.com/company/ ...
%    newsletters/news_notes/2011/ 

% Copyright 2011 MathWorks, Inc.
digits(d)
a = vpa(1,d);
b = 1/sqrt(vpa(2,d));
s = 1/vpa(4,d);
p = 1;
n = ceil(log2(d));
for k = 1:n
   c = (a+b)/2;
   b = sqrt(a*b);
   s = s - p*(c-a)^2;
   p = 2*p;
   a = c;
end
P = a^2/s;

http://en.wikipedia.org/wiki/Gauss–Legendre_algorithm
 
Last edited:
Physics news on Phys.org
Well, MATLAB by default represents numbers as double-precision floating points:
http://www.mathworks.com/help/techdoc/matlab_prog/f2-12135.html

That means that you only have 53 bits for your fraction (in decimal terms, about 16 significant figures):
http://en.wikipedia.org/wiki/Double_precision

You can see where the problem arises when you try to calculate beyond 16 decimal places using the implementation provided (and no, you can't try to do something sneaky like multiplying the answer by 10^16!)

I had a similar problem as a homework assignment in undergrad once (and I suspect that this is something similar)--the above insight provided the critical hint that I needed to come up with a solution, and may do the same for you.

HINT: how big can you make a vector (1 x n, or n x1 array) in MATLAB?
 
Last edited by a moderator:

Similar threads

  • · Replies 9 ·
Replies
9
Views
5K
Replies
2
Views
1K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 17 ·
Replies
17
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 13 ·
Replies
13
Views
7K
Replies
12
Views
3K
  • · Replies 2 ·
Replies
2
Views
4K
Replies
5
Views
8K
Replies
1
Views
9K