Want to play around with Pi yourself? These programs can help you get started...
A recent programming language invented by researchers at Google, called Go, is a fun way to play with Pi, because you can write and run short programs online entirely in your web browser using their "playground".
Next, we can use the simple Machin's formula to compute Pi:
First, we'll do it using native 64-bit floating point math, available in almost any programming language. The program is easy to understand, but note that it only computes Pi accurately to 16 digits!
In order to compute more digits of Pi, we have to turn to Go's support for "big" integers - those larger than 264, or 18,446,744,073,709,551,616. If we were limited to those, we'd only get about 20 digits of Pi, which you could memorize in just an hour. Far too small! Unfortunately, using big number support makes the programs a little harder to read, but it's not too bad. The only thing to remember is that a statement like sum.Div(unity, bigx) sets the value of Sum to be equal to unity / bigx.
Next, we compute the arccot using the taylor series expansion:
Machin's formula is great, but it's not that fast. A much faster way to compute Pi is to use the Chudnovsky formula. We'll build up to that in bits and pieces. In order to use it, though, we need to be able to compute square roots at high precision, something Go doesn't support natively. Onwards!
For more information about Go, check out their excellent online Go tutorial.
What better language to play with Pi than one that starts with Py? Nick Craig-Wood has written up a great series of articles about using Python to compute Pi, and the good learning site literateprograms.org offers a few as well:
Computing Pi can be useful as a benchmark of computer or program speed - or just fun. Several ways to compute Pi have been implemented in multiple languages if you want to compare across languages:
If you found some of these examples fun, you should immediately run off and check out Project Euler. Project Euler hosts an increasingly difficult series of problems that you can solve using a mix of mathematical knowledge and computer programming. They're great fun, and you might learn something if you're not careful. I (Dave) personally use the Project Euler problems as a springboard for playing around with new programming languages.