The title says it all!
(BigDecimal is a way to get more decimals than you do when using double or float)
I think it's a bit confusing, and there might be other uses of the Bigdecimal class, but here is one of them.

@tylers source is there now.
@mjrb4 i added the calculation to the act method, because I was worried about crashing if I used a while or for command.
@SPower you're right.

This is a BigDecimal demo, and I kinda like the way it calculates slowly.
I know that this page is in danish, but if you really wont the value of pi check the bottom of this pdf: http://www.matematiksider.dk/pi/tallet_pi.pdf

A new version of this scenario was uploaded on Wed Jul 04 17:06:26 UTC 2012
Made the calculation a little faster.
I'm abit slow, but I found out that the size of the BigDecimal scale, made the execute run slower if it was too high.
(I changed the scale from 10000 - I thought it would be more precise- to 120).
The program is still calculating slowly.

@SPower the way you calculate the arctan function is this arc(x) = x - x^3/3 + x^5/5 - x^7/7
Its just that pi/4 = arctan(1) or 1-1/3+1/5-1/7. So i think Zamoht's method is faster.
@Zamoht, I am impressed by your use of BigDecimals. I tried doing so but I had trouble with it. I will use some your code in my series formula scenario (I will give you credit and like your scenario).

Also, I think the only reason it takes so long to calculate pi is because you are only making one calculation per act cycle. I tried it out using a for loop to calculate 100 times per act frame and the laptop I'm using handled it like pi (pun intended).

2012/7/4

2012/7/4

2012/7/4

2012/7/4

2012/7/4

2012/7/4

2012/7/4

2012/7/4