Arbitrary Precision for decimals in C# help?

Question

Here is my current code for computing Pi using the chudnovsky method in c#:

using System;
using System.Diagnostics;
using System.IO;
using java.math;

namespace pi.chudnovsky
{
    public class Program
    {
        static Double Factorial(Double fact)
        {
            //begin factorial function
            if (fact <= 1)
                return 1;
            else
                return fact * Factorial(fact - 1); //loops multiplication until the factorial is reached
        }
        static Double doSummation(Double maxPower)
        {
            //begin chudnovsky summation function
            Double sum = 0;
            for (int i = 0; i <= maxPower; i++) //starts at i=0
            {
                sum += ((Math.Pow(-1, i)) * Factorial(6 * i) * (13591409 + 5451401 * i)) / (Factorial(3 * i) * Factorial(i) * Factorial(i) * Factorial(i) * Math.Pow(640320, (3 * i + 1.5))); //chudnovsky algorithm
            }
            return sum;
        }
        static void Main(string[] args)
        {
            int num;
            Console.WriteLine("Enter how many terms to compute Chudnovsky summation: ");
            //begin stopwatch
            Stopwatch stopwatch = new Stopwatch();
            stopwatch.Start();
            //parse user input
            num = Convert.ToInt32(Console.ReadLine());
            //perform calculation
            Double inv = 1 / (12 * doSummation(num));
            //stop stopwatch
            stopwatch.Stop();
            //display info
            Console.WriteLine(inv);
            Console.WriteLine("3.14159265358979323846264338327950288419716939937510");
            Console.WriteLine("Time elapsed: {0}", stopwatch.Elapsed.TotalMilliseconds);
            //write to pi.txt
            TextWriter pi = new StreamWriter("pi.txt");
            pi.WriteLine(inv);
            pi.Close();
            //write to stats.txt
            TextWriter stats = new StreamWriter("stats.txt");
            stats.WriteLine(stopwatch.Elapsed.TotalMilliseconds);
            stats.Close();
        }
    }
}

So, I've included the J# library, and included java.math. Now when I replace all the "double"s with "BigDecimal"s, I get these compile errors:

http://f.cl.ly/items/1r2X26470d0d0n260p0p/Image%202011-11-14%20at%206.16.19%20PM.png

I know that this isn't the problem with me using Int for the loops, as it worked perfectly with Doubles. My question is, how do you resolve these errors relating to int and BigDecimal, or can you recommend another arbitrary precision library?

I've tried using XMPIR, have gotten everything to compile, but I get:

http://f.cl.ly/items/1l3C371j2u3z3n2g3a0j/Image%202011-11-14%20at%206.20.24%20PM.png

So I can I use p/invoke to include xmpir so I can use whatever the bigdecimal class is?

Thank you for your time!

Answer 1

Can you not convert your int's to BigDecimal's before comparing them?

I assume you understand the problem here is that there is no operator overload for the greater, less than, etc. signs on the BigDecimal class that accepts an int.