#include <stdio.h>
#define eps 0.00000001

double A, B=((1<<31)-1), DMin=((1<<31)-1);
int N;

void bs()
{
        double p1 = A, p2 = 0, mij, a1, a2, a3, min;
        int i;

        while ((p1-p2)>eps)
        {
                min = A;
                mij  = (p1+p2)/2;
                a1 = A; a2 = mij;
                for (i = 2; i < N; i++)
                {
                        a3 = 2*(a2+1)-a1;
                        a1 = a2;
                        a2 = a3;
                        if (a3 < 0)
                        {
                                p2 = mij+eps;
                                break;
                        }
                        else
                            if (min > a3) min = a3;
                }
                if (i == N)
                {
                   if ((B-a3)<eps) p2=2*p1;
                   if ((DMin-min) > eps) DMin = min, B = a3;
                   p1 = mij-eps;
                }
        }
}

int main()
{
        scanf("%d %lf", &N, &A);
        bs();
        printf("%.2lf\n", B);

        return 0;

}

Can you just explain in simple terms the method you've used.

Edited by author 01.12.2007 02:29