3

I have a code that uses Gram-Schmidt inside a loop. I want to reduce the number of calls to this algorithm as much as possible, but the thing is that despite of getting the same result before and after the call, when I print the results of some operations using these values they are different. For example, on the code below the result of abs(muGS[k][0]) - abs(before2) should be 0 or very close to 0, since the printed values of this variable (before and after the call) are the same. However, this not what happens. muGS is a double matrix and its values are usually between 0 and 1. 

int k = 1;
double before2;

while(k < end) {

    before2 = muGS[k][0];

    gramSchmidt(b, muGS, cGS, k);

    //prints for debug
    if (abs(muGS[k][0]) - abs(before2) > 0.1) {

        if (abs(muGS[k][0]) - abs(before2) > 0.1)  {
            cout << "1 muGS[k] diff:" << abs(muGS[k][0]) - abs(before2) << endl;
            cout << "1 muGS[k] before:" << muGS[k][0] << endl;
            cout << "1 muGS[k] after:" << muGS[k][0] << endl;
            cout << "1 muGS[k] mult before:" << before2 * before2 << endl;
            cout << "1 muGS[k] mult after:" << muGS[k][0] * muGS[k][0] << endl;
            cout << "1 muGS[k] abs before:" << abs(before2) << endl;
            cout << "1 muGS[k] abs after:" << abs(muGS[k][0]) << endl;
        }
        getchar();
    }

    for (i = k-1; i >= 0; i--) {
        for (j = 0; j < i; j++) {
            muGS[k][j] -= round(muGS[k][i]) * muGS[i][j];
        }
    }

    //some other operations that don't change the value of muGS
    k++;
}

Output:

1 muGS[k] diff:0.157396
1 muGS[k] before:0.288172
1 muGS[k] after:0.288172
1 muGS[k] mult before:0.0171023
1 muGS[k] mult after:0.083043
1 muGS[k] abs before:0.130776
1 muGS[k] abs after:0.288172

Another thing that happens is that the absolute value of before2 is very different from the value of before2. Is it possible that I'm having some precision loss or why is this happening?

Thanks

James M
  • 18,506
  • 3
  • 48
  • 56
fc67
  • 409
  • 5
  • 17

1 Answers1

4

There is no precision loss. You just have a mistake in your code:

        cout << "1 muGS[k] before:" << muGS[k][0] << endl;
        cout << "1 muGS[k] after:" << muGS[k][0] << endl;

You print same value for both before and after. But shoulde be:

        cout << "1 muGS[k] before:" << before2 << endl;
        cout << "1 muGS[k] after:" << muGS[k][0] << endl;
Iurii
  • 56
  • 1