i was trying to write Gaussian elimination method on C

Code:

float M[4][5] = 
    {
        {99.69, 7.07, 53, -4, 155.76},
        {-138.41, 14.11, -6.31, 5.6, -124.74},
        {55, 35.174, 685, 1.4, 776.574},
        {33, 21, -3.28, -0.28, 50.44}
    };


   float t_arr[5] = {0,0,0,0,0};

    for(int j = 0; j < 5; j  )
        {   
            t_arr[j] = M[2][j];
            t_arr[j] *= (-(M[3][0]/M[2][0]));
            printf("t_arr[%d] = %f\n",j, t_arr[j]);
            M[3][j]  = t_arr[j]; 
        }

Error appears here:

t_arr[j] *= (-(M[3][0]/M[2][0]));

As i wanted it to be is that third row of matrix M[2][j] == t_arr[j] That works, but after multiplying values of `t_arr[j] to (-(M[3][j]/M[2][0]))

F.e t_arr[0] * (-(M[3][0]/M[2][0])) ==>> 55 * -33\55 = -33

It Outputs:

t_arr[0] = -33.000000
t_arr[1] = -0.000000
t_arr[2] = -0.000000
t_arr[3] = -0.000000
t_arr[4] = -0.000000

CodePudding user response：

M[3][j] = t_arr[j]; in the first iteration will turn the value of M[3][0] to zero because the initial value of M[3][0] is 33 and the calculated value of t_arr[0] is -33.000000.

If you want to use the initial value of M[3][0] and M[2][0] in all iterations, you should do the calculation of -(M[3][0]/M[2][0]) and save the result before the loop.

float M[4][5] = 
    {
        {99.69, 7.07, 53, -4, 155.76},
        {-138.41, 14.11, -6.31, 5.6, -124.74},
        {55, 35.174, 685, 1.4, 776.574},
        {33, 21, -3.28, -0.28, 50.44}
    };


float t_arr[5] = {0,0,0,0,0};
float multiplier = (-(M[3][0]/M[2][0]);

for(int j = 0; j < 5; j  )
    {
        t_arr[j] = M[2][j];
        t_arr[j] *= multiplier;
        printf("t_arr[%d] = %f\n",j, t_arr[j]);
        M[3][j]  = t_arr[j]; 
    }

CodePudding user response：

After this statement in the first iteration of the loop

M[3][j]  = t_arr[j]; 

the value of M[3][0] is equal to 0 ( 33.0 -33.0 ).

So this expression (-(M[3][0]/M[2][0])) will always evaluate to 0 in subsequent iterations of the loop.