Here is the $C$++ code that my compiler cannot process as the array size is 'too large'. I tried some online compilers and still I cannot get the product. I occasionally think about this problem on weekends. O dear, it is tormenting me.

```
#include<iostream.h>
#include<math.h>
int main()
int main()
{
/* Constructing the auxiliary table for enumerating the sign matrices*/
int Table[65536][16];
int k, j, i;
for (k = 0; k < 16; k++) // number of columns
{
for (j = 0; j < pow(2,(k + 1)); j++) // number of times sign alternates in a column
{
for (i = 0; i < 65536; i++) // number of rows
{
if (i >= pow(2,(16-k-1))*j && i < ( pow(2,(16-k-1))*(j + 1)) )
Table[i][k] = pow(-1,j);
}
}
}
```

/* Displaying the table*/
`/*N.B. Except for checking some rows, it is unwise to print the table as there are 65536 rows.*/`

```
for (i = 0; i < 65536; i++)
{
for (k = 0; k < 16; k++)
{
cout << Table[i][k] << " ";
}
cout << endl;
}
```

/*Computing $M_4$*/

//Entering the first sign matrix $M_{(4,1)}$

```
long double a = Table[0][0], b = Table[0][0], c = Table[0][0], d = Table[0][0],
e = Table[0][0], f = Table[0][0], g = Table[0][0], h = Table[0][0],
l = Table[0][0], m = Table[0][0], n = Table[0][0], o = Table[0][0],
p = Table[0][0], q = Table[0][0], r = Table[0][0], s = Table[0][0];
long double a1, b1, c1, d1, e1, f1, g1, h1, l1, m1, n1, o1, p1, q1, r1, s1;
```

`/*Formula for multiplying the matrices in the prescribed order*/`

```
for (i = 1; i < 65536; i++) // i starts from 1 since the first row has already been entered
{
a1 = a * Table[i][0] + b * Table[i][4] + c * Table[i][8] + d * Table[i][12];
b1 = a * Table[i][1] + b * Table[i][5] + c * Table[i][9] + d * Table[i][13];
c1 = a * Table[i][2] + b * Table[i][6] + c * Table[i][10] + d * Table[i][14];
d1 = a * Table[i][3] + b * Table[i][7] + c * Table[i][11] + d * Table[i][15];
e1 = e * Table[i][0] + f * Table[i][4] + g * Table[i][8] + h * Table[i][12];
f1 = e * Table[i][1] + f * Table[i][5] + g * Table[i][9] + h * Table[i][13];
g1 = e * Table[i][2] + f * Table[i][6] + g * Table[i][10] + h * Table[i][14];
h1 = e * Table[i][3] + f * Table[i][7] + g * Table[i][11] + h * Table[i][15];
l1 = l * Table[i][0] + m * Table[i][4] + n * Table[i][8] + o * Table[i][12];
m1 = l * Table[i][1] + m * Table[i][5] + n * Table[i][9] + o * Table[i][13];
n1 = l * Table[i][2] + m * Table[i][6] + n * Table[i][10] + o * Table[i][14];
o1 = l * Table[i][3] + m * Table[i][7] + n * Table[i][11] + o * Table[i][15];
p1 = p * Table[i][0] + q * Table[i][4] + r * Table[i][8] + s * Table[i][12];
q1 = p * Table[i][1] + q * Table[i][5] + r * Table[i][9] + s * Table[i][13];
r1 = p * Table[i][2] + q * Table[i][6] + r * Table[i][10] + s * Table[i][14];
s1 = p * Table[i][3] + q * Table[i][7] + r * Table[i][11] + s * Table[i][15];
a = a1, b = b1, c = c1, d = d1, e = e1, f = f1, g = g1, h = h1, l = l1,
m = m1, n = n1, o = o1, p = p1, q = q1, r = r1, s = s1;
}
```

/* Displaying the final product, i.e. $M_4$*/

```
cout << endl << endl;
cout << "\t\t" << a << "\t" << b << "\t" << c << "\t" << d << "\t" << endl;
cout << "M_4 =\t" << e << "\t" << f << "\t" << g << "\t" << h << "\t" << endl;
cout << "\t\t" << l << "\t" << m << "\t" << n << "\t" << o << "\t" << endl;
cout << "\t\t" << p << "\t" << q << "\t" << r << "\t" << s << "\t" << endl;
```

return 0;

}

sign pattern matrices, that is, matrices with entries in the set of three symbols $\{-,0,+\}$. They are meant to model applications in which only the sign of the entries of a matrix is known; in some cases, this is sufficient to deduce (non)singularity, rank, or irreducibility properties. A good starting point is chapter 33 inHandbook of Linear Algebraby Hogben. $\endgroup$ – Federico Poloni Sep 29 '10 at 12:20