I am assuming that you have a 2D array with 55 rows and 2 columns (as you have mentioned 110 values value of data above). First, you have to put this data in a file, say test.dat, (download from here) then you can read it in a gsl_matrix and calculate the covariance matrix using the gsl_stats_covariance function as follows:
#include <stdio.h>
#include <gsl/gsl_statistics.h>
#include <gsl/gsl_matrix.h>
int main (void)
{
int i, j;
int rows = 55, col = 2;
gsl_matrix * m = gsl_matrix_alloc (rows, col);
gsl_matrix * C = gsl_matrix_alloc (col, col);
{
FILE * f = fopen ("test.dat", "rb");
gsl_matrix_fscanf (f, m);
fclose (f);
}
for (i = 0; i < m->size2; i++)
{
for (j = 0; j < m->size2; j++)
{
gsl_vector_view col1, col2;
col1 = gsl_matrix_column (m, i);
col2 = gsl_matrix_column (m, j);
double cov = gsl_stats_covariance(col1.vector.data, col1.vector.stride,
col2.vector.data, col2.vector.stride,
col1.vector.size);
gsl_matrix_set (C, i, j, cov);
}
}
for (i = 0; i < C->size1; i++)
{
for (j = 0; j < C->size2; j++)
{
printf("%f ", gsl_matrix_get(C, i, j));
}
}
gsl_matrix_free (m);
gsl_matrix_free (C);
return 0;
}
You can see that the C matrix has been initialized as a 2x2 matrix because the covariance matrix is a square matrix. Each column of the matrix m is sliced as gsl_vector_view and utilized in the gsl_stats_covariance function. In the end, the covariance matrix is printed.
