We compute the covariance by
$$\text{cov}({x}: j, k)=\frac{\sum_i(x_i^{(j)}-\text{mean}(x^{(j)}))(x_i^{(k)}-\text{mean}(x^{(k)}))}{N}$$
where $N$ is the number of vectors in the dataset and $j$ and $k$ represent which components we are interested in.