Write calculate_covariance_matrix(vectors) that returns the sample covariance matrix of a set of features.

• vectors is a list of lists; inner list $i$ holds the $N$ observations of feature $i$ (all features have the same $N \ge 2$).
• With $\bar{x}_i$ the mean of feature $i$, the entry for features $i, j$ is $\text{cov}(i,j) = \frac{1}{N-1}\sum_{k=1}^{N}(x_{ik}-\bar{x}_i)(x_{jk}-\bar{x}_j)$ (divide by $N-1$, not $N$).
• Return a list of lists of floats of shape (n_features, n_features); the matrix is symmetric and the diagonal holds the sample variances.