# eigenvalue_decomposition¶

menpo.math.eigenvalue_decomposition(C, is_inverse=False, eps=1e-10)[source]

Eigenvalue decomposition of a given covariance (or scatter) matrix.

Parameters
• C ((N, N) ndarray or scipy.sparse) – The Covariance/Scatter matrix. If it is a numpy.array, then numpy.linalg.eigh is used. If it is an instance of scipy.sparse, then scipy.sparse.linalg.eigsh is used. If it is a precision matrix (inverse covariance), then set is_inverse=True.

• is_inverse (bool, optional) – It True, then it is assumed that C is a precision matrix ( inverse covariance). Thus, the eigenvalues will be inverted. If False, then it is assumed that C is a covariance matrix.

• eps (float, optional) –

Tolerance value for positive eigenvalue. Those eigenvalues smaller than the specified eps value, together with their corresponding eigenvectors, will be automatically discarded. The final limit is computed as

limit = np.max(np.abs(eigenvalues)) * eps


Returns

• pos_eigenvectors ((N, p) ndarray) – The matrix with the eigenvectors corresponding to positive eigenvalues.

• pos_eigenvalues ((p,) ndarray) – The array of positive eigenvalues.