3.1. tnpy.linalg

Functions

eigh(matrix[, k, backend])	param matrix Hermitian or real symmetric matrices whose eigenvalues and eigenvectors
eigshmv(linear_operator, v0[, k, which, tol])	param linear_operator
qr(matrix, cutoff)	param matrix Input Matrix.
svd(matrix, cutoff)	param matrix Input Matrix.

tnpy.linalg.svd(matrix, cutoff)

Parameters

• matrix (numpy.ndarray) – Input Matrix.

• cutoff (int) – Truncation dimensions.

Return type

Tuple[numpy.ndarray, numpy.ndarray, numpy.ndarray]

Returns:

tnpy.linalg.qr(matrix, cutoff)

Parameters

• matrix (numpy.ndarray) – Input Matrix.

• cutoff (int) – Truncation dimensions.

Return type

Tuple[numpy.ndarray, numpy.ndarray]

Returns:

tnpy.linalg.eigh(matrix, k=1, backend='numpy', **kwargs)

Parameters

• matrix (numpy.ndarray) – Hermitian or real symmetric matrices whose eigenvalues and eigenvectors are to be computed.

• k (int) – The number of eigenpairs to be computed.

• backend (str) – (‘numpy’ or ‘scipy’).

• **kwargs – Keyword arguments for scipy solver.

Return type

Tuple[float | numpy.ndarray, numpy.ndarray]

Returns:

tnpy.linalg.eigshmv(linear_operator, v0, k=1, which='SA', tol=0, **kwargs)

Parameters

• linear_operator (scipy.sparse.linalg._interface.LinearOperator) –

• v0 (numpy.ndarray) – Initial guesses to the eigenvectors.

• k (int) – The number of eigenpairs to be computed.

• which (str) – Which k eigenvectors and eigenvalues to find.

• tol (float) – Tolerance for eigenpairs (stopping criterion). The default value is sqrt of machine precision.

• **kwargs – Keyword arguments for primme solver.

Return type

Tuple[float | numpy.ndarray, numpy.ndarray]

Returns: