pykbe.misc module

pykbe.misc.accum_mat_into_tens3(mat, tensor, q, alpha=1.0, conj=False)

Accumulate a matrix into a tensor of rank 3.

Performs the operation tensor[:, q, :] += alpha * mat.

Parameters:

mat(N, M) array_like

2D array.

tensor(N, _, M) array_like

3D array.

qint

Second index of tens4.

alphascalar, optional

Scalar multiplier.

conjbool, optional

Whether to take the elementwise conjugate of mat.

pykbe.misc.accum_mat_into_tens4(mat, tensor, p, q, alpha=1.0, conj=False, beta=0.0)

Accumulate a matrix into a tensor of rank 4.

Performs the operation tensor[p, :, q, :] += alpha * mat.

Parameters:

mat(N, M) array_like

2D array.

tensor(_, N, _, M) array_like

4D array.

pint

First index of tens4.

qint

Third index of tens4.

alphascalar, optional

Scalar multiplier.

betascalar, optional

Scalar to add entrywise.

conjbool, optional

Whether to take the elementwise conjugate of mat.

pykbe.misc.fast_matexp(eigvals, eigvecs, out=None)

Compute the matrix exponential of a diagonalized matrix.

Parameters:

eigvals(N,) array_like

Eigenvalues.

eigvecs(N, N) array_like

Eigenvectors.

Returns:

matexp(N, N) ndarray

Matrix exponential.

pykbe.misc.parallel_set(arr, val)

Broadcast value to array.

Set all elements of a contiguous array to a given value. Fails if the array is not contiguous.

Parameters:

arrarray_like

Array to set.

valscalar

Value to set.

pykbe.misc.unpack_dim3_into_dim2(dest, src, q)

Unpack a slice of a 3D array into a 2D array.

Unpacks a slice of a 3D array into a 2D array. Make sure everything is row-major.

Parameters:

dest(N, M) array_like

2D array.

src(N, _, M) array_like

3D array.

qint

Second index of src.

pykbe.misc.unpack_tril(tril, symm, filltriu=True)

Unpack matrix lower triangle.

Unpacks the lower triangle of a matrix; optionally fills the upper triangle symmetrically.

Parameters:

tril(N*(N+1)/2,) ndarray

Packed lower triangle.

symm(N, N) ndarray

Unpacked matrix.

filltriubool, optional

Fill the upper triangle.