Projectors#

class Projectors(Plist, Pstruct, force_general=False)[source]#

Bases: object

Class to handle sparse shared projectors.

Parameters:

Plist (Sequence[ArrayLike]) – List of sparse projector matrices.
Pstruct (sp.csr_array) – Sparsity structure of the projectors.
force_general (bool, optional) – If true, treat all projectors as general sparse matrices even if diagonal.

Methods Summary

`allP_at_v`(v[, dagger])	Compute all P_j @ v (or P_j^† @ v) and return an (n, k) matrix.
`append`(Pnew)	Append a new projector.
`erase_leading`(m)	Remove the first m projection matrices.
`get_Pdata_column_stack`()	Extract all sparse P_j entries according to Pstruct.
`is_diagonal`()	Return True if all projectors are diagonal.
`set_Pdata_column_stack`(Pdata)	Set projectors from column-stacked sparse entries.
`validate_projector`(P)	Check if P is a valid projector (correct shape, subset of Pstruct).
`weighted_sum_on_vector`(v, weights[, dagger])	Compute Σ_j weights[j] * P_j^(†) @ v efficiently without forming Σ_j P_j.

Methods Documentation

allP_at_v(v, dagger=False)[source]#

Compute all P_j @ v (or P_j^† @ v) and return an (n, k) matrix.

Returns a matrix whose j-th column is P_j v (dagger=False) or P_j^† v (dagger=True). For diagonal projectors, dagger reduces to conjugation:

allP_at_v(v, dagger=True) == (Pdiags.conj().T * v).T (shape (n, k)).

Parameters:

Return type:

ndarray[tuple[int, …], dtype[complexfloating]]

append(Pnew)[source]#

Append a new projector.

Parameters:: Pnew (_Buffer | _SupportsArray[dtype[Any]] | _NestedSequence[_SupportsArray[dtype[Any]]] | bool | int | float | complex | str | bytes | _NestedSequence[bool | int | float | complex | str | bytes])
Return type:: None

erase_leading(m)[source]#

Remove the first m projection matrices.

get_Pdata_column_stack()[source]#

Extract all sparse P_j entries according to Pstruct.

Orders as columns of a (nnz,k) matrix. Returns a matrix whose j-th column is P_j[Pstruct]

is_diagonal()[source]#

Return True if all projectors are diagonal.

set_Pdata_column_stack(Pdata)[source]#

Set projectors from column-stacked sparse entries.

Use columns of Pdata as the sparse entries of each P_j with the current Pstruct.

Parameters:: Pdata (ndarray[tuple[int, ...], dtype[complexfloating]] | sparray)
Return type:: None

validate_projector(P)[source]#

Check if P is a valid projector (correct shape, subset of Pstruct).

weighted_sum_on_vector(v, weights, dagger=False)[source]#

Compute Σ_j weights[j] * P_j^(†) @ v efficiently without forming Σ_j P_j.

# Returns a vector of shape (n,).

Parameters:

Return type:

ndarray[tuple[int, …], dtype[complexfloating]]