Symmetry-Reduced QMB Operator Operations

Sparse operator-kernel utilities acting directly on symmetry-sector configuration tables, with optional momentum-basis projection support.

Sparse operator-application kernels inside symmetry-reduced bases.

This module builds triplet-form sparse data (row, column, value) for local and few-body operators acting on symmetry-sector configuration tables. It also supports optional momentum-basis projection through precomputed sparse factors.

edlgt.symmetries.sym_qmb_operations.nbody_term(op_list, op_sites_list, sector_configs, momentum_basis=None)[source]

Build triplets (row, col, value) for an n-body operator, optionally projected into a momentum sector.

Parameters:

  • op_list (numpy.ndarray) – Shape (n_ops, n_sites, d_loc, d_loc).

  • op_sites_list (numpy.ndarray) – Shape (n_ops,), int32 — the lattice sites the operator acts on.

  • sector_configs (numpy.ndarray) – Shape (n_states, n_sites), int32 — basis configurations in the (symmetry) sector.

  • momentum_basis (object, optional) – Momentum-projection data. If a dictionary is provided, it must contain the sparse left/right projection arrays with keys "L_col_ptr", "L_row_idx", "L_data", "R_row_ptr", "R_col_idx", and "R_data" ("n_rows"/"n_cols" may also be present). If an ndarray is provided (legacy path), it is interpreted as a dense projection basis of shape (n_states, basis_dim). If None, the operator is built in the real-space symmetry sector.

Returns:

(row_list, col_list, value_list) triplet arrays for the projected operator.

Return type:

tuple

edlgt.symmetries.sym_qmb_operations.nbody_data(op_list, op_sites_list, sector_configs)[source]

Build sparse triplets for a generic n-body operator in a symmetry basis.

Parameters:

  • op_list (numpy.ndarray) – Site-resolved operator matrices.

  • op_sites_list (numpy.ndarray) – Site indices on which the operator acts.

  • sector_configs (numpy.ndarray) – Symmetry-sector configurations, one row per basis state.

Returns:

(row_list, col_list, value_list) sparse triplet arrays.

Return type:

tuple

edlgt.symmetries.sym_qmb_operations.nbody_data_2sites(op_list, op_sites_list, sector_configs)[source]

Build sparse triplets for a two-site operator in a symmetry basis.

Parameters:

  • op_list (numpy.ndarray) – Site-resolved operator matrices.

  • op_sites_list (numpy.ndarray) – Two site indices on which the operator acts.

  • sector_configs (numpy.ndarray) – Symmetry-sector configurations, one row per basis state.

Returns:

(row_list, col_list, value_list) sparse triplet arrays.

Return type:

tuple

edlgt.symmetries.sym_qmb_operations.localbody_data_par(op, op_site, sector_configs)[source]

Build sparse triplets for a diagonal single-site operator.

Parameters:

  • op (numpy.ndarray) – Site-resolved diagonal operator matrices.

  • op_site (int) – Site index where the operator acts.

  • sector_configs (numpy.ndarray) – Symmetry-sector configurations, one row per basis state.

Returns:

(row_list, col_list, value_list) sparse triplet arrays for the diagonal operator restricted to nonzero entries.

Return type:

tuple