neuralqx.operators.computational.misc.numba package¶
- class NDU1Holonomy(H, edge_index, *, adjoint=False)¶
Bases:
ComputationalOperatorN-D (K-copy) holonomy on a single base edge index edge_index:
- Decomposed action:
h_1 ⊗ I ⊗ … + I ⊗ h_2 ⊗ … + … + I ⊗ … ⊗ h_K
For each gauge copy k ∈ {0, …, K-1}, produces one connected configuration σ’_k where only the k-th copy’s edge is shifted by ±1
Matrix elements are all 1.
adjoint=True flips the direction to lowering (-1), otherwise raising (+1).
- class U1Holonomy(H, edge_index, *, adjoint=False)¶
Bases:
ComputationalOperator- A pure U(1) holonomy on a single edge index edge_index:
Emits one connected configuration σ’ where that edge is shifted by ±1 (with wrap-mod).
Matrix element (weight) is identically 1.
- class MinimalLoopHolonomy(H, minimal_loop, *, copy=0, adjoint=False)¶
Bases:
ComputationalOperatorMinimal loop holonomy on a single gauge copy.
- Given a dressed minimal loop:
[ ((u, v, key), {“type”: …}), … ]
this operator applies the total shift of all loop edges on one chosen gauge copy (with wrap-mod), emitting a single connected configuration and unit matrix element
- class NDMinimalLoopHolonomy(H, minimal_loop, *, adjoint=0)¶
Bases:
ComputationalOperatorN-D minimal loop holonomy as a light-weight ComputationalOperator.
- This implements the decomposed action
h_1 ⊗ I ⊗ … + I ⊗ h_2 ⊗ … + … + I ⊗ … ⊗ h_N
Each h_k acts on a single gauge copy (k-th) by shifting the charges along the provided dressed minimal loop. The direction (+1 or -1) for each edge in the loop is deduced from the “type” entry in the dressed data.
- class OneMinusHalfHolonomy(H, minimal_loop, *, copy=0)¶
Bases:
ComputationalOperatorOperator: I - 0.5 * ( h_{a,k} + h_{a,k}^dagger )
- Acts on a single gauge copy k:
Connection 0: diagonal, σ (weight +1)
Connection 1: forward minimal-loop holonomy h_{a,k} (weight -1/2)
Connection 2: adjoint/backward minimal-loop holonomy h_{a,k}^dagger (weight -1/2)
If the dressed minimal loop is empty, this reduces to identity (one diagonal connection).
- class Number(H, edge)¶
Bases:
ComputationalOperatorDiagonal “number” operator for a single edge with U(1) DoF.
- class Identity(H, const=1.0)¶
Bases:
ComputationalOperatorDiagonal identity-like operator with an optional scalar factor const.
Exactly one connected component (diagonal).
Hermitian iff const is real.
- class Coloring(H, site, charge_vector)¶
Bases:
ComputationalOperatorCopy-resolved coloring on U(1)^gdim: δ_{(σ[site+k*nE])_k, charge_vector}
- This is a diagonal indicator on a base edge ‘site’ across gauge copies:
Works for generic gauge_dim (default use-case: U(1)^3 with gauge_dim=3).
- class ChargeColoring(H, site, charge)¶
Bases:
ComputationalOperatorSingle-site charge-coloring: δ_{σ[site], charge}.
- This is a diagonal indicator on one degree of freedom (edge-component):
(ChargeColoring at site=e, charge=q) |σ⟩ = [σ[e] == q] · |σ⟩
Returns a single diagonal connection (σ itself) with matrix element 1.0 when σ[…, site]==charge, otherwise 0.0.