MeasurementEfficientEnergyEstimatorDataCode

Scripts and data for the paper Measurement-efficient energy estimator for quantum chemistry of photodynamic materials.

Scripts

• Top-level make_*.py scripts generate plots and figures for the paper.
• fig*/ contains additional figure-specific plotting scripts.
• ai_smcs/ contains scripts for the AI-assisted SMCS data analysis; see its README for details.
• compute_uncompute.py and classical_shadows.py analyze hardware-shot data for the compute-uncompute and classical-shadow methods, respectively.
• guppy_overlaps_protocol.py builds overlap circuits for different protocols.
• qsys_result_sampling.py supports processing Qsys results, including leakage filtering, bootstrap resampling, and JSON input/output.
• selene_backend.py provides a local emulator function that takes a program and can run (noisy) shots.

Shared utilities

• utils.py contains shared utility functions for processing circuits and energies.
• inquanto_utils.py converts InQuanto objects into forms that can be used by the post-processing workflows.

Data

• data/circuit_resources/ contains compiled, optimized and transpiled resource estimates for state-preparation circuits at different truncations.
• data/energies/ contains CASCI, CS (classical shadows), CU (compute-uncompute), and SV (state-vector) energies used in the paper figures.
• data/hamiltonians/ contains the Hamiltonians used in the paper.
• data/<geo><cas>/ contains circuits and shot tables for the hardware runs, grouped by geometry and active-space size (for example, rx66 and t11010).
• data/pauli_averaging.json records the number of shots estimated to reach chemical accuracy with Pauli averaging (PA).
• data/tups_params.json contains the optimized tUPS parameters from VQE.

Some files make reference to the state "chi". This has been renamed to "lambda" in the manuscript.

Load a Hamiltonian into Zixy

The six .npz files in data/hamiltonians/ contain NumPy integral arrays and Pauli expansions. load_hamiltonians.py reads only constant, one_body, and two_body from an .npz file and constructs a normal-ordered Zixy fermion RealTermSum (Zixy >= 0.4.0), without InQuanto. Spatial integrals use chemist's notation, with the two-body factor of 1/2 and interleaved up/down spin orbitals:

from load_hamiltonians import load_hamiltonian

hamiltonian = load_hamiltonian("data/hamiltonians/ham_Rx_geom_triplet_spin_shciscf_basis_66_6e_6o.npz")