v0.0.85

The highlights of this release are:

• Efficient simulation of single-layer LUCJ ansatzes, contributed by @hkbelagali. See https://qiskit-community.github.io/ffsim/how-to-guides/ucj-energy.html.
• Fermionic fast Fourier transform Qiskit gate, contributed by @novitoll.
• Bug fixes and improvements in the initialization of UCJOpSpinUnbalanced from CCSD amplitudes.
• Double factorization routines in ffsim.linalg are now separated between basic and compressed functions. Enabling compression in the basic versions of the functions double_factorized and doubled_factorized_t2 is deprecated. Use double_factorized_compressed and double_factorized_t2_compressed instead. See #743.

What's Changed

• fix some bitstring bugs by @kevinsung in #699
• fix CI by @kevinsung in #701
• add jax CUDA extras to pyproject.toml by @kevinsung in #700
• bump ndarray to 0.17 by @kevinsung in #703
• add script to refresh notebook outputs by @kevinsung in #704
• add logm_unitary function by @kevinsung in #705
• Add "Create a variational ansatz" tutorial by @kevinsung in #706
• Improve handling of diagonal Coulomb indices by @kevinsung in #708
• improve random.py distributions by @kevinsung in #709
• fix eq for Qiskit gates by @kevinsung in #710
• set qiskit version cap by @kevinsung in #713
• raise ValueError in double_factorized_t2 with complex input by @kevinsung in #711
• fix FermionOperator trace overflow by @kevinsung in #714
• qiskit/gates: introduced Fermionic FFT with Cooley-Tukey alg. by @novitoll in #707
• add logm_special_orthogonal by @kevinsung in #715
• in Qiskit sim, support spinless gates in spinful circuit by @kevinsung in #716
• update Dockerfile by @kevinsung in #719
• make qiskit dependency optional by @kevinsung in #718
• use optimize_kwargs where scipy.optimize is used by @kevinsung in #720
• DOC: Add orbital optimization how-to guide by @Agnivesh-Anil-Warrier in #717
• improve orbital optimization by @kevinsung in #721
• add support for 3- and 4-RDMs by @kevinsung in #723
• add support for computing transition RDMs by @kevinsung in #725
• add protocols for SF Hamiltonian by @kevinsung in #726
• Add single-layer UCJ energy by @hkbelagali in #684
• improve UCJ energy implementation by @kevinsung in #730
• fix factor of 2 bug in UCJOpSpinUnbalanced.from_t_amplitudes by @kevinsung in #731
• support higher-order Slater RDMs by @kevinsung in #729
• add spinless UCCSD by @kevinsung in #733
• remove redundant parameters in unrestricted UCCSD by @kevinsung in #734
• validate t2 symmetry in restricted UCCSD by @kevinsung in #735
• make ucj spinless t2 match uccsd convention by @kevinsung in #736
• validate t2 amplitudes in UCJ by @kevinsung in #737
• change n_reps behavior in UCJOpSpinUnbalanced.from_t_amplitudes by @kevinsung in #738
• add unbalanced section to LUCJ notebook by @kevinsung in #740
• rewrite prose in UCJ energy notebook by @kevinsung in #741
• always optimize double factorization when optimize=True by @kevinsung in #742
• compute Slater RDMs by Laplace expansion by @kevinsung in #732
• split out real matrix parametrization by @kevinsung in #744
• ucj energy: skip two-body terms related by symmetry by @kevinsung in #745
• split double factorization into exact and compression functions by @kevinsung in #743

New Contributors

• @novitoll made their first contribution in #707
• @Agnivesh-Anil-Warrier made their first contribution in #717
• @hkbelagali made their first contribution in #684

Full Changelog: v0.0.84...v0.0.85