Skip to main content

QCSC Prefect SBD

Shared Python-side SBD Davidson solver integration for qcsc-prefect workflows.

This package provides:

  • SBDSolverJob as a Prefect block (block type slug sbd_solver_job)
  • SBDResult, the parsed solver output
  • create_sbd_blocks / register_sbd_block_types: create the command, execution-profile, HPC-profile and solver blocks for local, slurm, miyabi or fugaku
  • SBD input/output file handling and scheduler execution through qcsc-prefect blocks

It is intended to be reused by multiple algorithms. Installing this package does not build, download, vendor, or install the external SBD executable (diag). The executable must be compiled separately for each target HPC environment.

Source-tree native build helper assets live under:

packages/qcsc-prefect-sbd/native

They are not part of the Python package install. See native/README.md for the build scripts and how to point sbd_executable at the separately built binary.

Usage Example

from qcsc_prefect_sbd import SBDSolverJob, create_sbd_blocks

create_sbd_blocks(
    hpc_target="slurm",
    queue="normal",
    root_dir="/shared/sbd-jobs",
    sbd_executable="/abs/path/to/qcsc-prefect/packages/qcsc-prefect-sbd/native/diag",
    num_nodes=2,
    mpiprocs=4,
    mpi_options=["--mpi=pmix"],
)

# Inside a flow:
solver = await SBDSolverJob.load("sbd-solver")
result = await solver.run(
    ci_strings=(alpha_strings, beta_strings),
    one_body_tensor=h1,
    two_body_tensor=h2,
    norb=norb,
    nelec=nelec,
)
print(result.energy, result.carryover_bitstrings.shape)

create_sbd_blocks checks the request before it saves any block: local execution takes no modules or pre_commands, gpus_per_node must fit the target, solver_mode="gpu" needs resource_class="gpu", and task_comm_size x adet_comm_size x bdet_comm_size must divide the MPI rank count (num_nodes x mpiprocs). The launcher defaults to single (local), srun (Slurm), mpiexec.hydra (Miyabi) or mpiexec (Fugaku).

Closed and open shells

SBDSolverJob.run takes the same arguments as DiceSHCISolverJob.run. SBD diagonalizes in the product of the alpha and beta determinant sets with the restricted integrals, so it solves closed-shell and restricted open-shell (ROHF, nelec[0] != nelec[1]) problems.

  • When the alpha and beta sets are the same, SBD reads AlphaDets.bin for both spins.
  • When they differ (always for nelec[0] != nelec[1]), the beta set goes to BetaDets.bin (--bdetfile), and SBDResult.carryover_bitstrings_b holds the beta carryover. It is None when one set is shared.

run raises ValueError (see check_sbd_input) when:

  • the alpha set does not start with its Hartree-Fock determinant, (1 << nelec[0]) - 1, or the beta set with (1 << nelec[1]) - 1;
  • spin_sq is not None (SBD has no total spin target).

Each determinant must also have nelec[0] (alpha) or nelec[1] (beta) set bits and a set must not repeat a determinant. The FCIDUMP has no core energy, so energy is the electronic energy, as with the DICE solver.

The block refers to a CommandBlock, an ExecutionProfileBlock and an HPCProfileBlock by name. create_sbd_blocks creates all four. In the open-loop SQD workflow, sqd-create-blocks --solver sbd (algorithms/open_loop/sqd/sqd_sbd/create_blocks.py) creates them from sqd_blocks.toml as sqd-sbd-solver, cmd-sqd-sbd-solver, exec-sqd-sbd-<system> and hpc-<system>-sqd-sbd; see its README. algorithms/closed_loop/sqd_de/create_blocks.py creates them for the closed-loop workflow on Miyabi, Fugaku or Slurm.

Blocks saved before this package existed, when SBDSolverJob was defined in algorithms/closed_loop/sqd_de, keep working: the block type name (SBD Solver Job) and slug are unchanged, and carryover_type, which they lack, defaults to 1.

Metadata

Release files for qcsc-prefect-sbd 0.3.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for qcsc-prefect-sbd 0.3.1
File Size Uploaded Provenance Actions
qcsc_prefect_sbd-0.3.1.tar.gz 22.8 kB With provenance Details for qcsc_prefect_sbd-0.3.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for qcsc-prefect-sbd 0.3.1
File Interpreter ABI Platform Actions
qcsc_prefect_sbd-0.3.1-py3-none-any.whl Python 3 none any Details for qcsc_prefect_sbd-0.3.1-py3-none-any.whl

Total release size: 34.4 kB

Release files / qcsc_prefect_sbd-0.3.1.tar.gz

Download URL qcsc_prefect_sbd-0.3.1.tar.gz
Size 22.8 kB
Tags Source
SHA-256 checksum
How to use checksums
52cf141ae8eeb3f6e26b2cf282aea7f7bb17126a7aaecd97caa3a116ddb45a9d
BLAKE2b-256 checksum
How to use checksums
c223bf3e7bde5d061c3dc119358959b27c2fb5760d8f8961f37317e06b5ff303
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 10, 2026.

Transparency log

Release files / qcsc_prefect_sbd-0.3.1-py3-none-any.whl

Download URL qcsc_prefect_sbd-0.3.1-py3-none-any.whl
Size 11.6 kB
Tags Python 3
SHA-256 checksum
How to use checksums
08e589179bd1966c92dd49ec8afaabbdcb160d985f563a2f405944adece6c622
BLAKE2b-256 checksum
How to use checksums
a1b534efdc7a2c9c241c56ee1504b7ab26408448b3d99db4ea99ed1b78cbe741
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 10, 2026.

Transparency log

Security for qcsc-prefect-sbd 0.3.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

Both files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for qcsc_prefect_sbd-0.3.1.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 10, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.3.1 This release

2 release files Full provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page