Branch: refs/heads/somo-dev Home: https://github.com/ehb54/ultrascan3 Commit: e0d57bc99cd1ff7d35f3b5a0a03c310b4fd4745d https://github.com/ehb54/ultrascan3/commit/e0d57bc99cd1ff7d35f3b5a0a03c310b4... Author: ehb54 <brookes@uthscsa.edu> Date: 2026-07-18 (Sat, 18 Jul 2026) Changed paths: A us_somo/develop/grpy/INTEGRATION.md A us_somo/develop/grpy/README.md A us_somo/develop/grpy/grpy_api.hpp A us_somo/develop/grpy/grpy_core.hpp A us_somo/develop/grpy/grpy_report.hpp A us_somo/develop/grpy/linalg.hpp A us_somo/develop/grpy/parallel_qt.hpp A us_somo/develop/grpy/parallel_std.hpp A us_somo/develop/grpy/tests/data/1znf.bead_model A us_somo/develop/grpy/tests/data/1znf_golden.txt A us_somo/develop/grpy/tests/data/dumbbell_golden.txt A us_somo/develop/grpy/tests/qt_proof.cpp A us_somo/develop/grpy/tests/run.sh A us_somo/develop/grpy/tests/test_api.cpp A us_somo/develop/grpy/tests/test_assemble.cpp A us_somo/develop/grpy/tests/test_linalg.cpp A us_somo/develop/grpy/tests/test_ooc.cpp A us_somo/develop/grpy/tests/test_threaded.cpp M us_somo/develop/libus_somo.pro Log Message: ----------- somo/grpy: add self-contained in-process GRPY module (Eigen) Adds us_somo/develop/grpy/: a C++ port of GRPY (generalized Rotne-Prager-Yamakawa hydrodynamics) callable in-process, to replace the external Fortran binary run via QProcess with stdout scraping. The module is self-contained and isolated from the SOMO god classes: - header-only, own namespace (grpy), dependency-injected threading (la::Parallel; QtParallel over QThreadPool for SOMO, std::thread for the CLI/tests); - memory-lean tiled Cholesky (solve M X = T instead of forming the full inverse; upper-triangle, factored in place) with single-precision and out-of-core (mmap) options for large systems, and fine-grained in-inversion progress via callback; - grpy::Solver::run(beads, params, progressCb) returns structured scalars AND the full report text, so the on-disk results file is preserved as before. - tests/ validate against the captured GRPY golden (report + scalars) and prove the Qt in-process path; run standalone via tests/run.sh. Wires the qmake build (QT += concurrent, INCLUDEPATH += include grpy, HEADERS). No US_Hydrodyn changes yet; the QProcess call-site swap is a follow-up commit. Refs ehb54/ultrascan-tickets#972 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Commit: d3e4bc90e410221103f9e57d5f1aa0c0f5a0d8d7 https://github.com/ehb54/ultrascan3/commit/d3e4bc90e410221103f9e57d5f1aa0c0f... Author: ehb54 <brookes@uthscsa.edu> Date: 2026-07-18 (Sat, 18 Jul 2026) Changed paths: M us_somo/develop/grpy/grpy_api.hpp A us_somo/develop/grpy/tests/data/dumbbell.grpy M us_somo/develop/grpy/tests/test_api.cpp M us_somo/develop/src/us_hydrodyn_grpy.cpp Log Message: ----------- somo/grpy: run GRPY in-process, drop external binary + Docker path Wire the self-contained grpy module into US_Hydrodyn::calc_grpy_hydro / grpy_process_next: read the .grpy file SOMO already writes via grpy::read_native_file(), run grpy::Solver over a QtParallel backend on SOMO's thread pool, and fill grpy_stdout from Results::report so the existing grpy_finished() parsing and .grpy_res preservation are unchanged. Removes the QProcess launch, the US_Container_Grpy selection, and the Docker/parallel-image plumbing -- in-process is now the sole path. Progress is driven by the Solver callback (processEvents keeps the GUI painting). Adds grpy::read_native_file() + a native-reader test case (dumbbell.grpy) that reproduces the dumbbell golden. Refs ehb54/ultrascan-tickets#972 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Commit: 3647e4575c4e7785585ed676e543c891f0d3c98d https://github.com/ehb54/ultrascan3/commit/3647e4575c4e7785585ed676e543c891f... Author: ehb54 <brookes@uthscsa.edu> Date: 2026-07-18 (Sat, 18 Jul 2026) Changed paths: R us_somo/develop/include/us_container_grpy.h M us_somo/develop/include/us_hydrodyn.h M us_somo/develop/include/us_hydrodyn_misc.h M us_somo/develop/libus_somo.pro R us_somo/develop/src/us_container_grpy.cpp M us_somo/develop/src/us_hydrodyn.cpp M us_somo/develop/src/us_hydrodyn_grpy.cpp M us_somo/develop/src/us_hydrodyn_misc.cpp M us_somo/develop/src/us_hydrodyn_settings.cpp Log Message: ----------- somo/grpy: remove dead external/Docker GRPY plumbing Now that GRPY runs in-process (the external binary + Docker path is gone), delete the symbols that only served the old QProcess path: - US_Container_Grpy class (src/us_container_grpy.cpp, include/us_container_grpy.h) and its libus_somo.pro SOURCES/HEADERS entries; the us_container_grpy member, its includes/init, and grpy_parallel_pulled. - cb_parallel_grpy checkbox + set_parallel_grpy() slot and the misc.parallel_grpy field, incl. its settings save/load/default (old saved values are simply ignored). - the grpy QProcess* member, grpy_prog, and the grpy_readFromStdout/Stderr/started slots. stop_calc()'s process-terminate block is replaced by a note: cancellation now flows through stopFlag, honored by the Solver progress callback and grpy_finished(). Objects rebuild cleanly (us_hydrodyn, us_hydrodyn_grpy, us_hydrodyn_misc, us_hydrodyn_settings). Refs ehb54/ultrascan-tickets#972 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Commit: 7a83b64352e047e10fe6bc46d68e4e6a05bd9aee https://github.com/ehb54/ultrascan3/commit/7a83b64352e047e10fe6bc46d68e4e6a0... Author: ehb54 <brookes@uthscsa.edu> Date: 2026-07-19 (Sun, 19 Jul 2026) Changed paths: M us_somo/develop/src/us_hydrodyn_grpy.cpp Log Message: ----------- somo/grpy: fix in-process finish hand-off + resolve input path Two fixes found while validating the in-process path end-to-end via the batch gui_script on a real bead model: 1. grpy_finished() was never invoked. It was posted with QMetaObject::invokeMethod(..., Qt::QueuedConnection, Q_ARG(QProcess:: ExitStatus, ...)), but QProcess::ExitStatus is not a registered queued-connection metatype, so the invoke silently failed ("Unable to handle unregistered datatype 'QProcess::ExitStatus'"). grpy_running then never cleared and the batch's while(grpy_running) processEvents() loop hung forever. Replaced both call sites with QTimer::singleShot(0, this, lambda) which defers to the event loop (still no deep recursion through the model batch) but calls grpy_finished() directly, needing no metatype registration. 2. read_native_file() read grpy_last_processed (a bare filename) relative to the process CWD. The old QProcess set its working directory to get_somo_dir(); resolve the path against get_somo_dir() to match, and guard a missing file with a clear error instead of a silent empty solve. Validated: gui_script batch run on a 165-bead model produces a .grpy_res whose every physical observable (rotational diffusion, sedimentation, intrinsic viscosities, all 8 relaxation times, diagonal diffusion tensor) matches the Fortran GRPY_osx10.11 binary on the same .grpy input exactly; only the documented eigenvector sign-gauge and ~1e-9 coupling noise differ. Refs ehb54/ultrascan-tickets#972 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Commit: 89f4a3ddd7aa11db5166a243e2d5260aa746ccb4 https://github.com/ehb54/ultrascan3/commit/89f4a3ddd7aa11db5166a243e2d5260aa... Author: ehb54 <brookes@uthscsa.edu> Date: 2026-07-19 (Sun, 19 Jul 2026) Changed paths: M us_somo/develop/src/us_hydrodyn_grpy.cpp Log Message: ----------- somo/grpy: expose single-precision + out-of-core options Wire grpy::Options.single and .ooc_dir at the in-process call site. These matter only for very large bead models (memory-bound; run via batch/cluster), so they are surfaced through the lightweight, no-UI mechanisms rather than new god-class dialog controls: - gui_script `global grpy_single 1` / `global grpy_ooc_dir <dir>` - GRPY_SINGLE / GRPY_OOC_DIR environment-variable fallback (matches the standalone CLI's env knobs) Default off = in-core double precision, byte-identical to prior behavior. A "GRPY options: ..." editor message is emitted when either is active. Validated: gui_script run with `global grpy_single 1` on the 165-bead model completes and reproduces every labeled observable of the double-precision / Fortran-golden result (single matches to 4 sig figs by design). The module's single-precision and out-of-core paths themselves are covered by the grpy unit tests (test_linalg float, test_ooc). Refs ehb54/ultrascan-tickets#972 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Commit: f17026f45075de36cb25ba4601f2fa662255683c https://github.com/ehb54/ultrascan3/commit/f17026f45075de36cb25ba4601f2fa662... Author: ehb54 <brookes@uthscsa.edu> Date: 2026-07-19 (Sun, 19 Jul 2026) Changed paths: M us_somo/develop/include/us_hydrodyn_hydro.h M us_somo/develop/src/us_hydrodyn_grpy.cpp M us_somo/develop/src/us_hydrodyn_hydro.cpp M us_somo/develop/src/us_hydrodyn_settings.cpp Log Message: ----------- somo/grpy: add GRPY precision (double/float) control to Hydro Options Per review, surface the single-precision option as a proper control in the SOMO Hydrodynamic Calculation Options window rather than script/env only. Adds a "GRPY Numerical Precision:" group with "Double (default)" / "Float (for large systems)" radios to US_Hydrodyn_Hydro, right below the existing "Inclusion of Buried Beads (for GRPY)" group and mirroring its idiom. Backed by hydro.grpy_single (default false = double), persisted in the SOMO settings (save/load/default) and reported in display_default_differences() like the sibling GRPY setting. The in-process call site now reads hydro.grpy_single as the primary source; the gui_script `global grpy_single` param and GRPY_SINGLE env var still override for headless/batch automation. Out-of-core stays script/env only (a cluster-scale knob, no GUI control). Validated: full libus_somo rebuild clean (0 errors); default-double gui_script run reproduces the Fortran-golden scalars (no regression). Refs ehb54/ultrascan-tickets#972 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Commit: 2144ac85955171f354fcd3edd5402d67e4e4f4f5 https://github.com/ehb54/ultrascan3/commit/2144ac85955171f354fcd3edd5402d67e... Author: emre brookes <ehb54@users.noreply.github.com> Date: 2026-08-06 (Thu, 06 Aug 2026) Changed paths: A us_somo/develop/grpy/INTEGRATION.md A us_somo/develop/grpy/README.md A us_somo/develop/grpy/grpy_api.hpp A us_somo/develop/grpy/grpy_core.hpp A us_somo/develop/grpy/grpy_report.hpp A us_somo/develop/grpy/linalg.hpp A us_somo/develop/grpy/parallel_qt.hpp A us_somo/develop/grpy/parallel_std.hpp A us_somo/develop/grpy/tests/data/1znf.bead_model A us_somo/develop/grpy/tests/data/1znf_golden.txt A us_somo/develop/grpy/tests/data/dumbbell.grpy A us_somo/develop/grpy/tests/data/dumbbell_golden.txt A us_somo/develop/grpy/tests/qt_proof.cpp A us_somo/develop/grpy/tests/run.sh A us_somo/develop/grpy/tests/test_api.cpp A us_somo/develop/grpy/tests/test_assemble.cpp A us_somo/develop/grpy/tests/test_linalg.cpp A us_somo/develop/grpy/tests/test_ooc.cpp A us_somo/develop/grpy/tests/test_threaded.cpp R us_somo/develop/include/us_container_grpy.h M us_somo/develop/include/us_hydrodyn.h M us_somo/develop/include/us_hydrodyn_hydro.h M us_somo/develop/include/us_hydrodyn_misc.h M us_somo/develop/libus_somo.pro R us_somo/develop/src/us_container_grpy.cpp M us_somo/develop/src/us_hydrodyn.cpp M us_somo/develop/src/us_hydrodyn_grpy.cpp M us_somo/develop/src/us_hydrodyn_hydro.cpp M us_somo/develop/src/us_hydrodyn_misc.cpp M us_somo/develop/src/us_hydrodyn_settings.cpp Log Message: ----------- Merge pull request #487 from ehb54/ehb54-issue-972 somo: in-process GRPY (C++/Eigen), replacing the external binary Compare: https://github.com/ehb54/ultrascan3/compare/bf1c7305fc74...2144ac859551 To unsubscribe from these emails, change your notification settings at https://github.com/ehb54/ultrascan3/settings/notifications