v2.2.1

What's Changed

• feat: add universal one hot feature and sliding window for rop in it… by @Franklalalala in #231
• fix: add checks for basis existence in NNSK methods to prevent key err by @QG-phy in #257
• add manually input overlap matrix feature with example by @Lonya0 in #253
• Hotfix tojson by @QG-phy in #258
• feat: add support for multiple datasets input by @Franklalalala in #252
• add Trinity model and debug the lem by @floatingCatty in #256
• feat: Add support for eigenvalue solvers 'torch' and 'numpy' by @AsymmetryChou in #265
• hot fix: average loss for non-fast validation only by @Franklalalala in #261
• fix small type in doc ana update by @floatingCatty in #266
• feat: Add delete overlap option to pth2json and improve model saving … by @QG-phy in #268
• Fix orthogonalization in eigh by @AsymmetryChou in #270
• Fix bug so2 by @floatingCatty in #267
• Efficient Block-Diagonal Matrix Operations for Wigner D Computation by @Kai-Qi in #250
• make nintp large than 0 by @AsymmetryChou in #277
• feat: Add dptbplas example by @QG-phy in #281
• change to UV for dependency management by @QG-phy in #282
• add jupyter to toml files by @QG-phy in #283
• Fix: fix AtomicData.py mask handling and ensure proper numpy arrays by @AsymmetryChou in #286
• fix: Ensure proper tensor conversion for numpy solver in Eigenvalues … by @AsymmetryChou in #287
• feat: Support SOC for overlap, add --soc option in esk, add uniform_noref method for SOC and support s/p/d type for SOC by @xdzhu in #271
• refactor: Refactor trajectory data loading and improve the docstrings by @kirk0830 in #289
• Fix: Convert PBC array to list format by @kirk0830 in #295
• refactor: reorganize imports and improve dataset builder logic in build.py by @kirk0830 in #288
• build(deps): bump actions/checkout from 4 to 6 by @dependabot[bot] in #284
• build(deps): bump actions/setup-python from 5 to 6 by @dependabot[bot] in #274
• feat: refactor OPENMX orbital mapping constants by @QG-phy in #296
• SOC north by @QG-phy in #297
• Update_example by @QG-phy in #298
• feat(kpoints): add time symmetry reduction with correct weights by @QG-phy in #272
• Newlogo by @QG-phy in #299
• Added model export functionality to convert trained models to Wannier90 and PythTB formats. by @QG-phy in #300
• Fix(postprocess): Fixed overlap data initialization in electronic structure calculations Bug due to merged PR300 by @QG-phy in #303
• remove useless module of negf and v1 code by @QG-phy in #304
• Feat: Add new unifined postprocess class to handle DeePTB model. by @QG-phy in #301
• feat: basis transfer from small basis to larger ones by finetuning on the the eigenvalues. by @floatingCatty in #305
• feat(postprocess): Add functionality to export system data for Pardiso/Julia band structure calculations and include related tests. by @YJQ-7 in #309
• Perf: accelerate feature_to_block with torch_scatter by @AsymmetryChou in #302
• add two center integral of overlap computation by @floatingCatty in #311
• build(deps): bump docker/login-action from 2 to 4 by @dependabot[bot] in #318
• build(deps): bump docker/metadata-action from 4 to 6 by @dependabot[bot] in #317
• build(deps): bump docker/build-push-action from 6 to 7 by @dependabot[bot] in #316
• build(deps): bump docker/setup-buildx-action from 3 to 4 by @dependabot[bot] in #315
• Sparse Hr to HR forming by @floatingCatty in #312
• Make vbcsr an optional dependency by @QG-phy in #328
• fix: only swallow missing vbcsr import in Hr2HR by @QG-phy in #329
• fix the memory issue by @floatingCatty in #322
• feat: Add robust fallback for ill-conditioned overlap matrices in Eigenvalues solver by @Franklalalala in #320
• build(pyproject): adopt PEP 639 license metadata by @QG-phy in #333
• fix(postprocess): normalize unified eigen solver options by @QG-phy in #334
• Feat/modular julia pardiso by @QG-phy in #310
• CI: add maintainer review governance safeguards by @QG-phy in #337
• CI: improve PR review plan comment format by @QG-phy in #338
• CI: keep PR review plan comment checkout in sync by @QG-phy in #339

New Contributors

• @Kai-Qi made their first contribution in #250
• @xdzhu made their first contribution in #271
• @kirk0830 made their first contribution in #289
• @YJQ-7 made their first contribution in #309

Full Changelog: v2.2.0...v2.2.1