ee_adc_merge - #38

* Merging EE-ADC code (#38)

* Running autoflake to clean up flake errors

* Trying agressive autoflake

* Addressing flake errors in uadc_ip_cvs.py

* Fixing comparisons with boolean

* Addressing long line flake errors

* Flake is clean

* Fixing issues in periodic tests

* Renaming tests to skip

* Fixed API for the DF integrals

* Updates master from base fork

* Interface to libmsym (https://github.com/mcodev31/libmsym) for handeling non-abelian symmetry (pyscf#1477)

* ENH: Add interface to libmsym (https://github.com/mcodev31/libmsym)

* ENH: Add option to clean non-symmetric orbital parts in symm.symmetrize_space

* ENH: Add extrasym option to forbid certain orbital rotations in mcscf

* Write irrep ID if irrep name is not available

* ENH: Use axes and origin found by pyscf to generate SALCs with libmsym

* Add doc strings to the libmsym interface and verify it is not used with mol.cart

* Improvements to symm.addons.symmetrize_space

* Create find_symmetric_mo function out of existing code

* Add Lowdin orthogonalization steps into symm.addons.symmetrize_space

* Add function that can symmetrize orbitals with respect to multidimensional irreps

* Get rid of extra dicts in the multidimensional symmetrization function

* Make orbital irrep_id by libmsym interface reductable to the highest Abelian subgroup

* Fix tests, skip symmetrization analysis if there are no orbitals

* fix bad indentations (E111 pycodestyle warning)

* some pep8 fixes as suggested by pycodestyle (pyscf#1834)

* fix some typos (found using codespell) (pyscf#1828)

* some fixes for rst syntax in the doc (pyscf#1829)

* Recognize also numpy integer types in the density functional wrapper

* Implement another variant of the Hückel guess  (pyscf#1784)

* Implement another variant of the Hückel guess based on the updated GWH rule from doi:10.1021/ja00480a005

* Fix bug in symm cas methods (pyscf#1817)

* fix bug in symm cas methods

* Fix flake8 W293

* fixing a few more typos in gto/ and lo/

* Fix the edge case CAS(2,2) for Selected CI (issue pyscf#1808) (pyscf#1845)

* Fix the edge case CAS(2,2) for Selected CI (issue pyscf#1808)

* Fix test

* [Ready for review] Complete implementation of pbc/tdscf (pyscf#1793)

* Bug fix for RSDF due to Mole basis sorting

* Bug fix for pbc df Jbuild for hermi == 0

* improve TDSCF init guess degeneracy handle

* complete pbc tdscf

* b3lyp -> b3lyp5

* adjust values due to v2.3

* add examples; add finalize

* remove non-gamma point exception

* fix flake8 error

* remove numer instable tests (higher roots)

* update tests

* fix wrong num of states

* bug fix in response

* keep minimal changes for pbc _response

* better handle of exxdiv, including vcut

* add NotImplementedError for rsjk + non-zero kshift

* fix shifted kpt for TDRKS/UKS

* fix flake8 error

* revert CasidaTDDFT

* fix flake error

---------

Co-authored-by: hongzhouye <>

* added NVT BOMD and example (pyscf#1818)

* add NVT BOMD and example

* overload functions

* remove default NVT

* remove NVTBerendson

* update init

* update example

* add NVTBerendson unit test

* corrected Temperature()

* corrected temperature

* added temp rmsf check

* added temperature test

* removed pandas dependency

* added temperature test linear mol

* corrected flake8 errors

* Fix lz_adaptation example; needs to run in D2h symmetry to work

* new collection of typos everywhere

* Optimize grad_nuc

* Newton casscf hcc (issue pyscf#701) (pyscf#1869)

* Hessian_CC in newton_casscf (issue pyscf#731)

Hessian-vector product should have zero overlap w/ keyframe CI
vector at all times. Remove factors of 2 which spoiled this.

* update newton_casscf unit tests

---------

Co-authored-by: Matthew R Hermes <[email protected]>

* fix other rst issues

* Support Gaussian charge model for int1e_grids

* GHF with fractional occupancy (pyscf#1826)

* Fix bug in GHF with fractional occupancy

* Remove unused code in X2C

* Fix setup.py

* Implement fcidump.from_mcscf

* add pycodestyle to linter workflow

* trying to fix the flake8 linter

* get rid of TAB characters, and activate check for W191

* fix and activate pycodestyle W293

* Fixes pyscf#1846

* Optimize JK builder for short-range ERIs (pyscf#1882)

* Optimize get_jk for SR integrals; Update fill_ints_sr for cint 6.0

* Add tests

* Fix _vhf.direct_mapdm

* Refactor VHFOpt, to manage the memory of CVHFOpt in Python

* Updates for lint

* Libxc or XCFun should be cited for density functional calculations

* Support using DF-CCSD and frozen for FNO-CCSD (pyscf#1893)

* Bug fix for RSDF due to Mole basis sorting

* Bug fix for pbc df Jbuild for hermi == 0

* support frozen

* fix test name

---------

Co-authored-by: hongzhouye <>

* Assume 46 and 78 core configurations to be f-in-valence

* fix but do not activate W391 (pyscf#1885)

* Improve wheel (pyscf#1900)

* Improve setup.py; fix issue pyscf#973 pyscf#934

* Adjust CI test pipeline and pypi wheel build pipeline

* Adjust CI pipeline for python 3.12

* Fix CI job on MacOS

* Remove pytest-timer from python3.12 CI tests

* Pytest-cov uses a large amount of memory in CI jobs

* Fix ci.yml

* Remove deprecated pkg_resources library (for Python 3.12)

* Fix syntax error

* Disable pyberny version check

* New features and code refactoring (pyscf#1859)

* Add numint2c for pbc.dft; Update pbc-gks tests; Refactor pbc scf and dft methods

* Fix inheritance problems in k-DFT ksymm methods

* Add vv10 for pbc dft

* Fix kgks

* Improve k-dft initialization

* Update pbc-dft nlc tests

* Add undo method for dynamically generated classes

* Update the treatment of _keys

* Refactor smearing

* Refactor mcscf

* new GTO base class and conversion between mole and cell

* Refactor basis parser; merge gto.mole basis parser and pbcgto.cell basis parser

* Fix inconsistency between _bas and _basis (related to issue pyscf#1770)

* update __getattr__

* Remove unnecessary Mixins

* Update class structures of localization modules

* Add the shallow copy method in the base class StreamObject

* Remove copy statements in mcscf modules

* Fix basis parser and other tests

* Update mcscf addons and fci addons

* Update tests

* Make Mole and Cell picklable

* Adjust CI test pipelines and other updates

* Fix code merging conflicts

* Update rsdf _keys

* Restore r_numint tests

* Fix pp_int for Mole object

* Update tests

* Fix TDDFT initial guess (pyscf#1897)

* Remove trunc_workspace

* Fix initial guess for tdscf methods; Update tdscf tests

* Restore tests

* Update tests

* Fix tests, restore tests

* no need to inherit from object

* add pcm solvent models (pyscf#1884)

* added solvent models

* add example for RHF

* cleanup variables

* support casci casscf and ccsd

* uncomment unittests

* change example name

* update reset

* for flake8

* fixed a bug in soscf/newton_ah.py

* updated for recent master changes

* remove whitespace

* remove whitespace

* use yield from (ruff UP028)

* Release 2.4 (pyscf#1904)

* Release 2.4

* Fix macos pypi wheel

* Allow to turn off AO symmetry for PBC KS calculations (pyscf#1905)

* allow turning off AO symmetry for PBC DFT

* add test

* modernize super() (ruff UP008)

* Fix CASCI grad (pyscf#1914)

* casci grad crash for small max_memory. Fix pyscf#1909

* Fix tests

* Fix chunking dimension for pol_embed in  _compute_multipole_potential_integrals

* Lattice-sum cutoff (pyscf#1923)

* basis rcut bug for non-orthogonal cell. Fix pyscf#1911

* Slightly adjust estimate_rcut to improve PBC integral accuracy

* Fix adc tests

* fix get_bands with k-point symmetry

* added selected CI example for molecules

* apply "ruff --fix check --select=C40 pyscf"

* fix cc example 20

* Update Mole.loads method for backward compatibility

* Fix rsdf due to the basis sorting in 923d84c (issue 1942)

* Fix decontract_basis function when basis are missed on certain atoms. (Fix pyscf#1711) (pyscf#1943)

* Fix decontract_basis when basis are missed on certain atoms. (Fix pyscf#1711)

* 4th, and 5th order XC derivatives (pyscf#1930)

* Update conda build

* Arbitrary order to transform_xc; patch xcfun up to 5th order

* Update libxc itrf

* Configure max_deriv_order; arbitrary order for ud2ts and ts2ud

* Add 4th order XC deriv in CI job

* To gpu (pyscf#1901)

* Add to_gpu

* Remove multi-inheritances of pbc mean-field class

* Remove GKS.to_gpu

* Undefined variables

* Update to_gpu

* Add istype function to handle both the mol-SCF and gamma-point-SCF objects

* Update mf type check code in all post HF functions

* Improve the undo_soscf method in which unnecessary df objects are remvoed

* Fix tests

* Update tdscf initialization

* typo

* Missing remove_soscf in khf

* add 'max_cycle' to hessianbase

* Fix PBC DFT Becke grid rounding error (pyscf#1960)

* Explicit file encoding for cuda mixed environment due to issue NVIDIA/cuda-python#29

* Swith off sparsity algorithm for small system (issue pyscf#1956)

* native support dispersion correction (pyscf#1916)

* added solvent models

* add example for RHF

* cleanup variables

* support casci casscf and ccsd

* uncomment unittests

* change example name

* update reset

* for flake8

* fixed a bug in soscf/newton_ah.py

* updated for recent master changes

* remove whitespace

* remove whitespace

* native support dispersion correction

* fixed a bug in pcm

* Update hf.py

* move dispersion to addons

* remove disp in RKS

* call get_dispersion in kernels

* added unit test for d4

* fixed dispersion correction in testing h2o

* updated ci

* skip unittest if dftd3 or dftd4 is missing

* Fci str2addr (pyscf#1971)

* Update cistring module for issue pyscf#1886

* Slightly optimize the fci contraction performance

* fix a bug in KUGWAC that doesn't generate proper GDF when using LDA/GGA functionals

* Update periodic GW init and add example (pyscf#1967)

* Update periodic GW init and add example

* Update 22-k_points_gw.py

* Removing mf.to_rhf() and mf.to_uhf()

The `to_rhf()` seems to raise the same DF error as in the unrestricted case, so I've removed it. But with/without `to_uhf()` I still get the same DF error.

* Fix a bug in ci string function

* fix typo

* Bug fix in interactive solver for non-canonical UMP2 (pyscf#1991)

* Bug fix in interactive solver for non-canonical UMP2

* Fixed the test_ump2.py

---------

Co-authored-by: hung.pham <[email protected]>

* Bug fix for Pipek Mezey + Stability check using Jacobi sweep (pyscf#1990)

* PM loss/grad/hess for any exponent; add test

* add jacobi stability check for PM

* bug fix for boy pbc

* Add heading

---------

Co-authored-by: hongzhouye <>

* skip mol in to_gpu

* Fix basis order issue in molden.load (issue pyscf#1961)

* Update Mole and Cell for the hasattr error in to_gpu and to_cpu methods

* Fix post-hf Gradients with cartesian GTOs (issue pyscf#1985)

* Handle xcfun patch failure in cmake

* Fix the shape bug in make_eris_incore in gccsd.py

As described in issues pyscf#2002

* Enable numpy array of get_k_kpts in df_jk.py

Dear all,

I found that if the user gives a set of mo_occ or mo_coeff (in numpy array form) to the kmf with GDF, the current will raise an error. 
It assume the mo_coeff, mo_occ is list of tuple.

* fix indentation

* fix a typo (missing .ndim)

* Include publication year and DOI in ECP computation references

* Update cmake patch command (issue pyscf#1998)

* Warning messages, and other improvements (pyscf#2009)

* Warning messages for MGGA in Hessian calculations

* Cache dynamically generated class

* Update the HDF5 tempfile handler

* Ensure shape of lattice vectors (issue pyscf#1932)

* feat(gto.basis): improve final, generic error message for basis not found

* add nr_nlc_vxc to NumInt

* Chkfile initial guess problem in ROHF and ROKS (issue pyscf#1986) (pyscf#2010)

* Fix chkfile initial guess problem for ROHF and ROKS (issue pyscf#1986)

* Use UHF natural orbital for RHF chkfile initial guess

* Enable some xc_deriv tests

* Enable xcfun compilation

* Print b3lyp warning when needed (issue pyscf#2029) (pyscf#2040)

* Hide b3lyp warning (issue pyscf#2029)

* Backward compatibility for xcfun

* typo

* Optimize the memory usage of GDF builder (pyscf#2047)

* Control the memory footprint for rsdf_builder and rsjk_builder (issue pyscf#2017)

* Adjust omega value in rsdf and rsjk

* Improve cistring.addr2str for large address (issue pyscf#1886)

* Handle magmom in Mole serialization (pyscf#2049)

* magmom serialization problem (issue pyscf#2026)

* Handle mesh

* Add tests for cell.dumps

* Fix the class of supercell scf object for the k2gamma conversion (pyscf#2030) (pyscf#2041)

* Fix the class of supercell scf object for the k2gamma conversion (pyscf#2030)

* typo

* Update k2gamma tests

* Refactor: error checks in C code and updates for GPU4PySCF library (pyscf#2048)

* Update to c99 standard; update dependency to libcint-6.1.1

* Update SCF functions for corresponding code in GPU4PySCF

* Add checks for malloc

* Update 02-ghf example according to issue pyscf#1970

* Update libcint/qcint version

* Remove the implicit dependency to mpi4py in __init__.py (issue pyscf#1888) (pyscf#2059)

* Remove the implicit dependency to mpi4py in __init__.py (issue pyscf#1888)

* import error

* Import error in momgfccsd tests

* Fix df-grad for UHF when symmetry is enabled (issue pyscf#2054)

* Check cylindrical symmetry of the orbitals for FCI solver (issue pyscf#2022)

* Fix bug for init_guess=atom (issue pyscf#2056)
Fix dft get_veff tests

* Improve magmom initialization (issue pyscf#2055)

* DIIS with damping (pyscf#2053)

* update simple damping and add DIIS damping

* update get_fock in solvent and dynamic level shift

---------

Co-authored-by: hongzhouye <>

* Add level shift to CPHF solver

* fix cp2k basis parse (issue pyscf#2043) (pyscf#2045)

* fix issue pyscf#2043

* add test for parse_cp2k

* allow symbol choosing for GTH PP parser

* SA-CASSCF Nonadiabatic Coupling Vectors (pyscf#2046)

* add nacs stuff but need to rebuild

* fix io warning test

* add sacasscf nac files from mrh

* update and add nac_method

* fix for flake8

* add example file

* update to assert len and uniqueness of states

* udpate

* udpate

* udpate

* fix assert statements and handling of state[0]==state[1]

* Backward compatibility issue in fci.cistring

* Release 2.5

* fix magmom check in mol.copy()

If magmom is an np.ndarray,
`if mol.magmom` will raise error.
Need to explicitly use `is not None`

* Move LebedevGrid C code to Python code; fix issue pyscf#2073 (pyscf#2076)

* Move LebedevGrid C code to Python code; fix issue pyscf#2073

* Restore CxLebedevGrid.c and fix bug

* lint error

* Add an API to dump system info

* multigrid DFT version 2 (pyscf#2078)

* Fix smearing with predefined chemical potential (pyscf#2098)

* added hubbard model with UHF example, fixed chemical potential bug in addons.py

* fix pbc smearing

* add test for smearing with mu0

---------

Co-authored-by: fishjojo <[email protected]>

* Add tests for ndarray_pointer_2d function (pyscf#2101)

* Add tests for ndarray_pointer_2d function

* Fix ndarray_pointer_2d

* Fix transform_ci for more than 64 orbitals (pyscf#2095)

* Fix transform_ci for more than 64 orbitals

* Separate occ_masks into a function

* fix binomial function in fci_string

* Fix xcfun high-order derivatives

* More type checks in fci solver (pyscf#2122)

* float type checks

* Update mcscf docstring (issue pyscf#2119)

* Fix failed tests

* Improve to_gpu (pyscf#2124)

* Update to_gpu

* Improve to_ks/to_hf

* Update to_gpu for df, solvent, qmm, and other methods

* Fix fci initialization

* Improve memory usage for RPA (pyscf#2115)

* improve memory usage for RPA with outcore ao2mo; fix small bug

* raise something when loading libpbc.get_Gv

* fix flake8 issue

* fix flake8 issue

* used DF-RHF as EXX, changed ref values

* unrestricted case

* unrestricted case

* Restore pyscf/pbc/gto/cell.py

* DFT wrapper supporting conventional DFT methods (pyscf#2084)

* added solvent models

* add example for RHF

* cleanup variables

* support casci casscf and ccsd

* uncomment unittests

* change example name

* update reset

* for flake8

* fixed a bug in soscf/newton_ah.py

* updated for recent master changes

* remove whitespace

* remove whitespace

* native support dispersion correction

* fixed a bug in pcm

* Update hf.py

* move dispersion to addons

* remove disp in RKS

* call get_dispersion in kernels

* added unit test for d4

* fixed dispersion correction in testing h2o

* updated ci

* skip unittest if dftd3 or dftd4 is missing

* new high level dft wrapper

* clean up & move dispersion to energy_tot

* remove disp from dft _keys

* fixed import dftd3_xc_map

* fixed typo

* added a wrapper for is_nlc

* replaced more ni.libxc.is_nlc with ni.is_nlc

* added wb97x-d3 as a special case

* added disp_3body

* removesuffix -> replace

* address qiming's comments

* add dft_parser && change back to libxc.is_nlc

* fixed unit test

* disable wb97x-d3, wb97x-d

* skip dftd3 for py3.12 & skip wb97x-d3

* Update dft_parser; Add tests for dft_parser

* Add warning messages

* Update test_h2o.py

---------

Co-authored-by: Qiming Sun <[email protected]>

* fix dipole moment in sfx2c1e

* debug and simplify to_gpu (pyscf#2149)

* support and simplify to_gpu

* remove comments

* flake8

* bug fix for RCCSD(T) with complex orbitals (pyscf#2141)

* bug fix for vooo order

* fix flake8

---------

Co-authored-by: hongzhouye <>

* Fix if mf.istype('UHF') for to_uhf() conversion

* Solve conflicts between @Property and __getattr__

* move static configurations from setup.py to pyproject.toml (pyscf#2144)

* move testing options to pytest.ini

* move static configurations from setup.py to pyproject.toml

* Update CI release jobs

---------

Co-authored-by: Qiming Sun <[email protected]>

* fix pbc df with KPoints input

* Fix pip builder

* reset enuc in _build_supcell_ (pyscf#2164)

* reset enuc in _build_supcell_

* add test for enuc reset

---------

Co-authored-by: hongzhouye <>

* Bug fix for single k-point JK-build in PBC DF (pyscf#2165)

* bug fix for j-build in get_jk

* bug fix for k-build in get_jk

---------

Co-authored-by: hongzhouye <>

* add missing init file

* DFTD3 & DFTD4 builder (pyscf#2161)

* local build dftd3/dftd4

* flake8

* fixed issues apis

* add compilation tool for mac

* fixed bugs in unittest

* fixed an issue in unit test

* add fortran compiler

* fortran compiler

* Move dftd3 and dftd4 builder to a separated repo (pyscf-dispersion)

* enable dftd3 & dftd4 for python3.12

* Adjust tests

---------

Co-authored-by: Qiming Sun <[email protected]>

* Kpoint dft to_hf error (fix issue pyscf#2157)

* ancient GCC compatibility

* fix

* fix: infinite recursion atom_hf call when ECP with super-heavy atoms (>Cm) (pyscf#2183)

* bugfix: resolve recursive initial guess when atomic charge > 96 (> Cm) and ecp activated

* Adjust AGF2 tests

---------

Co-authored-by: Qiming Sun <[email protected]>

* fix x_id in parse_token

* fixing unit test for dftd3 and dftd4

* Run CASCI and CASSCF without symmetry if symmetry is C1 (pyscf#2195)

* Run CASCI and CASSCF without symmetry if symmetry is C1

* groupname, not symmetry!

* move auxbasis_response out of __init__ (pyscf#2192)

* move auxbasis_response out of __init__

* relax unit test in test_c_agf2

* import hessian

* Update __init__.py

* improve the consistency in DF modules (pyscf#2204)

* improve the consistency

* added comment to _pinv

* tight conv_tol_cpscf by default

* set linear dependence threshold to 1e-9

* remove comments

* update unit tests

* SMD model and PCM Hessian (pyscf#2173)

* local build dftd3/dftd4

* flake8

* fixed issues apis

* add compilation tool for mac

* fixed bugs in unittest

* fixed an issue in unit test

* add fortran compiler

* fortran compiler

* Move dftd3 and dftd4 builder to a separated repo (pyscf-dispersion)

* enable dftd3 & dftd4 for python3.12

* Adjust tests

* add PCM and SMD, gradient and Hessian

* added __init__ files

* debug for ci

* flake8

* add more unit tests && refactor code

* Fix if mf.istype('UHF') for to_uhf() conversion

* Solve conflicts between @Property and __getattr__

* move static configurations from setup.py to pyproject.toml (pyscf#2144)

* move testing options to pytest.ini

* move static configurations from setup.py to pyproject.toml

* Update CI release jobs

---------

Co-authored-by: Qiming Sun <[email protected]>

* fix pbc df with KPoints input

* Fix pip builder

* reset enuc in _build_supcell_ (pyscf#2164)

* reset enuc in _build_supcell_

* add test for enuc reset

---------

Co-authored-by: hongzhouye <>

* Bug fix for single k-point JK-build in PBC DF (pyscf#2165)

* bug fix for j-build in get_jk

* bug fix for k-build in get_jk

---------

Co-authored-by: hongzhouye <>

* add missing init file

* Kpoint dft to_hf error (fix issue pyscf#2157)

* ancient GCC compatibility

* fix

* local build dftd3/dftd4

* rebase

* fixed bugs in unittest

* add fortran compiler

* rebase

* delete dftd3 and dftd4 files

* legacy SMD

* fortran compiler for mac

* gcc

* which gcc

* found fortran compiler

* fortran compiler

* reinstall gcc

* update unit test for pcm

* relax unit test in agf2

* resolved some comments

* add statement for smd source code

* refactor smd

* fixed bugs

* bugfix

* Skip compiling SMD by default

* Add fortran compiler in Macos build;
Fix conda build.

* empty line at EOF

* Fix SMD configuration in linux wheel

* add an example for SMD

* SkipTest -> skipTest

* excited states for PCM

* revert the changes in dispersion

* EOF

* added warnings in solvent modules when finite difference is used

---------

Co-authored-by: Qiming Sun <[email protected]>
Co-authored-by: Michal Krompiec <[email protected]>
Co-authored-by: chillenb <[email protected]>
Co-authored-by: fishjojo <[email protected]>
Co-authored-by: Hong-Zhou Ye <[email protected]>
Co-authored-by: Maximilian Scheurer <[email protected]>
Co-authored-by: Matthew Hennefarth <[email protected]>

* Density fitting performance optimization (re issue pyscf#1915) (pyscf#2205)

* use trsm in df.incore

* use np.matmul in df/df_jk.py

* faster df.outcore with trsm

* avoid numpy mem copy with h5py read_direct

* clarify chained matmuls

* use lib.dot to avoid syrk, df/outcore.py swap buffers

* Restore functionality for CDFT example (pyscf#2210)

* Restore functionality for CDFT example

I'm sure this example worked when it was authored, but that is no longer the case. This patch restores functionality.

* replace deprecated xrange with range

* add V step size to verbose output

It is one of the convergence criteria and should be shown

* add commentable option to verify consistency of example

* make post and pre diis implementations of the inner loop consistent

* add arg that pyscf expects

On my machine this wasn't needed but on others it was

* Move comments to docstring

* factor out cdft inner loop

* remove unused things

* correct errors in the documentation

* improve DFT APIs (pyscf#2187)

* support nlc = False && enhance dispersion correction

* Add DFT d3 examples

* reorg xc, nlc, and disp

* flake8

* Update dispersion error messages and examples

* updated example for dftd3

* Fix dispersion funciton names

* Fix dft sanity check

* Fix super() in KohnSham.check_sanity()

* Fix dftd3 test

* Conservative import

* print info in dispersion corrections

* change apis

* bugfix

* remove print

* bugfix

* added do_nlc check

* clarify do_nlc

* raise conflict between mf.xc and mf.nlc

* Cleanup disp logics and update disp examples

* Update dft_parser module

* Fix undefined-variable errors

* Fix dispersion parser

* Update example

---------

Co-authored-by: Qiming Sun <[email protected]>

* reomve some traces of unicode (py2) and some unused imports

* Update install dependencies

* Fix an import issue for X2C_KS (issue pyscf#2223)

* The spin_samples attribute for DKS

* Fix frozen for FNO (pyscf#2217)

* fix frozen for FNO

* fix FNO-UCC

* fix flake8

---------

Co-authored-by: hongzhouye <>

* DF dataset naming issue for rsh xc (fix pyscf#2236) (pyscf#2240)

* df dataset naming issue for rsh xc (fix pyscf#2236)

* fix tests

* Release 2.6

* bugfix in dispersion hessian (pyscf#2242)

* bugfix in dispersion hessian

* fixed bug in dispersion hessian

* added more unit test for dispersion

* unit test

* allow passing in options when creating HDF5 files (pyscf#2218)

* allow passing in options when creating HDF5 files

* Undo typo in ao2mo/outcore

* fix typo in h5filewrite_options example

* change default basis linear dependence threshold (pyscf#2228)

* change default basis linear dependence threshold

* sync setting in outcore.py

* unit test & linear dependence threshold

* linter

* 1e-12 -> 1e-14 for linear dependence

* lindep = 1e-13

* relax unit test

* Update test_df.py

* Update incore.py

* Update test_linalg_helper.py

* Update linalg_helper.py

---------

Co-authored-by: Qiming Sun <[email protected]>

* Abandon ROHF gradients and ROHF hessian (pyscf#2259)

* Add ruff check (pyscf#2262)

* Updates for numpy 2.0 release; Add ruff check; Cleanup tests

* Degeneracy issue in KGKS tests

* Adjust tests

* Fix a knumint bug when handling mcol and ncol

* Update ruff config

* Update doc strings (pyscf#2264)

* Update doc of eval_gto function

* Update geomopt examples (fix pyscf#2233) and atom_coords examples

* Improve pypi build pipeline (pyscf#2265)

* Adjust build order in release pipeline

* Update aarch64 build

* Handle dynamically registered post-HF attributes

* Patch release 2.6.1

* use all conda-forge compilers

* Updates due to NumPy 2.0 API changes (pyscf#2274)

* Updates due to NumPy 2.0 API changes

* Fix __array_wrap__

* Fix a bug caused by an inplace operation in linalg_helper._qr

* Fix inplace operation bug in pbc.cc.kintermediates_uhf

* Retore a few np.array(copy=True) statement

* Patch release 2.6.2

* Fix wheel build for setuptools 70 (pyscf#2278)

* Fix platform issue in bdist_wheel

* In case fail to import bdist_wheel

* Remove the ase module bulk from pyscf_ase (fix pyscf#2271) (pyscf#2279)

* Remove the ase module bulk from pyscf_ase (fix pyscf#2271)

* Fix get_ab of tddft for RSH functional (Fix pyscf#2261)

* Missing symbols

* Add tests

* typo

* Fix integer overflow for large ano basis (pyscf#2282)

* Fix integer overflow for large ano basis (fix pyscf#2201)

* More int type issues

* Fix Hessian atmlst and example (pyscf#2293)

* add atmlst to examples; fix examples/hessian/01-scf_hessian.py

* fix atmlst for hessian

* fix lint issue

* fix scipy.sparse.linalg.cg issue

* fix scipy.sparse.linalg.cg issue

* fix scipy.sparse.linalg.cg issue

* fix scipy.sparse.linalg.cg issue

* allow modifying tol and nroots in stability (pyscf#2299)

* allow modifying tol and nroots

* add example for tight tol

* pass kwargs to mf.stability

* fix typo

* Fix bugs in complex-valued contractions for direct_nosym (fix pyscf#2302)

* Handle api changes in scipy-1.14

* typo

* Fix initial guess dimension problems when solving multiple roots (fix issue pyscf#2292)

* Fix import dependency issue in SCF.do_disp()

* Fix a dimension bug in direct_spin1_cyl_sym. Update example for issue pyscf#2291

* compare version as int, not str (pyscf#2309)

* fixing a few typos

* another little wave of typos

* Fix addons.smearing RHF electron number (pyscf#2290)

* typos and typos

* a few more typos fixed

* Fix to kscf for dft (pyscf#2336)

* fix bug in convert_to_kscf

* fix bug in convert_to_kscf

* fix kuhf case in convert_to_kscf

* Fix MP2 FNO bug when pct_occ=1 (pyscf#2337)

* Update mp2.py

* Update ump2.py

* Superposition of Atomic Potentials (SAP) initial guess implementation for HF-SCF calculations (pyscf#2301)

* Added VSAP initial guess implementation for Hartree-Fock SCF
calculations (described by S. Lehtola here:
https://doi.org/10.1021/acs.jctc.8b01089)
Modified fakemol_for_charges to be usable for SAP 3C2E integrals.
Supplied sap_grasp coefficients computed by S. Lehtola.

* Added 'large' and 'small' sap coefficient functionality

* Fixed some bugs in recognition of DFT and HF objects and their SAP
guesses. Made other cleanups and fixes proposed by S. Lehtola.

* Changed logger.error to logger.warn

* Renamed HF SAP guess functions from 'vsap' to 'sap' to avoid confusion
with DFT counterparts using a different method with same name.

* Fixed a bug I had created by accident.
Created new fakemol_for_cgtf_charge function to be used in forming SAP guess
and to avoid breaking other functionality which use the
fakemol_for_charges function.

* Removed redundant attribute check

* vsap -> sap

* charge sumcheck tolerance set to 1e-6

* Apply suggestions from code review

Co-authored-by: Susi Lehtola <[email protected]>

* Changed sap initial guess to accept basis set in internal format
(python dict) or as a string (will look for basis file locally and
from basis set exhcange if bse api is installed)

* Added SAP guess to restricted open-shell Hartree-Fock (ROHF)

* Added SAP guess to non-relativistic generalized Hartree-Fock (GHF)

* Tidy up SAP function in GHF, make sap_grasp_large to default SAP basis

* Add example for SAP inital guess.

* Added SAP guess to Dirac-Hartree-Fock (DHF)

* small cleanups

* SAP now accepts internally formatted python dictionary as basis. SCF object has new attribute sap_basis

* Apply suggestions from code review

Co-authored-by: Susi Lehtola <[email protected]>

* Clarified example using explicit dictionary format basis for SAP

* Apply suggestions from code review

Co-authored-by: Susi Lehtola <[email protected]>

* Comment for vsap added back

* Fixed  issue

* Fixed style issues

* Fix style issue with extraneous 'f' prefix

* fix ROHF SAP guess call tree, add SAP unittest for rhf and rohf

* Fixed accidental incorrect test

* Add SAP guess to UHF

* Add SAP unittest for uhf

* Add SAP unittest for GHF

* Add SAP unittest for DHF

* flake8 warning

---------

Co-authored-by: Susi Lehtola <[email protected]>
Co-authored-by: Qiming Sun <[email protected]>

* Use weakref for automatic deletion of tmpfiles (to fix pyscf#2318) (pyscf#2320)

* Use weakref for automatic deletion of tmpfiles

* github actions tests check number of leftover temporary files

* temporary file check pytest exit status

* Avoid h5f.flush() on closed file

---------

Co-authored-by: Qiming Sun <[email protected]>

* Updates to the EA/IP-ADC code

* Inteface changes

* Changes interface to the 1-RDM functions in IP/EA-RADC

* Fixed accidental copy

* Fixed accidental copy

* Fixed Contractions radc_ip.py

* Moves make_ref_rdm1 to the appropriate location

* IP tests 1RDM Ref and Exc. Up to ADC(3)

* Fixed bugs that were breaking EA tests

* Clean up

* Changes to the UADC class

* EA Tests Completed from SO code. Slight cleanup for IP tests

* Reference 1RDM tests N2 EA + IP up to ADC(3)

* N2 1RDM EA up to ADC(3) tests

* Completed RADC(3)  1RDM N2 IP tests

* Example file added for 1RDM dipole moment examples with various methods

* Small changes to comments in 1RDM test file

* Final Comment Change 1RDM examples

* Formatting changes and fixing approx_trans_moments option

* ROHF code cleanup

* Fixes a test for IP-ADC

* UADC class clean-up

* IP/EA classes clean-up

* Add singles correlation energy

* More class clean-up, added dipole moment integrals

* Author list update

* Addresses lint errors

* Addressing flake issues

* Few more flake issues addressed

* Adding the EE-ADC code from the old branch

* Fixing osc strengths calculation

* Testing oscillator strengths

* Oscillator strengths do not match the old implementation. Investigating...

* fixed osc problem, TY_xx.append(TY_x) not ordering things right for oscillators but this was not a problem for analyzing spec factors

* Changes in the UADC spin contamination interface

* fixed rohf bugg added ooov and vvvo density matrices and variants

* working on density efficient matrices

* added efficent density matrices

* deleted some lines

* added the radc_ee file that is most up to data

* DF low mem bug fixed

* made some updates, need to work on flake

* radc_ee.py passes flake8, working on uadc_ee.py file

* uadc_ee.py file passed flake

* Renames spin contamination functions

* files pass flake test

* Proper handling of compute_properties attribute in RADC

* Cleaning up imports

* Makes sure that CVS-UADC tests pass

* Updates to RADC tests and UADC default convergence thresholds

* updated convg thresholds

* Example updates and some other interface changes

* One more interface change for the RADC code

---------

Co-authored-by: Alexander Sokolov <[email protected]>
Co-authored-by: Alexander Sokolov <[email protected]>
Co-authored-by: sokolov-group <[email protected]>
Co-authored-by: Nikolay A. Bogdanov <[email protected]>
Co-authored-by: Frédéric Chapoton <[email protected]>
Co-authored-by: Susi Lehtola <[email protected]>
Co-authored-by: Junjie, Yang <[email protected]>
Co-authored-by: Qiming Sun <[email protected]>
Co-authored-by: Hongzhou Ye <[email protected]>
Co-authored-by: Aniruddha Seal <[email protected]>
Co-authored-by: Matthew R Hermes <[email protected]>
Co-authored-by: Matthew R Hermes <[email protected]>
Co-authored-by: Nikolay Bogdanov <[email protected]>
Co-authored-by: Max Nusspickel <[email protected]>
Co-authored-by: Xiaojie Wu <[email protected]>
Co-authored-by: Xing Zhang <[email protected]>
Co-authored-by: peter-reinholdt <[email protected]>
Co-authored-by: fishjojo <[email protected]>
Co-authored-by: Chong Sun <[email protected]>
Co-authored-by: Verena Neufeld <[email protected]>
Co-authored-by: Jiachen Li <[email protected]>
Co-authored-by: Timothy Berkelbach <[email protected]>
Co-authored-by: liwt31 <[email protected]>
Co-authored-by: Hung Pham <[email protected]>
Co-authored-by: hung.pham <[email protected]>
Co-authored-by: Zhihao Cui <[email protected]>
Co-authored-by: Felipe S. S. Schneider <[email protected]>
Co-authored-by: hongzhouye <unknown>
Co-authored-by: matthew-hennefarth <[email protected]>
Co-authored-by: Victor Yu <[email protected]>
Co-authored-by: Xubo Wang <[email protected]>
Co-authored-by: xubwa <[email protected]>
Co-authored-by: Michal Krompiec <[email protected]>
Co-authored-by: chillenb <[email protected]>
Co-authored-by: Maximilian Scheurer <[email protected]>
Co-authored-by: Zhenyu Zhu ajz34 <[email protected]>
Co-authored-by: jeanwsr <[email protected]>
Co-authored-by: scohenjanes5 <[email protected]>
Co-authored-by: Joonatan Huhtasalo <[email protected]>
Co-authored-by: Susi Lehtola <[email protected]>
Co-authored-by: serna.40 <[email protected]>
Co-authored-by: Terrence Stahl <[email protected]>

* Adding full out-of-core integral support for EE-RADC. Fixed a bug in EE-UADC out-of-core. Better handling of ladder contractions for the in-core code, some code cleanup. (#40)

* Code clean up, bug fixes, improved output formatting, new tests and examples (#41)

* Adding full out-of-core integral support for EE-RADC. Fixed a bug in EE-UADC out-of-core. Better handling of ladder contractions for the in-core code, some code cleanup.

* added uadc_ee open shell test, need to include spin square in test

* added radc_ee test, need to think about adding a spin square get flag

* all test file pass flake8

* Code cleanup, renaming tests to have consistent file names

* More test file renaming, updates doc strings

* Code clean up in EE-UADC

* added spin square values for open shell systems ee calculations

* A bit more cleaning in EE-UADC

* Looser test thesholds for spin square

* Remove redundant 1-RDM code from EE-ADC S^2 expectation value

* Making sure that the code works for approx_trans_moments=True. Improved output formatting

* New example for EE-ADC S^2

* Linting...

* Fixed subtle bug in approx_trans_moments for EE-UADC. Added tests to check for that.

* Small changes in tests

* Extra line removed

* Merge problem fix

---------

Co-authored-by: Alexander Sokolov <[email protected]>
Co-authored-by: Terrence Stahl <[email protected]>

---------

Co-authored-by: stahl225 <[email protected]>
Co-authored-by: Alexander Sokolov <[email protected]>
Co-authored-by: Nikolay A. Bogdanov <[email protected]>
Co-authored-by: Frédéric Chapoton <[email protected]>
Co-authored-by: Susi Lehtola <[email protected]>
Co-authored-by: Junjie, Yang <[email protected]>
Co-authored-by: Qiming Sun <[email protected]>
Co-authored-by: Hongzhou Ye <[email protected]>
Co-authored-by: Aniruddha Seal <[email protected]>
Co-authored-by: Matthew R Hermes <[email protected]>
Co-authored-by: Matthew R Hermes <[email protected]>
Co-authored-by: Nikolay Bogdanov <[email protected]>
Co-authored-by: Max Nusspickel <[email protected]>
Co-authored-by: Xiaojie Wu <[email protected]>
Co-authored-by: Xing Zhang <[email protected]>
Co-authored-by: peter-reinholdt <[email protected]>
Co-authored-by: fishjojo <[email protected]>
Co-authored-by: Chong Sun <[email protected]>
Co-authored-by: Verena Neufeld <[email protected]>
Co-authored-by: Jiachen Li <[email protected]>
Co-authored-by: Timothy Berkelbach <[email protected]>
Co-authored-by: liwt31 <[email protected]>
Co-authored-by: Hung Pham <[email protected]>
Co-authored-by: hung.pham <[email protected]>
Co-authored-by: Zhihao Cui <[email protected]>
Co-authored-by: Felipe S. S. Schneider <[email protected]>
Co-authored-by: matthew-hennefarth <[email protected]>
Co-authored-by: Victor Yu <[email protected]>
Co-authored-by: Xubo Wang <[email protected]>
Co-authored-by: xubwa <[email protected]>
Co-authored-by: Michal Krompiec <[email protected]>
Co-authored-by: chillenb <[email protected]>
Co-authored-by: Maximilian Scheurer <[email protected]>
Co-authored-by: Zhenyu Zhu ajz34 <[email protected]>
Co-authored-by: jeanwsr <[email protected]>
Co-authored-by: scohenjanes5 <[email protected]>
Co-authored-by: Joonatan Huhtasalo <[email protected]>
Co-authored-by: Susi Lehtola <[email protected]>
Co-authored-by: serna.40 <[email protected]>
Co-authored-by: Terrence Stahl <[email protected]>

* Running autoflake to clean up flake errors

* Trying agressive autoflake

* Addressing flake errors in uadc_ip_cvs.py

* Fixing comparisons with boolean

* Addressing long line flake errors

* Flake is clean

* Fixing issues in periodic tests

* Renaming tests to skip

* Fixed API for the DF integrals

* Updates master from base fork

* Interface to libmsym (https://github.com/mcodev31/libmsym) for handeling non-abelian symmetry (pyscf#1477)

* ENH: Add interface to libmsym (https://github.com/mcodev31/libmsym)

* ENH: Add option to clean non-symmetric orbital parts in symm.symmetrize_space

* ENH: Add extrasym option to forbid certain orbital rotations in mcscf

* Write irrep ID if irrep name is not available

* ENH: Use axes and origin found by pyscf to generate SALCs with libmsym

* Add doc strings to the libmsym interface and verify it is not used with mol.cart

* Improvements to symm.addons.symmetrize_space

* Create find_symmetric_mo function out of existing code

* Add Lowdin orthogonalization steps into symm.addons.symmetrize_space

* Add function that can symmetrize orbitals with respect to multidimensional irreps

* Get rid of extra dicts in the multidimensional symmetrization function

* Make orbital irrep_id by libmsym interface reductable to the highest Abelian subgroup

* Fix tests, skip symmetrization analysis if there are no orbitals

* fix bad indentations (E111 pycodestyle warning)

* some pep8 fixes as suggested by pycodestyle (pyscf#1834)

* fix some typos (found using codespell) (pyscf#1828)

* some fixes for rst syntax in the doc (pyscf#1829)

* Recognize also numpy integer types in the density functional wrapper

* Implement another variant of the Hückel guess  (pyscf#1784)

* Implement another variant of the Hückel guess based on the updated GWH rule from doi:10.1021/ja00480a005

* Fix bug in symm cas methods (pyscf#1817)

* fix bug in symm cas methods

* Fix flake8 W293

* fixing a few more typos in gto/ and lo/

* Fix the edge case CAS(2,2) for Selected CI (issue pyscf#1808) (pyscf#1845)

* Fix the edge case CAS(2,2) for Selected CI (issue pyscf#1808)

* Fix test

* [Ready for review] Complete implementation of pbc/tdscf (pyscf#1793)

* Bug fix for RSDF due to Mole basis sorting

* Bug fix for pbc df Jbuild for hermi == 0

* improve TDSCF init guess degeneracy handle

* complete pbc tdscf

* b3lyp -> b3lyp5

* adjust values due to v2.3

* add examples; add finalize

* remove non-gamma point exception

* fix flake8 error

* remove numer instable tests (higher roots)

* update tests

* fix wrong num of states

* bug fix in response

* keep minimal changes for pbc _response

* better handle of exxdiv, including vcut

* add NotImplementedError for rsjk + non-zero kshift

* fix shifted kpt for TDRKS/UKS

* fix flake8 error

* revert CasidaTDDFT

* fix flake error

---------

Co-authored-by: hongzhouye <>

* added NVT BOMD and example (pyscf#1818)

* add NVT BOMD and example

* overload functions

* remove default NVT

* remove NVTBerendson

* update init

* update example

* add NVTBerendson unit test

* corrected Temperature()

* corrected temperature

* added temp rmsf check

* added temperature test

* removed pandas dependency

* added temperature test linear mol

* corrected flake8 errors

* Fix lz_adaptation example; needs to run in D2h symmetry to work

* new collection of typos everywhere

* Optimize grad_nuc

* Newton casscf hcc (issue pyscf#701) (pyscf#1869)

* Hessian_CC in newton_casscf (issue pyscf#731)

Hessian-vector product should have zero overlap w/ keyframe CI
vector at all times. Remove factors of 2 which spoiled this.

* update newton_casscf unit tests

---------

Co-authored-by: Matthew R Hermes <[email protected]>

* fix other rst issues

* Support Gaussian charge model for int1e_grids

* GHF with fractional occupancy (pyscf#1826)

* Fix bug in GHF with fractional occupancy

* Remove unused code in X2C

* Fix setup.py

* Implement fcidump.from_mcscf

* add pycodestyle to linter workflow

* trying to fix the flake8 linter

* get rid of TAB characters, and activate check for W191

* fix and activate pycodestyle W293

* Fixes pyscf#1846

* Optimize JK builder for short-range ERIs (pyscf#1882)

* Optimize get_jk for SR integrals; Update fill_ints_sr for cint 6.0

* Add tests

* Fix _vhf.direct_mapdm

* Refactor VHFOpt, to manage the memory of CVHFOpt in Python

* Updates for lint

* Libxc or XCFun should be cited for density functional calculations

* Support using DF-CCSD and frozen for FNO-CCSD (pyscf#1893)

* Bug fix for RSDF due to Mole basis sorting

* Bug fix for pbc df Jbuild for hermi == 0

* support frozen

* fix test name

---------

Co-authored-by: hongzhouye <>

* Assume 46 and 78 core configurations to be f-in-valence

* fix but do not activate W391 (pyscf#1885)

* Improve wheel (pyscf#1900)

* Improve setup.py; fix issue pyscf#973 pyscf#934

* Adjust CI test pipeline and pypi wheel build pipeline

* Adjust CI pipeline for python 3.12

* Fix CI job on MacOS

* Remove pytest-timer from python3.12 CI tests

* Pytest-cov uses a large amount of memory in CI jobs

* Fix ci.yml

* Remove deprecated pkg_resources library (for Python 3.12)

* Fix syntax error

* Disable pyberny version check

* New features and code refactoring (pyscf#1859)

* Add numint2c for pbc.dft; Update pbc-gks tests; Refactor pbc scf and dft methods

* Fix inheritance problems in k-DFT ksymm methods

* Add vv10 for pbc dft

* Fix kgks

* Improve k-dft initialization

* Update pbc-dft nlc tests

* Add undo method for dynamically generated classes

* Update the treatment of _keys

* Refactor smearing

* Refactor mcscf

* new GTO base class and conversion between mole and cell

* Refactor basis parser; merge gto.mole basis parser and pbcgto.cell basis parser

* Fix inconsistency between _bas and _basis (related to issue pyscf#1770)

* update __getattr__

* Remove unnecessary Mixins

* Update class structures of localization modules

* Add the shallow copy method in the base class StreamObject

* Remove copy statements in mcscf modules

* Fix basis parser and other tests

* Update mcscf addons and fci addons

* Update tests

* Make Mole and Cell picklable

* Adjust CI test pipelines and other updates

* Fix code merging conflicts

* Update rsdf _keys

* Restore r_numint tests

* Fix pp_int for Mole object

* Update tests

* Fix TDDFT initial guess (pyscf#1897)

* Remove trunc_workspace

* Fix initial guess for tdscf methods; Update tdscf tests

* Restore tests

* Update tests

* Fix tests, restore tests

* no need to inherit from object

* add pcm solvent models (pyscf#1884)

* added solvent models

* add example for RHF

* cleanup variables

* support casci casscf and ccsd

* uncomment unittests

* change example name

* update reset

* for flake8

* fixed a bug in soscf/newton_ah.py

* updated for recent master changes

* remove whitespace

* remove whitespace

* use yield from (ruff UP028)

* Release 2.4 (pyscf#1904)

* Release 2.4

* Fix macos pypi wheel

* Allow to turn off AO symmetry for PBC KS calculations (pyscf#1905)

* allow turning off AO symmetry for PBC DFT

* add test

* modernize super() (ruff UP008)

* Fix CASCI grad (pyscf#1914)

* casci grad crash for small max_memory. Fix pyscf#1909

* Fix tests

* Fix chunking dimension for pol_embed in  _compute_multipole_potential_integrals

* Lattice-sum cutoff (pyscf#1923)

* basis rcut bug for non-orthogonal cell. Fix pyscf#1911

* Slightly adjust estimate_rcut to improve PBC integral accuracy

* Fix adc tests

* fix get_bands with k-point symmetry

* added selected CI example for molecules

* apply "ruff --fix check --select=C40 pyscf"

* fix cc example 20

* Update Mole.loads method for backward compatibility

* Fix rsdf due to the basis sorting in 923d84c (issue 1942)

* Fix decontract_basis function when basis are missed on certain atoms. (Fix pyscf#1711) (pyscf#1943)

* Fix decontract_basis when basis are missed on certain atoms. (Fix pyscf#1711)

* 4th, and 5th order XC derivatives (pyscf#1930)

* Update conda build

* Arbitrary order to transform_xc; patch xcfun up to 5th order

* Update libxc itrf

* Configure max_deriv_order; arbitrary order for ud2ts and ts2ud

* Add 4th order XC deriv in CI job

* To gpu (pyscf#1901)

* Add to_gpu

* Remove multi-inheritances of pbc mean-field class

* Remove GKS.to_gpu

* Undefined variables

* Update to_gpu

* Add istype function to handle both the mol-SCF and gamma-point-SCF objects

* Update mf type check code in all post HF functions

* Improve the undo_soscf method in which unnecessary df objects are remvoed

* Fix tests

* Update tdscf initialization

* typo

* Missing remove_soscf in khf

* add 'max_cycle' to hessianbase

* Fix PBC DFT Becke grid rounding error (pyscf#1960)

* Explicit file encoding for cuda mixed environment due to issue NVIDIA/cuda-python#29

* Swith off sparsity algorithm for small system (issue pyscf#1956)

* native support dispersion correction (pyscf#1916)

* added solvent models

* add example for RHF

* cleanup variables

* support casci casscf and ccsd

* uncomment unittests

* change example name

* update reset

* for flake8

* fixed a bug in soscf/newton_ah.py

* updated for recent master changes

* remove whitespace

* remove whitespace

* native support dispersion correction

* fixed a bug in pcm

* Update hf.py

* move dispersion to addons

* remove disp in RKS

* call get_dispersion in kernels

* added unit test for d4

* fixed dispersion correction in testing h2o

* updated ci

* skip unittest if dftd3 or dftd4 is missing

* Fci str2addr (pyscf#1971)

* Update cistring module for issue pyscf#1886

* Slightly optimize the fci contraction performance

* fix a bug in KUGWAC that doesn't generate proper GDF when using LDA/GGA functionals

* Update periodic GW init and add example (pyscf#1967)

* Update periodic GW init and add example

* Update 22-k_points_gw.py

* Removing mf.to_rhf() and mf.to_uhf()

The `to_rhf()` seems to raise the same DF error as in the unrestricted case, so I've removed it. But with/without `to_uhf()` I still get the same DF error.

* Fix a bug in ci string function

* fix typo

* Bug fix in interactive solver for non-canonical UMP2 (pyscf#1991)

* Bug fix in interactive solver for non-canonical UMP2

* Fixed the test_ump2.py

---------

Co-authored-by: hung.pham <[email protected]>

* Bug fix for Pipek Mezey + Stability check using Jacobi sweep (pyscf#1990)

* PM loss/grad/hess for any exponent; add test

* add jacobi stability check for PM

* bug fix for boy pbc

* Add heading

---------

Co-authored-by: hongzhouye <>

* skip mol in to_gpu

* Fix basis order issue in molden.load (issue pyscf#1961)

* Update Mole and Cell for the hasattr error in to_gpu and to_cpu methods

* Fix post-hf Gradients with cartesian GTOs (issue pyscf#1985)

* Handle xcfun patch failure in cmake

* Fix the shape bug in make_eris_incore in gccsd.py

As described in issues pyscf#2002

* Enable numpy array of get_k_kpts in df_jk.py

Dear all,

I found that if the user gives a set of mo_occ or mo_coeff (in numpy array form) to the kmf with GDF, the current will raise an error.
It assume the mo_coeff, mo_occ is list of tuple.

* fix indentation

* fix a typo (missing .ndim)

* Include publication year and DOI in ECP computation references

* Update cmake patch command (issue pyscf#1998)

* Warning messages, and other improvements (pyscf#2009)

* Warning messages for MGGA in Hessian calculations

* Cache dynamically generated class

* Update the HDF5 tempfile handler

* Ensure shape of lattice vectors (issue pyscf#1932)

* feat(gto.basis): improve final, generic error message for basis not found

* add nr_nlc_vxc to NumInt

* Chkfile initial guess problem in ROHF and ROKS (issue pyscf#1986) (pyscf#2010)

* Fix chkfile initial guess problem for ROHF and ROKS (issue pyscf#1986)

* Use UHF natural orbital for RHF chkfile initial guess

* Enable some xc_deriv tests

* Enable xcfun compilation

* Print b3lyp warning when needed (issue pyscf#2029) (pyscf#2040)

* Hide b3lyp warning (issue pyscf#2029)

* Backward compatibility for xcfun

* typo

* Optimize the memory usage of GDF builder (pyscf#2047)

* Control the memory footprint for rsdf_builder and rsjk_builder (issue pyscf#2017)

* Adjust omega value in rsdf and rsjk

* Improve cistring.addr2str for large address (issue pyscf#1886)

* Handle magmom in Mole serialization (pyscf#2049)

* magmom serialization problem (issue pyscf#2026)

* Handle mesh

* Add tests for cell.dumps

* Fix the class of supercell scf object for the k2gamma conversion (pyscf#2030) (pyscf#2041)

* Fix the class of supercell scf object for the k2gamma conversion (pyscf#2030)

* typo

* Update k2gamma tests

* Refactor: error checks in C code and updates for GPU4PySCF library (pyscf#2048)

* Update to c99 standard; update dependency to libcint-6.1.1

* Update SCF functions for corresponding code in GPU4PySCF

* Add checks for malloc

* Update 02-ghf example according to issue pyscf#1970

* Update libcint/qcint version

* Remove the implicit dependency to mpi4py in __init__.py (issue pyscf#1888) (pyscf#2059)

* Remove the implicit dependency to mpi4py in __init__.py (issue pyscf#1888)

* import error

* Import error in momgfccsd tests

* Fix df-grad for UHF when symmetry is enabled (issue pyscf#2054)

* Check cylindrical symmetry of the orbitals for FCI solver (issue pyscf#2022)

* Fix bug for init_guess=atom (issue pyscf#2056)
Fix dft get_veff tests

* Improve magmom initialization (issue pyscf#2055)

* DIIS with damping (pyscf#2053)

* update simple damping and add DIIS damping

* update get_fock in solvent and dynamic level shift

---------

Co-authored-by: hongzhouye <>

* Add level shift to CPHF solver

* fix cp2k basis parse (issue pyscf#2043) (pyscf#2045)

* fix issue pyscf#2043

* add test for parse_cp2k

* allow symbol choosing for GTH PP parser

* SA-CASSCF Nonadiabatic Coupling Vectors (pyscf#2046)

* add nacs stuff but need to rebuild

* fix io warning test

* add sacasscf nac files from mrh

* update and add nac_method

* fix for flake8

* add example file

* update to assert len and uniqueness of states

* udpate

* udpate

* udpate

* fix assert statements and handling of state[0]==state[1]

* Backward compatibility issue in fci.cistring

* Release 2.5

* fix magmom check in mol.copy()

If magmom is an np.ndarray,
`if mol.magmom` will raise error.
Need to explicitly use `is not None`

* Move LebedevGrid C code to Python code; fix issue pyscf#2073 (pyscf#2076)

* Move LebedevGrid C code to Python code; fix issue pyscf#2073

* Restore CxLebedevGrid.c and fix bug

* lint error

* Add an API to dump system info

* multigrid DFT version 2 (pyscf#2078)

* Fix smearing with predefined chemical potential (pyscf#2098)

* added hubbard model with UHF example, fixed chemical potential bug in addons.py

* fix pbc smearing

* add test for smearing with mu0

---------

Co-authored-by: fishjojo <[email protected]>

* Add tests for ndarray_pointer_2d function (pyscf#2101)

* Add tests for ndarray_pointer_2d function

* Fix ndarray_pointer_2d

* Fix transform_ci for more than 64 orbitals (pyscf#2095)

* Fix transform_ci for more than 64 orbitals

* Separate occ_masks into a function

* fix binomial function in fci_string

* Fix xcfun high-order derivatives

* More type checks in fci solver (pyscf#2122)

* float type checks

* Update mcscf docstring (issue pyscf#2119)

* Fix failed tests

* Improve to_gpu (pyscf#2124)

* Update to_gpu

* Improve to_ks/to_hf

* Update to_gpu for df, solvent, qmm, and other methods

* Fix fci initialization

* Improve memory usage for RPA (pyscf#2115)

* improve memory usage for RPA with outcore ao2mo; fix small bug

* raise something when loading libpbc.get_Gv

* fix flake8 issue

* fix flake8 issue

* used DF-RHF as EXX, changed ref values

* unrestricted case

* unrestricted case

* Restore pyscf/pbc/gto/cell.py

* DFT wrapper supporting conventional DFT methods (pyscf#2084)

* added solvent models

* add example for RHF

* cleanup variables

* support casci casscf and ccsd

* uncomment unittests

* change example name

* update reset

* for flake8

* fixed a bug in soscf/newton_ah.py

* updated for recent master changes

* remove whitespace

* remove whitespace

* native support dispersion correction

* fixed a bug in pcm

* Update hf.py

* move dispersion to addons

* remove disp in RKS

* call get_dispersion in kernels

* added unit test for d4

* fixed dispersion correction in testing h2o

* updated ci

* skip unittest if dftd3 or dftd4 is missing

* new high level dft wrapper

* clean up & move dispersion to energy_tot

* remove disp from dft _keys

* fixed import dftd3_xc_map

* fixed typo

* added a wrapper for is_nlc

* replaced more ni.libxc.is_nlc with ni.is_nlc

* added wb97x-d3 as a special case

* added disp_3body

* removesuffix -> replace

* address qiming's comments

* add dft_parser && change back to libxc.is_nlc

* fixed unit test

* disable wb97x-d3, wb97x-d

* skip dftd3 for py3.12 & skip wb97x-d3

* Update dft_parser; Add tests for dft_parser

* Add warning messages

* Update test_h2o.py

---------

Co-authored-by: Qiming Sun <[email protected]>

* fix dipole moment in sfx2c1e

* debug and simplify to_gpu (pyscf#2149)

* support and simplify to_gpu

* remove comments

* flake8

* bug fix for RCCSD(T) with complex orbitals (pyscf#2141)

* bug fix for vooo order

* fix flake8

---------

Co-authored-by: hongzhouye <>

* Fix if mf.istype('UHF') for to_uhf() conversion

* Solve conflicts between @Property and __getattr__

* move static configurations from setup.py to pyproject.toml (pyscf#2144)

* move testing options to pytest.ini

* move static configurations from setup.py to pyproject.toml

* Update CI release jobs

---------

Co-authored-by: Qiming Sun <[email protected]>

* fix pbc df with KPoints input

* Fix pip builder

* reset enuc in _build_supcell_ (pyscf#2164)

* reset enuc in _build_supcell_

* add test for enuc reset

---------

Co-authored-by: hongzhouye <>

* Bug fix for single k-point JK-build in PBC DF (pyscf#2165)

* bug fix for j-build in get_jk

* bug fix for k-build in get_jk

---------

Co-authored-by: hongzhouye <>

* add missing init file

* DFTD3 & DFTD4 builder (pyscf#2161)

* local build dftd3/dftd4

* flake8

* fixed issues apis

* add compilation tool for mac

* fixed bugs in unittest

* fixed an issue in unit test

* add fortran compiler

* fortran compiler

* Move dftd3 and dftd4 builder to a separated repo (pyscf-dispersion)

* enable dftd3 & dftd4 for python3.12

* Adjust tests

---------

Co-authored-by: Qiming Sun <[email protected]>

* Kpoint dft to_hf error (fix issue pyscf#2157)

* ancient GCC compatibility

* fix

* fix: infinite recursion atom_hf call when ECP with super-heavy atoms (>Cm) (pyscf#2183)

* bugfix: resolve recursive initial guess when atomic charge > 96 (> Cm) and ecp activated

* Adjust AGF2 tests

---------

Co-authored-by: Qiming Sun <[email protected]>

* fix x_id in parse_token

* fixing unit test for dftd3 and dftd4

* Run CASCI and CASSCF without symmetry if symmetry is C1 (pyscf#2195)

* Run CASCI and CASSCF without symmetry if symmetry is C1

* groupname, not symmetry!

* move auxbasis_response out of __init__ (pyscf#2192)

* move auxbasis_response out of __init__

* relax unit test in test_c_agf2

* import hessian

* Update __init__.py

* improve the consistency in DF modules (pyscf#2204)

* improve the consistency

* added comment to _pinv

* tight conv_tol_cpscf by default

* set linear dependence threshold to 1e-9

* remove comments

* update unit tests

* SMD model and PCM Hessian (pyscf#2173)

* local build dftd3/dftd4

* flake8

* fixed issues apis

* add compilation tool for mac

* fixed bugs in unittest

* fixed an issue in unit test

* add fortran compiler

* fortran compiler

* Move dftd3 and dftd4 builder to a separated repo (pyscf-dispersion)

* enable dftd3 & dftd4 for python3.12

* Adjust tests

* add PCM and SMD, gradient and Hessian

* added __init__ files

* debug for ci

* flake8

* add more unit tests && refactor code

* Fix if mf.istype('UHF') for to_uhf() conversion

* Solve conflicts between @Property and __getattr__

* move static configurations from setup.py to pyproject.toml (pyscf#2144)

* move testing options to pytest.ini

* move static configurations from setup.py to pyproject.toml

* Update CI release jobs

---------

Co-authored-by: Qiming Sun <[email protected]>

* fix pbc df with KPoints input

* Fix pip builder

* reset enuc in _build_supcell_ (pyscf#2164)

* reset enuc in _build_supcell_

* add test for enuc reset

---------

Co-authored-by: hongzhouye <>

* Bug fix for single k-point JK-build in PBC DF (pyscf#2165)

* bug fix for j-build in get_jk

* bug fix for k-build in get_jk

---------

Co-authored-by: hongzhouye <>

* add missing init file

* Kpoint dft to_hf error (fix issue pyscf#2157)

* ancient GCC compatibility

* fix

* local build dftd3/dftd4

* rebase

* fixed bugs in unittest

* add fortran compiler

* rebase

* delete dftd3 and dftd4 files

* legacy SMD

* fortran compiler for mac

* gcc

* which gcc

* found fortran compiler

* fortran compiler

* reinstall gcc

* update unit test for pcm

* relax unit test in agf2

* resolved some comments

* add statement for smd source code

* refactor smd

* fixed bugs

* bugfix

* Skip compiling SMD by default

* Add fortran compiler in Macos build;
Fix conda build.

* empty line at EOF

* Fix SMD configuration in linux wheel

* add an example for SMD

* SkipTest -> skipTest

* excited states for PCM

* revert the changes in dispersion

* EOF

* added warnings in solvent modules when finite difference is used

---------

Co-authored-by: Qiming Sun <[email protected]>
Co-authored-by: Michal Krompiec <[email protected]>
Co-authored-by: chillenb <[email protected]>
Co-authored-by: fishjojo <[email protected]>
Co-authored-by: Hong-Zhou Ye <[email protected]>
Co-authored-by: Maximilian Scheurer <[email protected]>
Co-authored-by: Matthew Hennefarth <[email protected]>

* Density fitting performance optimization (re issue pyscf#1915) (pyscf#2205)

* use trsm in df.incore

* use np.matmul in df/df_jk.py

* faster df.outcore with trsm

* avoid numpy mem copy with h5py read_direct

* clarify chained matmuls

* use lib.dot to avoid syrk, df/outcore.py swap buffers

* Restore functionality for CDFT example (pyscf#2210)

* Restore functionality for CDFT example

I'm sure this example worked when it was authored, but that is no longer the case. This patch restores functionality.

* replace deprecated xrange with range

* add V step size to verbose output

It is one of the convergence criteria and should be shown

* add commentable option to verify consistency of example

* make post and pre diis implementations of the inner loop consistent

* add arg that pyscf expects

On my machine this wasn't needed but on others it was

* Move comments to docstring

* factor out cdft inner loop

* remove unused things

* correct errors in the documentation

* improve DFT APIs (pyscf#2187)

* support nlc = False && enhance dispersion correction

* Add DFT d3 examples

* reorg xc, nlc, and disp

* flake8

* Update dispersion error messages and examples

* updated example for dftd3

* Fix dispersion funciton names

* Fix dft sanity check

* Fix super() in KohnSham.check_sanity()

* Fix dftd3 test

* Conservative import

* print info in dispersion corrections

* change apis

* bugfix

* remove print

* bugfix

* added do_nlc check

* clarify do_nlc

* raise conflict between mf.xc and mf.nlc

* Cleanup disp logics and update disp examples

* Update dft_parser module

* Fix undefined-variable errors

* Fix dispersion parser

* Update example

---------

Co-authored-by: Qiming Sun <[email protected]>

* reomve some traces of unicode (py2) and some unused imports

* Update install dependencies

* Fix an import issue for X2C_KS (issue pyscf#2223)

* The spin_samples attribute for DKS

* Fix frozen for FNO (pyscf#2217)

* fix frozen for FNO

* fix FNO-UCC

* fix flake8

---------

Co-authored-by: hongzhouye <>

* DF dataset naming issue for rsh xc (fix pyscf#2236) (pyscf#2240)

* df dataset naming issue for rsh xc (fix pyscf#2236)

* fix tests

* Release 2.6

* bugfix in dispersion hessian (pyscf#2242)

* bugfix in dispersion hessian

* fixed bug in dispersion hessian

* added more unit test for dispersion

* unit test

* allow passing in options when creating HDF5 files (pyscf#2218)

* allow passing in options when creating HDF5 files

* Undo typo in ao2mo/outcore

* fix typo in h5filewrite_options example

* change default basis linear dependence threshold (pyscf#2228)

* change default basis linear dependence threshold

* sync setting in outcore.py

* unit test & linear dependence threshold

* linter

* 1e-12 -> 1e-14 for linear dependence

* lindep = 1e-13

* relax unit test

* Update test_df.py

* Update incore.py

* Update test_linalg_helper.py

* Update linalg_helper.py

---------

Co-authored-by: Qiming Sun <[email protected]>

* Abandon ROHF gradients and ROHF hessian (pyscf#2259)

* Add ruff check (pyscf#2262)

* Updates for numpy 2.0 release; Add ruff check; Cleanup tests

* Degeneracy issue in KGKS tests

* Adjust tests

* Fix a knumint bug when handling mcol and ncol

* Update ruff config

* Update doc strings (pyscf#2264)

* Update doc of eval_gto function

* Update geomopt examples (fix pyscf#2233) and atom_coords examples

* Improve pypi build pipeline (pyscf#2265)

* Adjust build order in release pipeline

* Update aarch64 build

* Handle dynamically registered post-HF attributes

* Patch release 2.6.1

* use all conda-forge compilers

* Updates due to NumPy 2.0 API changes (pyscf#2274)

* Updates due to NumPy 2.0 API changes

* Fix __array_wrap__

* Fix a bug caused by an inplace operation in linalg_helper._qr

* Fix inplace operation bug in pbc.cc.kintermediates_uhf

* Retore a few np.array(copy=True) statement

* Patch release 2.6.2

* Fix wheel build for setuptools 70 (pyscf#2278)

* Fix platform issue in bdist_wheel

* In case fail to import bdist_wheel

* Remove the ase module bulk from pyscf_ase (fix pyscf#2271) (pyscf#2279)

* Remove the ase module bulk from pyscf_ase (fix pyscf#2271)

* Fix get_ab of tddft for RSH functional (Fix pyscf#2261)

* Missing symbols

* Add tests

* typo

* Fix integer overflow for large ano basis (pyscf#2282)

* Fix integer overflow for large ano basis (fix pyscf#2201)

* More int type issues

* Fix Hessian atmlst and example (pyscf#2293)

* add atmlst to examples; fix examples/hessian/01-scf_hessian.py

* fix atmlst for hessian

* fix lint issue

* fix scipy.sparse.linalg.cg issue

* fix scipy.sparse.linalg.cg issue

* fix scipy.sparse.linalg.cg issue

* fix scipy.sparse.linalg.cg issue

* allow modifying tol and nroots in stability (pyscf#2299)

* allow modifying tol and nroots

* add example for tight tol

* pass kwargs to mf.stability

* fix typo

* Fix bugs in complex-valued contractions for direct_nosym (fix pyscf#2302)

* Handle api changes in scipy-1.14

* typo

* Fix initial guess dimension problems when solving multiple roots (fix issue pyscf#2292)

* Fix import dependency issue in SCF.do_disp()

* Fix a dimension bug in direct_spin1_cyl_sym. Update example for issue pyscf#2291

* compare version as int, not str (pyscf#2309)

* fixing a few typos

* another little wave of typos

* Fix addons.smearing RHF electron number (pyscf#2290)

* typos and typos

* a few more typos fixed

* Fix to kscf for dft (pyscf#2336)

* fix bug in convert_to_kscf

* fix bug in convert_to_kscf

* fix kuhf case in convert_to_kscf

* Fix MP2 FNO bug when pct_occ=1 (pyscf#2337)

* Update mp2.py

* Update ump2.py

* Superposition of Atomic Potentials (SAP) initial guess implementation for HF-SCF calculations (pyscf#2301)

* Added VSAP initial guess implementation for Hartree-Fock SCF
calculations (described by S. Lehtola here:
https://doi.org/10.1021/acs.jctc.8b01089)
Modified fakemol_for_charges to be usable for SAP 3C2E integrals.
Supplied sap_grasp coefficients computed by S. Lehtola.

* Added 'large' and 'small' sap coefficient functionality

* Fixed some bugs in recognition of DFT and HF objects and their SAP
guesses. Made other cleanups and fixes proposed by S. Lehtola.

* Changed logger.error to logger.warn

* Renamed HF SAP guess functions from 'vsap' to 'sap' to avoid confusion
with DFT counterparts using a different method with same name.

* Fixed a bug I had created by accident.
Created new fakemol_for_cgtf_charge function to be used in forming SAP guess
and to avoid breaking other functionality which use the
fakemol_for_charges function.

* Removed redundant attribute check

* vsap -> sap

* charge sumcheck tolerance set to 1e-6

* Apply suggestions from code review

Co-authored-by: Susi Lehtola <[email protected]>

* Changed sap initial guess to accept basis set in internal format
(python dict) or as a string (will look for basis file locally and
from basis set exhcange if bse api is installed)

* Added SAP guess to restricted open-shell Hartree-Fock (ROHF)

* Added SAP guess to non-relativistic generalized Hartree-Fock (GHF)

* Tidy up SAP function in GHF, make sap_grasp_large to default SAP basis

* Add example for SAP inital guess.

* Added SAP guess to Dirac-Hartree-Fock (DHF)

* small cleanups

* SAP now accepts internally formatted python dictionary as basis. SCF object has new attribute sap_basis

* Apply suggestions from code review

Co-authored-by: Susi Lehtola <[email protected]>

* Clarified example using explicit dictionary format basis for SAP

* Apply suggestions from code review

Co-authored-by: Susi Lehtola <[email protected]>

* Comment for vsap added back

* Fixed  issue

* Fixed style issues

* Fix style issue with extraneous 'f' prefix

* fix ROHF SAP guess call tree, add SAP unittest for rhf and rohf

* Fixed accidental incorrect test

* Add SAP guess to UHF

* Add SAP unittest for uhf

* Add SAP unittest for GHF

* Add SAP unittest for DHF

* flake8 warning

---------

Co-authored-by: Susi Lehtola <[email protected]>
Co-authored-by: Qiming Sun <[email protected]>

* Use weakref for automatic deletion of tmpfiles (to fix pyscf#2318) (pyscf#2320)

* Use weakref for automatic deletion of tmpfiles

* github actions tests check number of leftover temporary files

* temporary file check pytest exit status

* Avoid h5f.flush() on closed file

---------

Co-authored-by: Qiming Sun <[email protected]>

* Updates to the EA/IP-ADC code

* Inteface changes

* Changes interface to the 1-RDM functions in IP/EA-RADC

* Fixed accidental copy

* Fixed accidental copy

* Fixed Contractions radc_ip.py

* Moves make_ref_rdm1 to the appropriate location

* IP tests 1RDM Ref and Exc. Up to ADC(3)

* Fixed bugs that were breaking EA tests

* Clean up

* Changes to the UADC class

* EA Tests Completed from SO code. Slight cleanup for IP tests

* Reference 1RDM tests N2 EA + IP up to ADC(3)

* N2 1RDM EA up to ADC(3) tests

* Completed RADC(3)  1RDM N2 IP tests

* Example file added for 1RDM dipole moment examples with various methods

* Small changes to comments in 1RDM test file

* Final Comment Change 1RDM examples

* Formatting changes and fixing approx_trans_moments option

* ROHF code cleanup

* Fixes a test for IP-ADC

* UADC class clean-up

* IP/EA classes clean-up

* Add singles correlation energy

* More class clean-up, added dipole moment integrals

* Author list update

* Addresses lint errors

* Addressing flake issues

* Few more flake issues addressed

* Adding the EE-ADC code from the old branch

* Fixing osc strengths calculation

* Testing oscillator strengths

* Oscillator strengths do not match the old implementation. Investigating...

* fixed osc problem, TY_xx.append(TY_x) not ordering things right for oscillators but this was not a problem for analyzing spec factors

* Changes in the UADC spin contamination interface

* fixed rohf bugg added ooov and vvvo density matrices and variants

* working on density efficient matrices

* added efficent density matrices

* deleted some lines

* added the radc_ee file that is most up to data

* DF low mem bug fixed

* made some updates, need to work on flake

* radc_ee.py passes flake8, working on uadc_ee.py file

* uadc_ee.py file passed flake

* Renames spin contamination functions

* files pass flake test

* Proper handling of compute_properties attribute in RADC

* Cleaning up imports

* Makes sure that CVS-UADC tests pass

* Updates to RADC tests and UADC default convergence thresholds

* updated convg thresholds

* Example updates and some other interface changes

* One more interface change for the RADC code

---------

Co-authored-by: Alexander Sokolov <[email protected]>
Co-authored-by: Alexander Sokolov <[email protected]>
Co-authored-by: sokolov-group <[email protected]>
Co-authored-by: Nikolay A. Bogdanov <[email protected]>
Co-authored-by: Frédéric Chapoton <[email protected]>
Co-authored-by: Susi Lehtola <[email protected]>
Co-authored-by: Junjie, Yang <[email protected]>
Co-authored-by: Qiming Sun <[email protected]>
Co-authored-by: Hongzhou Ye <[email protected]>
Co-authored-by: Aniruddha Seal <[email protected]>
Co-authored-by: Matthew R Hermes <[email protected]>
Co-authored-by: Matthew R Hermes <[email protected]>
Co-authored-by: Nikolay Bogdanov <[email protected]>
Co-authored-by: Max Nusspickel <[email protected]>
Co-authored-by: Xiaojie Wu <[email protected]>
Co-authored-by: Xing Zhang <[email protected]>
Co-authored-by: peter-reinholdt <[email protected]>
Co-authored-by: fishjojo <[email protected]>
Co-authored-by: Chong Sun <[email protected]>
Co-authored-by: Verena Neufeld <[email protected]>
Co-authored-by: Jiachen Li <[email protected]>
Co-authored-by: Timothy Berkelbach <[email protected]>
Co-authored-by: liwt31 <[email protected]>
Co-authored-by: Hung Pham <[email protected]>
Co-authored-by: hung.pham <[email protected]>
Co-authored-by: Zhihao Cui <[email protected]>
Co-authored-by: Felipe S. S. Schneider <[email protected]>
Co-authored-by: hongzhouye <unknown>
Co-authored-by: matthew-hennefarth <[email protected]>
Co-authored-by: Victor Yu <[email protected]>
Co-authored-by: Xubo Wang <[email protected]>
Co-authored-by: xubwa <[email protected]>
Co-authored-by: Michal Krompiec <[email protected]>
Co-authored-by: chillenb <[email protected]>
Co-authored-by: Maximilian Scheurer <[email protected]>
Co-authored-by: Zhenyu Zhu ajz34 <[email protected]>
Co-authored-by: jeanwsr <[email protected]>
Co-authored-by: scohenjanes5 <[email protected]>
Co-authored-by: Joonatan Huhtasalo <[email protected]>
Co-authored-by: Susi Lehtola <[email protected]>
Co-authored-by: serna.40 <[email protected]>
Co-authored-by: Terrence Stahl <[email protected]>