v0.1.0

What's Changed

• Migrate to json as input file format over my home cooked formatting by @JorgeG94 in #116
• read effective core potentials from the BSE basis files by @JorgeG94 in #119
• feat/python interface by @JorgeG94 in #120
• fix/scf convergence by @JorgeG94 in #121
• feat/checkpoint restart by @JorgeG94 in #122
• feat/hdf5 checkpoints by @JorgeG94 in #123
• feat/properties by @JorgeG94 in #124
• feat/cpu backend by @JorgeG94 in #125
• feat/cpu fragmented hf by @JorgeG94 in #126
• density-fit J and K, the way cuEST always does by @JorgeG94 in #127
• feat/libxc by @JorgeG94 in #128
• Feat/scf direct by @JorgeG94 in #130
• route density-fitted exchange through the occupied orbitals by @JorgeG94 in #131
• exercise the Python interface in CI, and document it by @JorgeG94 in #132
• add Lebedev-Laikov angular quadrature by @JorgeG94 in #134
• Feat/dft grid radial by @JorgeG94 in #135
• Feat/dft grid partition by @JorgeG94 in #136
• Feat/dft grid assembly by @JorgeG94 in #137
• Feat/dft grid pruning by @JorgeG94 in #138
• Chore/backends dir by @JorgeG94 in #139
• make density fitting reachable, and stop losing backend errors by @JorgeG94 in #141
• clean by @JorgeG94 in #142
• Docs/readme python by @JorgeG94 in #143
• let the release architecture flags be overridden by @JorgeG94 in #118
• ignore CMake build trees under any name by @JorgeG94 in #144
• route Cartesian basis sets instead of reading them as spherical by @JorgeG94 in #145
• Feat/cpu validation breadth by @JorgeG94 in #146
• Perf/general contraction by @JorgeG94 in #148
• Give the DF 3c2e fitting integrals an optimizer by @JorgeG94 in #149
• Add unrestricted Hartree-Fock to the CPU backend by @JorgeG94 in #150
• Add conventional MP2 with spin-component scaling by @JorgeG94 in #151
• Add RI-MP2 by @JorgeG94 in #152
• Speed up the three-centre driver and RI-MP2 energy loop by @JorgeG94 in #153
• Optimize the MP2 four-index transform by @JorgeG94 in #154
• Fix Python-driven fragmented runs: broken bonds, threading, and per-fragment charge by @JorgeG94 in #156
• Feat/cpu atomic guess by @JorgeG94 in #157
• Add spin-orbital CCSD and (T) on the CPU by @JorgeG94 in #158
• Wire coupled cluster into the input deck by @JorgeG94 in #159
• Add RI-CCSD and RI-CCSD(T) on the CPU by @JorgeG94 in #160
• Chore/validation inputs tree by @JorgeG94 in #161
• document the methods that arrived without docs by @JorgeG94 in #162
• Feat/cpu dft by @JorgeG94 in #163
• Feat/dft wiring by @JorgeG94 in #164
• unrestricted Kohn-Sham, LDA through range-separated hybrid by @JorgeG94 in #165
• wire cuEST's polarizable continuum into the SCF by @JorgeG94 in #166
• Fix/cray by @JorgeG94 in #167
• Feat/efp makefp by @JorgeG94 in #168
• Mqc geoopt by @JorgeG94 in #170
• uptick deps by @JorgeG94 in #171
• Feat/efp efp by @JorgeG94 in #172
• Feat/cpu integral perf by @JorgeG94 in #173
• Fix CPHF non-convergence: keep density-weighted screening out of the response operator by @JorgeG94 in #184
• Fix flaky distributed-polarizability test: tie on the yy diagonal by @JorgeG94 in #185
• Weight integral screening by the density it multiplies by @JorgeG94 in #174
• Build the Fock matrix from the density change, not the density by @JorgeG94 in #175
• Carry a converged small-basis density into a larger basis as a guess by @JorgeG94 in #176
• Analytic RHF and UHF gradients on the CPU backend by @JorgeG94 in #177
• Density-fitted RHF gradients by @JorgeG94 in #178
• Kohn-Sham gradients, LDA by @JorgeG94 in #179
• GGA and hybrid-GGA Kohn-Sham gradients by @JorgeG94 in #180
• validation by @JorgeG94 in #186
• mp2 grads by @JorgeG94 in #187
• mgga grads by @JorgeG94 in #190
• rimp2 grads by @JorgeG94 in #188
• Feat/cpu gradients mgga by @JorgeG94 in #191
• Feat/cpu gradients df dft by @JorgeG94 in #192
• Feat/sapt by @JorgeG94 in #193
• load each basis set once instead of once per element by @JorgeG94 in #194
• keep hand-maintained cases in the manifest across a regeneration by @JorgeG94 in #195
• Wiberg–Mayer bond orders from xTB by @JorgeG94 in #197
• Mulliken and CHELPG atomic charges by @JorgeG94 in #198
• FMO and EE-MBE to arbitrary order by @JorgeG94 in #199
• Verbose convergence tables (SCF/FMO/CCSD/geo-opt), makefp SCF fixes, and friends by @JorgeG94 in #201
• Start the SCF from Wolfsberg-Helmholz, not from the core Hamiltonian by @JorgeG94 in #202
• Feat/cpu xc kernel by @JorgeG94 in #203
• Feat/counterpoise by @JorgeG94 in #204
• Chore/validation to testdrive by @JorgeG94 in #206
• enable DF for DFT and RI gradients by @JorgeG94 in #208
• Docs/counterpoise by @JorgeG94 in #209
• amek ricc fast by @JorgeG94 in #210
• scale diis error by @JorgeG94 in #211
• make makeefp good by @JorgeG94 in #212
• Feat/mgga kernel by @JorgeG94 in #214
• Give the continuum a gradient by @JorgeG94 in #216
• add optional cuest option the json and uptick pic-mpi for a cray filaure by @JorgeG94 in #205
• make HF faster by @JorgeG94 in #217
• Feat/aambs basis by @JorgeG94 in #218
• Document the quasi-atomic bonding analysis by @JorgeG94 in #221
• Feat/mcscf by @JorgeG94 in #219
• Add a spatial-orbital coupled cluster path by @JorgeG94 in #223
• Turn the spatial CCSD iteration into matrix products by @JorgeG94 in #224
• Batch the perturbative triples over the innermost virtual by @JorgeG94 in #225
• Cover RI-CCSD(T) on the spatial path by @JorgeG94 in #226
• Give the fitted ladder both algorithms and choose between them by @JorgeG94 in #227
• Thread the perturbative triples over virtual pairs by @JorgeG94 in #228
• Thread the CCSD iteration, conventional and fitted by @JorgeG94 in #229
• ormas by @JorgeG94 in #231
• Feat/ieda by @JorgeG94 in #233
• Feat/optional mpi by @JorgeG94 in #236
• Feat/ieda complete by @JorgeG94 in #235
• Chore/python interface by @JorgeG94 in #234
• build with NVHPC: three fixes by @JorgeG94 in #232
• make mcscf safee by @JorgeG94 in #237
• make thresholds powerful by @JorgeG94 in #238
• Move over to libfint over libcint by @JorgeG94 in #213
• Feat/fukui indices by @JorgeG94 in #239
• Take pic-blas v0.3.1 in the fpm build too by @JorgeG94 in #240
• Share the constants and helpers the backends had copied by @JorgeG94 in #183
• Put the two EFP2 cases back in the manifest by @JorgeG94 in #241
• Feat/analytic hessian rhf by @JorgeG94 in #243
• enable Fukui on DFT exceptn on dh by @JorgeG94 in #245
• table header by @JorgeG94 in #248
• ump2 and urimp2 by @JorgeG94 in #247
• Stop billing the Fock build to the exchange-correlation quadrature by @JorgeG94 in #253
• Chore/unit tests by @JorgeG94 in #242
• Feat/quao ci transform by @JorgeG94 in #244
• cc timings but good by @JorgeG94 in #255
• pcm by @JorgeG94 in #254
• afo by @JorgeG94 in #256
• agents MD for testing copilot reviews by @JorgeG94 in #258
• Perf/iterative tdhf by @JorgeG94 in #249
• Feat/efp keywords by @JorgeG94 in #250
• sapt2 by @JorgeG94 in #260
• Build HDF5 in CI, which nothing here was doing by @JorgeG94 in #263
• Tell the review bot where the bugs actually are by @JorgeG94 in #264
• Analytic MCSCF gradients, and stop the block size deciding the thread count by @JorgeG94 in #262
• Fix/unconverged fragment report by @JorgeG94 in #257
• Stop superseded CI runs, except on main by @JorgeG94 in #266
• unrestricted couples cluster by @JorgeG94 in #259
• Show the ion SCFs a Fukui run solves by @JorgeG94 in #270
• Feat/charges any reference by @JorgeG94 in #272
• Let the deck decide whether the Fukui ions print their iterations by @JorgeG94 in #273
• Perf/dft benchmarks by @JorgeG94 in #276
• Build DL-FIND in CI, and make it take a step by @JorgeG94 in #277
• make DFT faster by @JorgeG94 in #278
• Fukui with PCM by @JorgeG94 in #279
• fix hardness bug by @JorgeG94 in #280
• Feat/scf level shift by @JorgeG94 in #271
• Stop a Fortran comment from breaking the Ninja build by @JorgeG94 in #281
• Feat/linear dependence report by @JorgeG94 in #274
• Perf/xc uks kernel by @JorgeG94 in #283
• Fix/deltascf ion guess by @JorgeG94 in #282
• Give the unrestricted SCF a clock, so its XC column means something by @JorgeG94 in #285
• Store coupled-cluster amplitudes in HDF5, so a killed run can resume by @JorgeG94 in #265
• Feat/crest by @JorgeG94 in #286
• Feat/opt hess end by @JorgeG94 in #287
• Document how the Kohn-Sham energies are checked, and pin the reference by @JorgeG94 in #288
• Feat/vv10 by @JorgeG94 in #289
• screen in DFT gradients by @JorgeG94 in #291
• Report linear dependence once for a fragmented run, not once per fragment by @JorgeG94 in #290
• Let a deck say it is looking for a saddle, and read the result that way by @JorgeG94 in #292
• Run a band between two structures, and let its climbing image find the top by @JorgeG94 in #293
• Hunt a saddle without ever building a Hessian by @JorgeG94 in #294
• Say what the saddle connects, by walking off it in both directions by @JorgeG94 in #295
• Stop P-RFO from climbing a rotation by @JorgeG94 in #296
• Feat/mulliken charges by @JorgeG94 in #300
• Call cuEST's VV10 by @JorgeG94 in #298
• pr reviewer skill by @JorgeG94 in #310
• Analytic exchange-correlation second derivatives, LDA through meta-GGA by @JorgeG94 in #302
• Range-separated Hessians: the omega that never reached the integral by @JorgeG94 in #303
• Let a deck reach the Kohn-Sham Hessian by @JorgeG94 in #304
• model.cartesian: ask for the angular form rather than inherit it by @JorgeG94 in #305
• The VV10 gradient, so a -V functional can be optimized by @JorgeG94 in #306
• The VV10 Hessian, and the analytic path opened to -V functionals by @JorgeG94 in #307
• MP2 and RI-MP2 gradients learn a frozen core by @JorgeG94 in #308
• Fix/fukui level shift by @JorgeG94 in #311
• make rimp2 fast by @JorgeG94 in #312
• Feat/mp2 hessian by @JorgeG94 in #314
• enable ecp in cpu by @JorgeG94 in #313
• some light opt for mp2 hessian by @JorgeG94 in #315
• cake got awful by @JorgeG94 in #322
• different grids for the vv correction by @JorgeG94 in #320
• not show xc in scf if not used by @JorgeG94 in #321
• refactor scf config by @JorgeG94 in #318
• refactor by @JorgeG94 in #323
• diis but better by @JorgeG94 in #317
• use diis for conv by @JorgeG94 in #324
• Feat/cuest cuda13 by @JorgeG94 in #325
• Evaluate VV10 once for an unrestricted Kohn-Sham step, not twice by @JorgeG94 in #326
• Make the SCAN family reachable, and refuse half a functional by @JorgeG94 in #327
• refactor delta scf config by @JorgeG94 in #329
• remove hard coding of hess displacement by @JorgeG94 in #330
• Build the atomic guess the deltaSCF ions ask for by @JorgeG94 in #332
• refactor scf passing arguments into makeefp and other scf routiens by @JorgeG94 in #331
• Docs/scf guess by @JorgeG94 in #334
• Feat/scf convergence metric by @JorgeG94 in #333
• analytical dh DFT Hessians by @JorgeG94 in #335
• frag method in fragmentation was being stupid by @JorgeG94 in #336
• Fix/cuest dropped settings by @JorgeG94 in #339
• Feat/analytic ir intensities by @JorgeG94 in #337
• make CI faster by @JorgeG94 in #340
• Fix/verified defects by @JorgeG94 in #338
• cleanup by @JorgeG94 in #342
• Name the CPU backend cenzontle by @JorgeG94 in #343
• Put the SCF and the CI solvers inside methods/, not beside it by @JorgeG94 in #345
• catch the python api up by @JorgeG94 in #344
• Use the ECOP by @JorgeG94 in #346
• run by @JorgeG94 in #328
• make gradient and gessian fast by @JorgeG94 in #347
• refactor rhf iteration by @JorgeG94 in #348
• spherically average the Cartesian guess by @JorgeG94 in #349
• use threads in crest by @JorgeG94 in #350
• use large_info logger levels added in pic 0.6.1 by @JorgeG94 in #352
• enable terco backend by @JorgeG94 in #351

Full Changelog: v0.0.1...v0.1.0