1-------------------------------------------------------------------------------- 2rel1_73_3 3-------------------------------------------------------------------------------- 4dff558a Make delta_ member of LineMinimizer, set to 1.1 5641deae Fix RSHFunctional Fx enhancement factor 6550005c Fix ekin_e for nspin=2 CP dynamics 7493bdb0 Change delta to 0.3 in LineMinimizer.cpp (undone later) 89bd517d Add util/qbox_vavg.py for average potential calcs 9-------------------------------------------------------------------------------- 10rel1_73_2 11-------------------------------------------------------------------------------- 12Fix testSlaterDet output 13f87029e Add numroc0 function in Matrix 14d1d39b5 Use MPIdata in testMatrix.cpp 151ccde59 Rewrite Matrix::nrm2() implementations 167d1079e Clean up syntax in JDWavefunctionStepper.cpp 1790d6d8c Fix nrm2() of ComplexMatrix 18db88c3f Fix print format in testMatrix.cpp 19-------------------------------------------------------------------------------- 20rel1_73_1 21-------------------------------------------------------------------------------- 22148f560 Fix debug variable. Fix debug STRESS in tests. 23-------------------------------------------------------------------------------- 24rel1_73_0 25-------------------------------------------------------------------------------- 26c50970e Fix Makefile, testSample, testSampleReader 279222a00 Add ANDIonicStepper, update AndersonMixer 281a0ed95 Add DELTA_RATIO debug var in BOSampleStepper 29828eb57 Implement debug map 30-------------------------------------------------------------------------------- 31rel1_72_3 32-------------------------------------------------------------------------------- 335780a4f Fix printing of MLWF data 347620e48 Fix printing of bisection size and pair fraction 35-------------------------------------------------------------------------------- 36rel1_72_2 37-------------------------------------------------------------------------------- 38f9a03e2 Fix G=0 component of wkerker 398df4f96 Add srand48 call to Wavefunction::randomize 40533fcf8 Fix printing of occ in update_occ 41-------------------------------------------------------------------------------- 42rel1_72_1 43-------------------------------------------------------------------------------- 442067ca0 Fix Occ.h for -nspb 2 case 4548200a9 Fix SpectrumCmd for new data layout 46-------------------------------------------------------------------------------- 47rel1_72_0 48-------------------------------------------------------------------------------- 492396ba7 merged comm branch, implementing new data layout 50New command line parameters: -nstb -nkpb -nspb 51The nrowmax variable is removed 52-------------------------------------------------------------------------------- 53rel1_71_5 54-------------------------------------------------------------------------------- 553635f3a Exit response calc if fractionally occ states 569d34423 Fix ecut in vext test.i file 570933f63 Add function3d.xsd schema 5806767fe Fix upf2qso nlcc core charge 59-------------------------------------------------------------------------------- 60rel1_71_4 61-------------------------------------------------------------------------------- 6251217ee Fix BasisMapping single-task map for np0!=np1 63712c6f8 Set nrowmax to ntasks if larger than ntasks 644c3f1b1 Fix precision in bisection time output 65-------------------------------------------------------------------------------- 66rel1_71_3 67-------------------------------------------------------------------------------- 68e7ea41f Make test_fftw no-op if USE_FFTW2 not defined 6985c9d9a Remove Id: strings 7092b4f86 Update Makefile, change .C to .cpp in src files 71a8f7b58 Update Makefiles for .C to .cpp rename 7251e5bf5 Rename .C files to .cpp 73b3e26db update release.C 74-------------------------------------------------------------------------------- 75rel1_71_2 76-------------------------------------------------------------------------------- 7700ce1b0 Optimize single-task transpose in BasisMapping 78a28add8 Add sge and slurm job scripts for multiple runs 79-------------------------------------------------------------------------------- 80rel1_71_1 81-------------------------------------------------------------------------------- 820b54b01 Merge branch 'develop' 8304bf383 Fix timing msg in Bisection.C 848ad14a3 Fix bisection timer for nspin==2 854b2de68 Fix constness in BasisMapping 861a2fb01 Enable spin-polarized CP dynamics 872aad1f5 Add cstdlib header for abs() 882d6f542 Use vector swap in BasisMapping for single-node 89e9d9631 Use memset to zero arrays in BasisMapping 9090f6ddc Update gather/scatter in BasisMapping for MKL 914f1caa4 Fix timers in testFourierTransform 922311fdd Add fxy and fz functions in FourierTransform 93984676c Cleanup FourierTransform and BasisMapping 948dc7100 Rewrite BasisMapping and FourierTransform 9557f8d91 Switch names of transpose fns in BasisMapping 96-------------------------------------------------------------------------------- 97rel1_71_0 98-------------------------------------------------------------------------------- 99a26cbfa Update constraints before ionic step 10088181a8 Change sign of force on constraints and pair force 101-------------------------------------------------------------------------------- 102rel1_70_0 103-------------------------------------------------------------------------------- 104c790914 Fix xsd links in upf2qso dir 1052348554 Update test/Makefile 106785af78 Add test/siatom test of custom occupation numbers 1070af0f85 Add occ variable for custom occupation numbers 108-------------------------------------------------------------------------------- 109rel1_69_1 110-------------------------------------------------------------------------------- 111cb15c6c Fix output formatting in SpectrumCmd.C 11294915e6 Fix gaussian normalization in SpectrumCmd.C 1139f4d98f Remove tabs in ExchangeOperator.C 114-------------------------------------------------------------------------------- 115rel1_69_0 116-------------------------------------------------------------------------------- 117beb8bee Cleanup trailing spaces in util scripts 1183137cbd Add scripts qbox_reduce.sh, qbox_move_subsample.py 1198d434af Fix spectrum file for nspin==2 1207a00811 Update spectrum command 1216ba92e5 Update spectrum command help msg 12280c343d Initialize eig with zero in SlaterDet 1234031b74 Modify spectrum command parameters 124539499d write spectrum on file, fix parallel SpectrumCmd 12554058fe Add spectrum cmd to Makefile and qb.C 126fc34625 Add spectrum command 127-------------------------------------------------------------------------------- 128rel1_68_4 129-------------------------------------------------------------------------------- 130d9bac45 Use syevd/heevd in JD stepper, workaround lwork size calculation 131-------------------------------------------------------------------------------- 132rel1_68_3 133-------------------------------------------------------------------------------- 13441edad7 Fix issue fail to restart MD with multiple k-points 135-------------------------------------------------------------------------------- 136rel1_68_2 137-------------------------------------------------------------------------------- 138260b6be Remove debug print in ExchangeOperator.C 1390d66277 Fix Matrix.C calc of lwork, return values 140a0ff3b7 Fix D3tensor.h 14139fd798 Changed default to all atomsets in qbox_xyz.py 142-------------------------------------------------------------------------------- 143rel1_68_1 144-------------------------------------------------------------------------------- 145b2326e0 Remove unneeded loop in XCPotential.C 1466ce2309 Fix white space, reorder tests in SCANFunctional.C 147186b7a8 Cleanup ExchangeOperator functions params and return values 148e1a0d16 Modified SCANFunctional to not crash with grad=0, or negative density 14952c16a2 Set echo=true in qb.C to echo commands in server mode 150-------------------------------------------------------------------------------- 151rel1_68_0 152-------------------------------------------------------------------------------- 15305cd227 Merge branch 'scan' into develop 1545c8ff65 Make Sample const in ExchangeOperator 15515649b6 Make nlevels const in Bisection 1568afc2a0 Add qbox_species_temp.sh utility script 1574f46485 Add scripts to extract position,velocity,force 158-------------------------------------------------------------------------------- 159rel1_67_4 160-------------------------------------------------------------------------------- 1616e23b88 Merge branch 'develop' 162c1261b7 Fix units of stress_tol test in BOSampleStepper 163-------------------------------------------------------------------------------- 164rel1_67_3 165-------------------------------------------------------------------------------- 16648c0d4b Merge branch 'develop' 167211d894 Fix Harris-Foulkes estimate for NLCC when nspin=2 168-------------------------------------------------------------------------------- 169rel1_67_2 170-------------------------------------------------------------------------------- 1719459eed Merge branch 'develop' 1721d85485 Use daxpy in XCPotential loops 173136f424 Fix total energy and stress calc with NLCC 1744e6d0f5 Fix non-linear core correction 175c59340b Add rhocore_r and rhocore_g to ChargeDensity 176d010476 Fix initialization of core correction in Species.C 177593b1d1 update release string to 1.9 178ddcb699 Restore 1.0 coefficient of non-local projectors to follow QE definitions 179715e140 Add fast return in NonLocalPotential if no non-local species 18072caf01 Add factor 0.5 in upf2qso.C for consistency of energy units 181a52c966 Add <algorithm> header in upf2qso.C for g++ compilation 182797585d Added ISO date stamp in upf2qso output 1834f93df7 Fix upf2qso.C to work with UPF 2 pseudopotentials 184-------------------------------------------------------------------------------- 185rel1_67_1 186-------------------------------------------------------------------------------- 18700fb07e Add missing files ForceTol.h and StressTol.h 188-------------------------------------------------------------------------------- 189rel1_67_0 190-------------------------------------------------------------------------------- 191c1bad59 Restore scaling of alpha_start in CGIonicStepper.C 1928ab5bfa Fix force_tol and stress_tol convergence test in BOSampleStepper.C 193a7c7157 Fix test convergence test of force_tol and stress_tol 19461ac1c1 Add optional VERSION macro in qb.C and Makefile 1950ed5ed8 Add force_tol and stress_tol variables 196ef0e7eb Adjust params in CGIonicStepper, remove rescaling of alpha_start 1974a664cb Change beta_max in testCGOptimizer 19866305fd Add set_nstep_max function to LineMinimizer 199e800eeb Fix PlotCmd argument parsing for -vlocal with nspin=2 200-------------------------------------------------------------------------------- 201rel1_66_3 202-------------------------------------------------------------------------------- 20323da694 Adjust atomic radii. Update logic of non-scf loop termination. 20424ec148 Fix pure virtual functions declarations in SampleStepper.h 2050cae884 Fix MPI_Comm leak in SlaterDet copy ctor 206ffa583a Remove old kpgen.C, replaced by util/kpgen/kpgen.cpp 2072e5500e Update util/kpgen files 208c67d4c3 Fix units of ext_stress 209648574e Update mk file for MacBook Pro Mojave 210d51cdb9 Fix precision of eigenvalue_sum output 211-------------------------------------------------------------------------------- 212rel1_66_2 213-------------------------------------------------------------------------------- 2142fdf056 Fix Function3d byte swapping on big endian systems 215f31f37e Fix check of argc in Vext.h 216-------------------------------------------------------------------------------- 217rel1_66_1 218-------------------------------------------------------------------------------- 2192d07308 Add cstring header in FourierTransform.C 220-------------------------------------------------------------------------------- 221rel1_66_0 222-------------------------------------------------------------------------------- 223124fb28 Fix test causing fail on assert in FCC cell 224Merge vext branch 225-------------------------------------------------------------------------------- 226rel1_65_1 227-------------------------------------------------------------------------------- 228a23e645 Remove static_cast and cleanup RSH and HSE functionals 229de21bb8 Cleanup test programs 230a312251 Add error msg in BisectionCmd if nkp > 1 231cd77e32 Print np0v,np1v,np2v as XML elements in EnergyFunctional 2325abca0d Compute and print np0v,np1v,np2v in StatusCmd 233dc43a96 Add README and COPYING files 2340827954 Rewrite and simplify UserInterface 235218356e Update Makefile dependencies 236df25ed5 Fix missing continuation character 23758547ce Add basis_fits_in_grid function 238e66b7c0 Replace zeroing loops with memset in FourierTransform.C 2395af642d Add stress tests to test Makefile 240-------------------------------------------------------------------------------- 241rel1_65_0 242-------------------------------------------------------------------------------- 243fb22565 Merge branch 'develop' 244939c153 Add Mac OS makefile 2453ffc3f0 Fix -atomic_charge initialization when net_charge!=0 246e997363 Fix UserInterface processing of empty lines 247ca27b6e Add centos7 and centos7_intel makefiles 248707eadd Use StrX in XML handler classes to avoid mem leaks 249c4e49cf Fixed calls to align wfs in BOSampleStepper.C 2500436991 Add BHandHLYP hybrid functional 251f3b32ed Add qbox_move_to.py script 2525298a69 Added qso_plot_species.py plot utility 253fbf9477 Removed unused code from HSE and RSH functionals 25475919ee Test negative density in HSE and RSH functionals 255c940023 Fix text-form in SlaterDet::write 256-------------------------------------------------------------------------------- 257rel1_64_0 258-------------------------------------------------------------------------------- 259d4c5c81 Merge branch 'develop' 26028c1fc2 Merge branch 'rsh-dev' into develop 26138843c5 Removed unused variables 26252cb12a Implemented stress for general RSH case 263211d90a adjust blocking factors of square matrices 2647c62abc Implement reset_rotation command 2650b1e4fa fix help msg in MoveCmd.h 26684416bf Reorganized stress tests as si4stress and si2stress 2677e0b69a Added tests for invalid bisection parameters 2685f7c1c7 Adjust strain values in test/stress files 2690d593a2 Cleanup files in test/stress dir 2709c5654f Add test of stress tensor 2716f049b7 Fixed x_coeff in RSHFunctional.C Cleanup & comment 272af533ce Fix errors in ExchangeOperator at gamma 2735e24b92 Cleanup ExchangeOperator 274ccd020c First RSH implementation 27552e6976 Merged HSE implementation 276fbb7421 Merge branch 'hse-dev' into develop 2775b93ad8 Add cstring header in ConstraintSet.C 27872e55a8 Fix force normalization in PairExtForce::energy 279000b3db Fixed charge integral in ChargeDensity::update_rhor 28080f2730 Fix long lines 2811c14353 Merge branch 'develop' into hse-dev 282e9d66bf Add zero volume test to run command 283667cce0 Updated centos6.mk 284f9b90ad Updated notes 285-------------------------------------------------------------------------------- 286rel1_63_11 287-------------------------------------------------------------------------------- 288be7d4cc rel1_63_11 2894e6e1b6 adjust blocking factors of square matrices 290-------------------------------------------------------------------------------- 291rel1_63_10 292-------------------------------------------------------------------------------- 293058730d Implemented PSDA for complex wave functions 29426c3495 Update printing of electronic charge & net charge 295c102a32 Add testSampleReader program 2968f85701 Add option to define constraint with current value 29735f2374 changed output format of constraint value, force 298e85f85d Changed logic of shifts: zero shifts include Gamma 29925f2f96 Generate xyz file from kpoint file 300718452c Changed bisection tolerance from 1.0 to 0.001 301-------------------------------------------------------------------------------- 302rel1_63_9 303-------------------------------------------------------------------------------- 3045310d81 Fix discontinuity in torsion constraint function 305ee9efe5 Fix test for negative scf_tol value 306f47d0c1 New implementation of kpgen 30786be185 Added kpoint move functionality 308dd1a11b Merge branch 'upf2' into develop 30945aea94 Added NLCC functionality 310ac8d7e6 Implemented NL potential in UPF 2 format 311ff8a484 Implemented UPF 2 format for SL potentials 312feb88ae Fix end of line in rseed help msg. 3136309a0a Removed enthalpy_int.plt 314e19b136 Add util/enthalpy.plt script 3158d051ce Added -range option to cell.plt and econste_cmp.plt 316f030ffc Fix file headers 317-------------------------------------------------------------------------------- 318rel1_63_8 319-------------------------------------------------------------------------------- 3204ee30f9 Fix discontinuity in torsion constraint function 321d2a3c52 Removed $Id strings 322-------------------------------------------------------------------------------- 323rel1_63_7 324-------------------------------------------------------------------------------- 325r1920: Modified UserInterface.C to use C library, fsync() 326to ensure synchronization. 327r1912-1919: improvements in util/qbdriver files, twin.C 328Added test programs in util/qbdriver 329Added runtwin.sh script for interactive test of client-server mode 330-------------------------------------------------------------------------------- 331rel1_63_6 332-------------------------------------------------------------------------------- 333r1907: add sync in UserInterface.C 334-------------------------------------------------------------------------------- 335rel1_63_5 336-------------------------------------------------------------------------------- 337r1886: fix coeff of divergence correction for nspin=2 and multiple k-points 338r1881,r1883: qbox_eig.py, xyz2qbox.py scripts 339r1871: fix etotal_int calc with multiple k-points 340r1869: removed blanks from timing name attribute 341Enabled iter_cmd with modifications of positions or velocities 342-------------------------------------------------------------------------------- 343rel1_63_4 344-------------------------------------------------------------------------------- 345r1844,r1845,r1846: add qbox_maxforce.py script 346r1835-r1843: B3LYP and BLYP spin-polarized implementation, SVWN updated 347r1829-r1832: qbox_angle.py and qbox_torsion.py scripts 348-------------------------------------------------------------------------------- 349rel1_63_3 350-------------------------------------------------------------------------------- 351r1825: TorsionCmd.h: revert to previous sign convention for compatibility 352with VMD and consistency with torsion constraint. 353-------------------------------------------------------------------------------- 354r1823: Revert to use of wf (rather than dwf) in preconditioner update 355-------------------------------------------------------------------------------- 356r1818: Added set_velocity command. 357-------------------------------------------------------------------------------- 358r1809-r1813: modified Context to accept MPI communicators as ctor argument. 359Removed pimpl idiom from implementation of Context. 360-------------------------------------------------------------------------------- 361r1801 moved sync_positions and sync_velocities to AtomSet set_positions and 362set_velocities functions 363-------------------------------------------------------------------------------- 364rel1_63_2 365-------------------------------------------------------------------------------- 366r1797 BOSampleStepper.C removed timer used in onpe0 section 367r1796 JDWavefunctionStepper.C fixed timer name 368-------------------------------------------------------------------------------- 369r1792 rel1_63_1 370-------------------------------------------------------------------------------- 371r1792 CGIonicStepper.C: reset velocities at end of CG step 372r1791 BOSampleStepper.C: modified etotal_int convergence criterion 373r1788 added timers in JDWavefunctionStepper.C 374-------------------------------------------------------------------------------- 375r1786 rel1_63_0 376-------------------------------------------------------------------------------- 377r1779-r1782 Implementation of Harris-Foulkes functional for calc of etotal_int 378-------------------------------------------------------------------------------- 379rel1_62_7 r1776 380-------------------------------------------------------------------------------- 381r1774: Fixed output of partial_charge command. 382Increased max iter in ConstraintSet (note: two changes in r1774. accidental) 383-------------------------------------------------------------------------------- 384r1771 rel1_62_6 385Implementation of partial_charge command 386Implementation of iter_cmd and iter_cmd_period variables 387JDWavefunctionStepper.C Fix: use dwf for preconditioning. 388-------------------------------------------------------------------------------- 389rel1_62_5 r1760 390r1758 GGIonicStepper.C 391Reduced alpha_max to 5.0 392Rescale alpha when displacement exceeds limit 393r1757 testFourierTransform.C 394Added MPI barriers for more accurate timing. 395Added print of grid sizes. 396r1756 AtomSet.C fix reset_vcm in randomize_velocities 397r1754 r1755 ElectricEnthalpy.C remove omp pragmas to avoid race condition 398-------------------------------------------------------------------------------- 399rel1_62_4 r1748 400r1748 qb.C: Change URL in header to qboxcode.org 401r1742-r1744 ExchangeOperator.C Removed implementation of quadratic correction 402to exchange. Correction does not improve accuracy when far from asymptotic 403convergence of kpoint sampling. 404-------------------------------------------------------------------------------- 405rel1_62_3 r1735 406r1733 Species.C: Fix bug in implementation of nonlocal potential f-projector 407r1735 NonLocalPotential.C: Fix bug in nonlocal potential stress contribution 408from f projectors 409-------------------------------------------------------------------------------- 410rel1_62_2 r1725 411ElectricEnthalpy.[Ch] Changed "polarization" to "dipole" to reflect the 412correct physical definition of output. Adjusted output format. 413-------------------------------------------------------------------------------- 414rel1_62_1 r1722 415Species.C use 12 digit precision 416Fix uninitialized variables 417-------------------------------------------------------------------------------- 418rel1_62_0 419Implementation of electric field 420-------------------------------------------------------------------------------- 421rel1_61_0 422Implementation of ONCV potentials 423/xml/species.xsd 424/xml/sample.xsd 425/src/NonLocalPotential.C 426/src/SpeciesHandler.C 427/src/NonLocalPotential.h 428/src/EnergyFunctional.C 429/src/SpeciesHandler.h 430/src/EnergyFunctional.h 431/src/Species.C 432/src/Species.h 433/src/ChargeDensity.C 434/src/ChargeDensity.h 435-------------------------------------------------------------------------------- 436rel1_60_9 437r1656 ExchangeOperator.C: changed default bisection tolerance to 1.0 438-------------------------------------------------------------------------------- 439rel1_60_8 440r1655 ExchangeOperator print exch time only if DEBUG defined 441r1652 Bisection.C print nsweep only if TIMING defined 442r1651 ExchangeOperator.C added omp directives in jade.C 443r1649 ExchangeOperator.C, Bisection.C added BISECTION_NSWEEP debug option 444-------------------------------------------------------------------------------- 445rel1_60_7 446-------------------------------------------------------------------------------- 447r1645: Bisection.C: modified pair_fraction to reduce calls to overlap. 448r1644: optimization of Bisection::overlap speeds up exchange on large problems 449r1644: added detailed timers to ExchangeOperator.C 450-------------------------------------------------------------------------------- 451rel1_60_6 452-------------------------------------------------------------------------------- 453r1605: added -vlocal option to the plot command. Plot Vloc+VHart+Vxc. 454r1570: cleanup testXMLGFPreprocessor.C 455r1569: cleanup tab ExchangeOperator.C 456-------------------------------------------------------------------------------- 457rel1_60_5 458-------------------------------------------------------------------------------- 459r1564: modify FFTW3 use of threads 460Added -pbc option to distance, angle and torsion functions 461PlotCmd.C: fixed -wfs option. 462r1552 CPSampleStepper.C: added stepper kinetic energy to econst 463Added reset_vcm at end of randomize_v 464-------------------------------------------------------------------------------- 465rel1_60_4 r1550 466-------------------------------------------------------------------------------- 467added FFTW3 option and restored ESSL_FFT functionality 468r1506: Fixed typo in AlphaPBE0.h 469-------------------------------------------------------------------------------- 470rel1_60_2 471CGCellStepper.C: revert to 1.60.0 use of matrix transformations 472Use reduced alpha_init (0.002) 473-------------------------------------------------------------------------------- 474rel1_60_1 475CGCellStepper.C: fixed incorrect use of amat_inv() in calc of gradient. 476Note: use was correct. later restored. 477CGCellStepper.C: reduced initial alpha value to 0.01, max to 0.5. 478-------------------------------------------------------------------------------- 479qbox_msd.py: print msd value only, not step index 480-------------------------------------------------------------------------------- 481Context.C: fixed deprecated cast to const char* 482-------------------------------------------------------------------------------- 483rel1_60_0 484Added alpha_PBE0 variable 485-------------------------------------------------------------------------------- 486rel1_59_0 487Added scf_tol variable 488-------------------------------------------------------------------------------- 489rel1_58_0 490Added utility scripts 491-------------------------------------------------------------------------------- 492rel1_57_15 493-------------------------------------------------------------------------------- 494r1409: ExchangeOperator.C: fixed coeffs of divergent terms for empty states 495r1407: XMLGFPreprocessor.C: add unistd header for close function 496-------------------------------------------------------------------------------- 497rel1_57_14 498-------------------------------------------------------------------------------- 499r1396: util/kpgen/kpgen.C: fixed odd numbered case and added test. 500r1395: Species.C: fix for case of missing radial functions 501r1394: BOSampleStepper.C: include weight info in eigenvalues 502r1390,r1391: Fix exchange calculation with multiple k-points 503r1388: Species.[Ch]: ensure that print outputs the same data as was read 504from the pseudopotential file. 505r1363: XMLGFPreprocessor.C: implemented sample loading using http connection. 506-------------------------------------------------------------------------------- 507rel1_57_13 508-------------------------------------------------------------------------------- 509r1358: ExchangeOperator.C: Fixed exchange contributions to stress. 510KPGridConnectivity.[Ch]: Added cell_moved() member to update cell-dependent 511quantities during cell optimization. 512-------------------------------------------------------------------------------- 513rel1_57_12 514-------------------------------------------------------------------------------- 515r1354: Added cell_moved() member in ExchangeOperator to resize vbasis 516-------------------------------------------------------------------------------- 517rel1_57_11 518-------------------------------------------------------------------------------- 519r1349: modified Wavefunction::operator= to allow assignment between wfs with 520different cell sizes. Needed for variable cell calculations. 521-------------------------------------------------------------------------------- 522rel1_57_10 523-------------------------------------------------------------------------------- 524r1344: removed dependency on Context in multiple classes 525 SlaterDet uses MPI column communicator instead of Context 526-------------------------------------------------------------------------------- 527rel1_57_9 528-------------------------------------------------------------------------------- 529r1326: WavefunctionHandler.[Ch]: fixed current_gfdata_pos variable 530-------------------------------------------------------------------------------- 531rel1_57_8 532-------------------------------------------------------------------------------- 533r1317: undo r1312: keep positive constraint on delta_spin 534r1316: Fixed loading of gfdata into slater_det with non-zero delta_spin 535r1315: Restored correct preconditioning when using stress confinement 536r1314: EnergyFunctional.[Ch]: removed dependency on ChargeDensity.h from header 537r1313: XCOperator.[Ch]: removed dependency on ChargeDensity.h from header 538r1312: allow negative delta_spin 539r1311: MDWavefunctionStepper.C: removed unused vars 540r1310: NonLocalPotential.C: removed unused vars 541r1309: removed deprecated conversions from string const to char* in headers 542-------------------------------------------------------------------------------- 543rel1_57_7 544-------------------------------------------------------------------------------- 545r1304: BOSampleStepper.C: removed printing of delta_ehart and delta_eigsum. 546Renamed variable nscf_converged. 547r1303: Wavefunction.C: fixed add_kpoint to avoid modifying occ of existing kpts 548This caused incorrect occupation numbers when loading wf with multiple kpts. 549r1301: BOSampleStepper.C: print RC1 override msg on pe0 only 550-------------------------------------------------------------------------------- 551rel1_57_6 552-------------------------------------------------------------------------------- 553r1296: SampleHandler.C: set wfv to use same ctxt as s_.wf 554-------------------------------------------------------------------------------- 555rel1_57_5 556-------------------------------------------------------------------------------- 557r1292: fixed mlwf calc for spin-polarized case 558r1291: ComputeMLWFCmd.C: added spin implementation 559r1290: Anderson charge mixing on both spins simultaneously 560r1288: BOSampleStepper: moved enthalpy definition out of onpe0 test 561-------------------------------------------------------------------------------- 562r1286: rel1_57_4 563r1283: PSDAWavefunctionStepper.C: fixed extrapolate flag 564r1282: rewrite of kpgen.C 565-------------------------------------------------------------------------------- 566rel1_57_3 567r1275: fix sample read/write for nspin=2. Save cmd writes full species info. 568-------------------------------------------------------------------------------- 569rel1_57_2 570r1272: BOSampleStepper.C: fixed nite loop test for nite==0 case 571-------------------------------------------------------------------------------- 572rel1_57_1 573r1265: BOSampleStepper.C: changed behavior for nite=1, use eigenvalue_sum for 574convergence test of nite loop, changed threshold value to 0.01*delta_ehart 575r1264: RunCmd.C: changed default nite value to 0 576r1268: ExchangeOperator.C: Added EXCHANGE_NOQUAD debug string. Cleanup. 577-------------------------------------------------------------------------------- 578rel1_57_0 r1262 579Cleanup of XCOperator and ExchangeOperator to include stress at gamma point 580-------------------------------------------------------------------------------- 581rel1_56_2 r1259 582r1257: EnergyFunctional: fix nspin=2 forces, stress. 583BOSampleStepper: fixed bug in mixing with charge_mix_ndim=0 584r1212: Adaptive preconditioner. 585r1211: tolerance for non-self consistent iterations 586-------------------------------------------------------------------------------- 587rel1_55_6 588Redesign of Bisection and ExchangeOperator separates bisection transformation 589and load-balancing permutation. 590Fixed KPGridConnectivity 591Added unit_cell info in BOSampleStepper output. 592SlatedDet: added occ_ and eig_ to operator= 593-------------------------------------------------------------------------------- 594rel1_55_5 595Added polar decomposition in Matrix class. 596Use polar in ortho_align, align, lowdin ortho. 597-------------------------------------------------------------------------------- 598rel1_55_4 599Added OpenMP pragmas 600-------------------------------------------------------------------------------- 601rel1_55_0 602merge of hfb functionality 603-------------------------------------------------------------------------------- 604rel1_54_6 605Added USE_DFFTW macro for platforms providing the dfftw.h header only 606-------------------------------------------------------------------------------- 607rel1_54_5 608Added enthalpy in BOSampleStepper for use in cell optimization. 609BasisMapping.C: fix bug of serial version: MLWF calc of dipole wrong. 610Added randomize_r command 611removed printing on stdout in rseed command 612Changed number of iterations in SlaterDet::{ortho_align,align,riccati} 613-------------------------------------------------------------------------------- 614rel1_54_4 615Added randomize_v command to set velocities to a Maxwell-Boltzmann distribution 616Modified CGIonicStepper: project forces on constraints 617Modified BMDIonicStepper: rescale velocity rather than stop 618Other steppers: added synchronization calls to ensure consistency of positions 619and unit cell values across tasks. 620-------------------------------------------------------------------------------- 621rel1_54_3 622Rewrite of CellStepper, SDCellStepper, CGCellStepper classes. 623CGOptimizer and LineMinimizer restructured. 624-------------------------------------------------------------------------------- 625rel1_54_2 626fix CG algorithm in CGCellStepper.C 627-------------------------------------------------------------------------------- 628rel1_54_1 629CGCellStepper.C: fixed CG algorithm. 630-------------------------------------------------------------------------------- 631rel1_54_1 632Basis.C: fixed bug 26: cannot build basis for kpoint close to FCC BZ boundary 633fixed in r924 634-------------------------------------------------------------------------------- 635rel1_54_0 636LineMinimizer.h: limited the number of iterations in bracketing mode to 4 637UnitCell.C: accept unit cells with orientation leading to negative volume and 638define the volume as fabs(a0*(a1^a2)) 639CellStepper.C, CGCellStepper.C: implemented CG algorithm to optimize both 640cell parameters and atomic positions. Moved implementation of common CellStepper 641member functions to CellStepper.C. 642-------------------------------------------------------------------------------- 643rel1_53_0 644RseedCmd.h: rseed command to initialize random number generator 645BOSampleStepper.C: reintroduced wf extrapolation when nite>1 (leads to 646much improved energy conservation in liquid Si) 647Matrix.[Ch]: added IntegerMatrix, permutations 648ComputeMLWF.C: test if not at Gamma point 649StructureFactor.C, SampleWriter.C: added <cstring> for memset 650XCPotential.h: moved header dependencies to XCPotential.C 651-------------------------------------------------------------------------------- 652rel1_52_4 653Introduced a USE_UUID macro in qb.C, uuid_str.[Ch] (some platforms have no 654uuid library) 655Fix bug 23: Wavefunction::set_nempty: added update occ when nempty=0. 656-------------------------------------------------------------------------------- 657rel1_52_3 658Fixed StructuredDocumentHandler.[Ch] and WavefunctionHandler.C for compilation 659with Xerces-C 3.x.x 660SampleReader.C: removed debug print 661-------------------------------------------------------------------------------- 662rel1_52_2 663RunCmd.h: updated help msg 664rel1_52_2 665RunCmd.h: updated help msg 666LoadCmd.C: fix bug in reading a sample with atomset only. (unit_cell not set). 667Add <cmd> </cmd> tags when echoing commands executed by qbox. 668Add <eigenset> tag around <eigenvalues> for each kpoint and spin. 669BOSampleStepper.C: Disabled wf extrapolation if nite>1 (unstable) 670-------------------------------------------------------------------------------- 671rel1_52_1 672SlaterDet::gram use beta=0 in herk, for consistency (complex case). 673(does not affect results since Matrix ctor clears data). 674SlaterDet: removed constraint on Im(c(G=0)) for complex case. 675Note: this fixes a bug introduced in 1.51.1 which incorrectly set Im=0 for k!=0. 676This fixes bug 2010-04-04T01:14:19Z 677Matrix.C: parameters alpha and beta of pzherk and zherk must be double* 678Moved files uuid.[Ch] to uuid_str.[Ch] to avoid name conflicts with uuid/uuid.h 679-------------------------------------------------------------------------------- 680rel1_52_0 681Added the rescale_v command (RescaleVCmd.h) to rescale all velocities. 682Implemented the Bussi-Donadio-Parrinello thermostat. 683Added files sampling.C sampling.h 684Modified IonicStepper.h to add ekin_stepper member 685BOSampleStepper.C: <econst> includes stepper kinetic energy when using BDP. 686FourierTransform.C: included some OpenMP pragmas 687Known bug: 2010-04-04T01:14:19Z 688-------------------------------------------------------------------------------- 689rel1_51_1 690Attempt to fix bug 2010-04-04T01:14:19Z 691Small imaginary part in c(G=0), initialization in SlaterDet::resize. 692Incorrect fix of 2010-04-04T01:14:19Z causes incorrect results in k!=0 wfs 693Later corrected in 1.52.1 694Known bug: 2010-04-04T01:14:19Z 695-------------------------------------------------------------------------------- 696rel1_51_0 697Added external forces: atomic, pair or global. 698Added capability to load samples multiple times 699Added capability to redefine species 700Added uuid identifier in output 701SlaterDet.C, Wavefunction.C modified clear, reset, init members 702Species.C: modified Species::info to print XML elements 703-------------------------------------------------------------------------------- 704rel1_50_4 705ConstraintSet.C: Fixed uninitialized ndofs_ in ctor. 706-------------------------------------------------------------------------------- 707rel1_50_3 708SampleWriter.C: Fixed size calculation for serial version 709-------------------------------------------------------------------------------- 710rel1_50_2 711Updated several files to enable serial version (without MPI or ScaLAPACK) 712Added include<cstring> in various files to get correct memcpy declaration. 713(caused trouble with gcc 4.4.1 on Ubuntu). 714-------------------------------------------------------------------------------- 715rel1_50_1 716EnergyFunctional.C:732: use sigma only if compute_stress==true 717-------------------------------------------------------------------------------- 718rel1_50_0 719BOSampleStepper.C, RunCmd.C: added [-atomic_density] option to the run cmd. 720AndersonMixer.C: iterative LS problem with adaptive Tikhonov parameter. 721Matrix.[Ch]: added getsub for arbitrary block destination 722JDWavefunctionStepper.[Ch]: Jacobi-Davidson WavefunctionStepper 723-------------------------------------------------------------------------------- 724rel1_49_1 725EnergyFunctional.C: increased criterion for determination of neighboring 726cells in esr calculation. 727-------------------------------------------------------------------------------- 728rel1_49_0 729BOSampleStepper.C: introduced separate row-weighting of Anderson LS problem 730with fixed cutoff rc1=3.0 a.u. 731AndersonMixer.C: modified regularization 732PlotCmd.C: white space 733SlaterDet.C: white space 734-------------------------------------------------------------------------------- 735rel1_48_1 736SlaterDet.C: fix bug in SlaterDet::write for -text mode 737BOSampleStepper.C: modified Kerker weights. Added mixer.reset() call. 738AtomCmd.h: fixed text of help message. 739AndersonMixer.C: added Tikhonov regularization. Bcast results from task 0. 740PlotCmd.C: cleanup include. 741-------------------------------------------------------------------------------- 742rel1_48_0 743PlotCmd.[Ch]: added plot command. Generates cube format or xyz 744{LDA,PBE,BLYP}Functional: modified API for compat. with future implem. of PBE0 745SlaterDet.C: added timing ifdefs in gram ortho. 746-------------------------------------------------------------------------------- 747rel1_47_1 748Added a virtual member function in Constraint that returns the number of DOFs to 749be subtracted due to that constraint. That number is 1 for all constraints, 750except for PositionConstraint, for which it is 3. This function is used 751by ConstraintSet to keep track of the total number of DOFs to be removed 752when computing the system temperature. The ConstraintSet class maintains the 753total number of DOFs and modifies it only when constraints are added or deleted. 754MDIonicStepper::ekin() modified to compute kinetic energy correctly in the 755presence of constraints. 756Emass.h: allow reset to zero 757BOSampleStepper.C: fix print format of etotal_int 758-------------------------------------------------------------------------------- 759rel1_47_0 760Add BMDIonicStepper class. 761ConstraintCmd.h ConstraintSet.C: Added PositionConstraint class to lock atoms. 762Species.C: modified calculation of ndft to ensure rmax>40 a.u. 763This corrects the problem that ndft would depend on the number of mesh points 764given in the species file. 765WavefunctionHandler.C: fix XMLString::release() call for Xerces 3.0.1 766-------------------------------------------------------------------------------- 767rel1_46_0 768AndersonMixer.C, BOSampleStepper.C: Implemented multidimensional Anderson 769acceleration of charge mixing. 770Added charge_mix_ndim to specify dimension of Anderson acceleration of 771charge mixing. charge_mix_ndim=0 results in simple mixing. 772-------------------------------------------------------------------------------- 773rel1_45_3 774UserInterface.C: modified server mode to allow multiple commands in input file. 775Use file "<inputfile>.lock" to synchronize with client. 776Output file in server mode is now a valid XML fragment, with fpmd namespace. 777-------------------------------------------------------------------------------- 778rel1_45_2 779qb.C, UserInterface.C: implemented -server option, cleaned up destructors. 780AtomSet.C: implemented destructor (fix leaks) 781qb.C: added delete s at end (fix leaks) 782ConstraintSet: implemented destructor (fix leaks) 783Constraint.h and derived classes: implemented virtual destructors. 784StrainCmd.h: fix error in usage message. 785XMLGFPreprocessor.C: new macro PARALLEL_FS. Limit the number of simultaneous 786fread calls if PARALLEL_FS is not defined. 787-------------------------------------------------------------------------------- 788rel1_45_1: fixed SDCellStepper.C: make local copy of sigma to avoid 789PGI compiler error on franklin: "must be lvalue or...". 790-------------------------------------------------------------------------------- 791rel1_45_0 792Added strain command for deformation of the cell. 793UnitCell.C: removed member compute_deda (moved calc to SDCellStepper). Added 794symmetric matrix - vector product function. 795SDCellStepper.C fixed calculation of cell correction from stress tensor. 796BOSampleStepper.C: reintroduced wf extrapolation when nempty>0 for MD. 797RunCmd.h: fixed help msg. 798-------------------------------------------------------------------------------- 799rel1_44_5 800BOSampleStepper.C: Fix uninitialized mlwft ptr. Fixed delete statements. 801-------------------------------------------------------------------------------- 802rel1_44_4 803BOSampleStepper.C: reintroduced feature that keeps memory of wfv between run 804commands. Allows for smooth continuation of multiple runs. Added MLWF and 805MLWFC calculation in the niter loop. 806Modified wf extrapolation to include alignment in all cases. Use lowdin 807orthogonalization after extrapolation. 808WfDiag.h: Added MLWF and MLWFC option for the wf_diag variable: computes MLWFs 809or MLWF centers at each ionic step. 810SaveCmd.[Ch], SampleWriter.[Ch]: added -no_wfv option: save without wf velocity. 811-------------------------------------------------------------------------------- 812rel1_44_3 813UserInterface.h, Base64Transcoder.C: added include <cstring> 814removed trailing blanks. 815-------------------------------------------------------------------------------- 816rel1_44_2 817Added copyright headers. 818-------------------------------------------------------------------------------- 819rel1_44_1 820EnergyFunctional.C: fixed calculation of Esr and Esr stress in cases where the 821unit cell is small compared with rcps (initialized in AtomSet.C). 822-------------------------------------------------------------------------------- 823rel1_44_0 824WavefunctionHandler.[Ch], SampleReader.[Ch]: fixed compatibility with older 825versions when reading samples with smaller grid sizes. 826-------------------------------------------------------------------------------- 827rel1_43_5 828BOSampleStepper.C: print energy at end of wf optimization when using niter=0. 829Basis.C: modified grid size to cover all grids for any k-point 830-------------------------------------------------------------------------------- 831rel1_43_4 832Basis.C: np() yields same grid size for all kpoints 833XMLGFPreprocessor.C, SampleReader.C: fixed reading samples with both real and 834complex wavefunctions 835Wavefunction.C: removed unneeded resize in add_kpoint 836-------------------------------------------------------------------------------- 837rel1_43_3 838TorsionCmd.h: fixed sign of torsion angle. 839SlaterDet.[Ch]: fixed const-ness of some members. 840Wavefunction.C: modif allocation in add_kpoint to avoid O(nkp**2) time 841x8664_gcc.mk: link statically to gfortran libs for better portability of exec. 842-------------------------------------------------------------------------------- 843rel1_43_2 844SlaterDet.C: rewrite of the parallel I/O section, using less memory. 845-------------------------------------------------------------------------------- 846rel1_43_1 847BOSampleStepper.C, CPSampleStepper.C: fixed initialization of ekin_ion,temp_ion. 848SampleWriter.C, Wavefunction.C, SlaterDet.C: implementation of parallel I/O 849using SharedFilePtr class. 850-------------------------------------------------------------------------------- 851rel1_43_0 852SampleWriter.C, Wavefunction.C, SlaterDet.C: implemented MPI-IO parallel I/O 853using individual file pointers, no shared pointers. 854SlaterDet.C: fixed bug: base64 transcoding done twice on big endian platforms 855when writing in parallel. 856-------------------------------------------------------------------------------- 857rel1_42_0 858SlaterDet.C, Wavefunction.C: implemented dynamic resize of ecut or nempty 859without loss of data. Changes of cell also preserve wavefunctions. 860-------------------------------------------------------------------------------- 861rel1_41_0 862BOSampleStepper.C, CPSampleStepper.C: remove redundant printing of unitcell 863LineMinimizer.h: new LineMinimizer class used in optimization algorithms 864SDIonicStepper.C: back to simple dt^2/m steepest descent (back to rel1_39_0) 865SDAIonicStepper.C: steepest descent with line minimization. 866CGIonicStepper.C: conjugate gradient with line minimization. 867-------------------------------------------------------------------------------- 868rel1_40_0 869AtomsDyn.h,CGIonicStepper.[Ch]: added CG ionic stepper. 870BOSampleStepper.C: enabled niter=0 in run command. 871MDIonicStepper.C: removed trailing blanks. 872SDAIonicStepper: reduced theta_max, test for energy increase. 873SDIonicStepper: implemented line search, removed dt, mass dependency. 874Known problems: SD, SDA, CG algorithms are not working properly 875-------------------------------------------------------------------------------- 876rel1_39_0 877SampleReader.C, SpeciesReader.C: set validation scheme to Val_Auto to allow 878successful parsing in the absence of a local copy of sample.xsd or species.xsd. 879Added fold_in_ws command: fold all atoms in the WS cell. 880MDIonicStepper.C: removed collision debug info printout. 881MDIonicStepper.C: make calc of collisions on task 0, then sync in LOWE and 882ANDERSEN thermostats. Ensures that all velocities are consistent on all nodes. 883PSDAWavefunctionStepper.C: removed psda_enhancement_factor which caused 884oscillations in some PSDA optimizations. 885BOSampleStepper.C: use wf velocity only inside BOSampleStepper. wfv is not 886saved at end of steps. wfv is not used at first step for wf extrapolation. 887-------------------------------------------------------------------------------- 888rel1_38_0 889MDIonicStepper.C: Modified minimum image convention in LOWE thermostat to 890conserve angular momentum in clusters. Fixed LOWE and ANDERSEN thermostats to 891have correct collision rate. 892UserInterface.C: fixed endl in output to avoid double endl. 893Basis.C: remove pimpl implementation. 894ChargeDensity.C: use factorizable sizes for density grid. 895EnergyFunctional.C: corrected upper bound of ia1 in esr calculation. 896EnergyFunctional.C: implemented multiple cell in esr calculation. 897AtomSet.C: changed rcps value to 1.5. Changed interface to access cell. 898-------------------------------------------------------------------------------- 899rel1_37_0 900Cleaned up XML formatting of output. Removed comments. Use <timing> tag. 901-------------------------------------------------------------------------------- 902rel1_36_0 903Added unit cell to the sample schema. Modified AtomSet to contain a unit cell. 904Included unit cell in saved samples. 905-------------------------------------------------------------------------------- 906rel1_35_0 907Implemented parallel I/O for writing sample files. Uses MPI-IO. 908This does not work on BG/L with PVFS since PVFS does not support the 909collective calls MPI_File_write_ordered and MPI_File_write_shared. 910Option -serial added to the save command to use the old serial mode. 911-------------------------------------------------------------------------------- 912rel1_34_4 913Base64Transcoder.C: modif print functions to use larger string buffer for I/O. 914SampleReader.C, SampleWriter.C: added timer. 915Corrected lda arg in various BLACS calls. Did cause errors on 916AIX/pessl. (Not a bug when using blacs lib compiled from netlib distribution). 917-------------------------------------------------------------------------------- 918rel1_34_3 919Wavefunction.C: fixed incorrect XML comments 920Basis.C: fixed warning message printed by all tasks when k out of BZ. 921AtomSet.C, SpeciesCmd.C: modif to always print species info when a species is 922defined, i.e. also when loading from a sample file. 923-------------------------------------------------------------------------------- 924rel1_34_2 925WavefunctionHandler.C: fixed for compilation on BG/L 926SampleReader.C: fixed for case where loading multiple k-points with wfv 927NonLocalPotential.C: fixed calc of forces with multiple k-points 928EnergyFunctional.C: fixed forces and stress with multiple k-points 929MDWavefunctionStepper.C: fixed CP dynamics with k-points 930MLWFTransform.C: changed syntax of blas call (zcopy instead of zcopy_) 931NonLocalPotential.C: fix forces with k-point, fix bug introduced in 1.34.0. 932SampleReader.C: multiple k-point and wf velocity functionality. 933SlaterDet.C: fixed print() and write(). 934Wavefunction.C: changed context allocation. Fix update_occ, entropy. 935PSDWavefunctionStepper.C: fixed complex case (k!=0) 936-------------------------------------------------------------------------------- 937rel1_34_1 938Several bug fixes for kpoint functionality. 939CPSampleStepper.C: create wfv if absent. 940ChargeDensity.C: increased grid size to avoid aliasing with k!=0 941-------------------------------------------------------------------------------- 942rel1_34_0 (not for distribution) 943Added kpoints. This release still has bugs. Not for distribution. 944-------------------------------------------------------------------------------- 945rel1_33_5 946removed trailing blanks in *.[Ch] *.mk Makefile 947-------------------------------------------------------------------------------- 948rel1_33_4 949ComputeMLWFCmd.C: fixed memory leak. 950blas.h: added zgemm, dnrm2. 951BOSampleStepper.C: disabled Anderson acceleration of charge mixing by default, 952can be enabled with debug="AND_CHMIX" 953AtomSet.C: removed unused variables. 954-------------------------------------------------------------------------------- 955rel1_33_3 956BOSampleStepper.C: extrapolation: restored alignment before extrapolation 957ChargeMixCoeff.h: changed default value back to 0.5 958Matrix.h: added iglobal and jglobal functions 959SDAIonicStepper.h: modified default theta_max and theta_nc values 960Species.[Ch]: added phi function 961-------------------------------------------------------------------------------- 962rel1_33_2 963Implemented calculation of total electronic and ionic dipole in MLWF. 964-------------------------------------------------------------------------------- 965rel1_33_1 966MLWFTransform.C: fixed bug in MLWF calc when cell sizes in x and y direction 967differ. 968-------------------------------------------------------------------------------- 969rel1_33_0 candidate 970Implemented compute_mlwf command 971Changed spline.C and sinft.C with new implementations. 972Modified Species.C to use new spline and sinft interface. 973-------------------------------------------------------------------------------- 974rel1_32_0 975Dt.h, MDIonicStepper.C: modified to allow for negative time step. 976Removed Sample.h dependency in WavefunctionStepper and derived classes. 977Files affected are 978BOSampleStepper.C 979SDWavefunctionStepper.C 980PSDWavefunctionStepper.C 981PSDAWavefunctionStepper.C 982EnergyFunctional.C 983CPSampleStepper.C 984MDWavefunctionStepper.C 985WavefunctionStepper.h 986SDWavefunctionStepper.h 987PSDWavefunctionStepper.h 988PSDAWavefunctionStepper.h 989MDWavefunctionStepper.h 990Removed facs, sigmas from Control.h. Moved vars to ConfinementPotential. 991Use hardwired values 2.0 and 0.5 in 992BOSampleStepper.C, CPSampleStepper.C 993Removed namespace declarations in header files. 994-------------------------------------------------------------------------------- 995rel1_31_0 996Added class SampleWriter to separate write function from SaveCmd class. 997-------------------------------------------------------------------------------- 998rel1_30_1 999Wavefunction.C: typos 1000XMLGFPreprocessor.C: fixed bug in boundary adjustment. 1001jacobi.C: implement special case of fewer blocks than process columns. 1002BOSampleStepper.C: Moved Kerker preconditioning of charge density update 1003before Anderson acceleration. 1004ChargeMixCoeff.h: Changed default value to 1.0 1005ExtStress.h: initialize with zero in ctor 1006MDIonicStepper.C, Thermostat.h: corrected misspelled ANDERSEN 1007-------------------------------------------------------------------------------- 1008rel1_30_0 1009D3vector.h: modified ctor to avoid accidental use with only one argument. 1010AtomSet.[Ch]: added functions to get/reset center of mass velocity. 1011StatusCmd.h: status cmd prints center of mass velocity. 1012MDIonicStepper.[Ch]: new thermostats ANDERSON LOWE. Old thermo renamed 1013to SCALING. 1014-------------------------------------------------------------------------------- 1015rel1_29_1 1016BOSampleStepper.C: use wfmm only when needed. ASP extrapolation implemented. 1017Activate NTC or ASP extrapolation using debug variable. 1018XMLGFPreprocessor.C: added tests to avoid div by zero if no local grid 1019-------------------------------------------------------------------------------- 1020rel1_29_0 1021BOSampleStepper.C: major rewrite: 1022added a wf_stepper->preprocess before the itscf loop to 1023get correct extrapolation when ions move. 1024Modified NTC extrapolation: activated if NTC_EXTRAPOLATION is found in the 1025debug variable. 1026ReadTSC.h: added missing #define statement 1027Debug.h: modified to take arbitrary strings 1028jacobi.C: new jacobi diagonalization. activate with ifdef in Wavefunction.C 1029-------------------------------------------------------------------------------- 1030rel1_28_2 1031AtomSet.h: initialize nel_ in ctor. 1032XMLGFPreprocessor.C: use char buffer to read file. 1033Include timers in SlaterDet::ortho_align with #if TIMING 1034qb.C: remove <> brackets in processor names for XML compatibility (BGL) 1035AtomSetHandler.C, SpeciesHandler.C: removed msg about xlink, replaced with href 1036Species.C, SaveCmd.C: removed xlink namespace declaration in sample 1037SlaterDet.C: added #if TIMING and timers in ortho_align 1038MDIonicStepper.h: corrected comments 1039-------------------------------------------------------------------------------- 1040rel1_28_1e{1234567} experimental versions containing 1041various modifications leading to rel1_28_2 1042-------------------------------------------------------------------------------- 1043rel1_28_0 1044qb.C, SaveCmd.C, qbox_xmlns.C, xmlSpecies.C: change namespace to fpmd 1045-------------------------------------------------------------------------------- 1046rel1_27_1 1047minor updates to make single-node version work. 1048-------------------------------------------------------------------------------- 1049rel1_26_1e 1050redefine constraint command and subcommands, output format. 1051Introduce constraint names. 1052Remove ListConstraintsCmd.h 1053SDAIonicStepper.h: fixed bug: 1054Use of ndofs_ incorrect when ndofs<3*na (when constraints are defined). 1055Use CHOLESKY_REMAP macro value for size of square context in SlaterDet.C (bgl). 1056-------------------------------------------------------------------------------- 1057rel1_26_0 1058Species.C: changed spline BC for local potential. 1059SlaterDet.C: randomize: return if basis size is zero. 1060Implementation of a new logic in all IonicStepper classes. 1061Implementation of the move command. 1062Implementation of constraints. 1063Implementation of distance, angle and torsion measurement commands. 1064-------------------------------------------------------------------------------- 1065rel1_25_0 1066implementation of netcharge. 1067SpeciesCmd.C: print species information when adding species. 1068bug fix: MDWavefunctionStepper.C: correct dt2bye when ecuts>0. 1069-------------------------------------------------------------------------------- 1070rel1_24_3 1071FourierTransform.C: removed unused code. 1072NonLocalPotential.C: cleanup calculation of G=0 in update_twnl. 1073BOSampleStepper.C: added APC counters 1074qb.C: added APC counters 1075-------------------------------------------------------------------------------- 1076rel1_24_2 1077NonLocalPotential.C: Fixed bug in G=0 component of l=0 non-local projector in 1078semilocal form (merged bugfix from rel1_23_4) 1079-------------------------------------------------------------------------------- 1080rel1_24_1 1081SlaterDet.C: fixed bug introduced in 1.24.0: Cholesky decomp was skipped if 1082CHOLESKY_REMAP is not defined. 1083-------------------------------------------------------------------------------- 1084rel1_24_0 1085UnitCell.C: added space in output of lattice vectors. 1086Wavefunction.C: added syevx option for diagonalization (not active) 1087BOSampleStepper.C: added diag timer. 1088SlaterDet.C: Added CHOLESKY_REMAP macro to enable remapping to a square context 1089in SlaterDet::gram. 1090SlaterDet.C: rewritten communication in SlaterDet::print and SlaterDet::write 1091to avoid a deadlock on BGL. 1092Matrix.C: Added syevx and syevd functions 1093Note: modif of SlaterDet.C introduced a bug: see 1.24.1 1094-------------------------------------------------------------------------------- 1095rel1_23_4 (on branch rel1_23_3_bf) 1096Fixed bug in NonLocalPotential.C: G=0 component of l=0 semilocal projector. 1097-------------------------------------------------------------------------------- 1098rel1_23_3 1099NonLocalPotential.C: fixed bug: double fnl_size should be int fnl_size. 1100-------------------------------------------------------------------------------- 1101rel1_23_2 1102FourierTransform.C: Fixed bug #39: When unit cells are non-orthogonal, the 1103transforms along x are not done correctly. Bounds including all non-zero 1104transforms are redefined. 1105Basis.C: Simplified basis set construction. 1106NonLocalPotential.C: removed variables used in old implementation (pre 1.23.0) 1107-------------------------------------------------------------------------------- 1108rel1_23_1 1109PSDAWavefunctionStepper.C: Reset default theta for negative curvature to 0.0. 1110AndersonMixer: added settable default theta value for negative curvature. 1111-------------------------------------------------------------------------------- 1112rel1_23_0 1113Added SDA algorithm for atoms dynamics: Anderson acceleration of SD. 1114New implementation of NonLocalPotential::energy, recomputing anl array at 1115every call. 1116Modified IonicStepper to make ekin and temp virtual. Implemented only in 1117MDIonicStepper. 1118-------------------------------------------------------------------------------- 1119rel1_22_0 1120BOSampleStepper.C: new implementation of Kerker-Anderson charge mixing. 1121new variables charge_mix_rcut and charge_mix_coeff introduced: determine 1122the Kerker cutoff and mixing coefficient parameters. 1123-------------------------------------------------------------------------------- 1124rel1_21_0 1125Context.[Ch]: redesigned Context to be column-major in all cases. 1126ChargeDensity.C: vcontext definition modified. 1127SlaterDet.C: new my_col_ctxt member: store single column Context. 1128Use modified Context allocation for my_col_ctxt to ensure that calls to 1129MPI_Comm_create are done with same arguments on all tasks. 1130Wavefunction.C: use simpler Context constructor (default is col-major). 1131UserInterface.C: removed printing of newline when echo==true. 1132Cleaned up timers in wf steppers. 1133-------------------------------------------------------------------------------- 1134rel1_20_1 1135NonLocalPotential.C: fixed index calculation (bug 38). 1136-------------------------------------------------------------------------------- 1137rel1_20_0 1138Wavefunction.C: Changed allocation of sd context to column-major order to get 1139contiguous nodes in context columns. 1140Wavefunction.C: Changed maxiter to 500 and dmu to 2 eV in update_occ. 1141Nrowmax.h: add update_occ when changing nrowmax. 1142Matrix.C:219 and similar: removed assert. Some valptr may be 0 on some blocks. 1143SlaterDet.C: print sdcontext dimensions in SlaterDet::info. 1144ChargeDensity.C: added (void*) cast to delete statements (for xlC 7.0). 1145Known bug #38: incorrect forces with nrowmax=16 with 512 tasks. Occurs when 1146nrowmax <= (ntasks/nrowmax)/2. 1147-------------------------------------------------------------------------------- 1148rel1_19_0 2004-10-04 1149BOSampleStepper.C: Implemented simple charge mixing for SCF iterations. 1150Disabled potential mixing. 1151SpeciesReader.C: fixed parsing of species files. 1152SlaterDet.C in print and write: fixed incorrect allocation of char buffer b: 1153caused segfaults when saving sample. 1154Base64Transcoder.[Ch]: added cstdio style interface for print function. 1155FourierTransform.[Ch]: implemented timers, activated if TIMING is defined. 1156testFourierTransform.C: updated to print timing info in readable form. 1157Known bug: Wavefunction::update_occ iteration on mu does not converge in some 1158systems. 1159-------------------------------------------------------------------------------- 1160rel1_18_0 2004-09-14 (not for distribution) 1161BOSampleStepper modified for scf iterations. Update of charge density and 1162calculation of VHXC is taken out of EnergyFunctional. 1163RunCmd modified. Use run niter nitscf nite. 1164Added FermiTemp variable. 1165Modifications for calculation of occupation numbers if nempty > 0. 1166Added USE_XERCES macro. If not defined, the Xerces library is not used, and 1167only species files can be read. Loading of sample files without the Xerces lib 1168is not implemented yet. 1169SlaterDet.C modified to remove dependency on xercesc. Uses Base64Transcoder. 1170Makefile modified for libqb.a target for BGL. 1171Known bug: SpeciesReader.C incorrect if not USE_XERCES 1172Known bug: Base64 encoding not done in SlaterDet.C 1173-------------------------------------------------------------------------------- 1174rel1_17_1 2004-08-18 1175Fixed bug #27: overrun of variable tmpfion in NonLocalPotential.C is due to 1176incorrect index calculation in situations where na[is] % (ntasks/nrowmax) != 0. 1177This is the case for (CdSe)_110 on 128 tasks, nrowmax=32, since na[is]/4 = 27.5. 1178Note: bug #27 does not affect calculations with a single non-local projector 1179(npr[is]==1). 1180Fixed bug #33 LDAFunctional.C: incorrect use of Perdew-Zunger caused small 1181differences (~1.e-3 a.u.) in total energy when compared to other codes. 1182The adjustment of the constant gamma to make the PZ energy function continuous 1183was incorrect. Instead, the constants C and D must be adjusted to ensure 1184continuity of Ec and its derivative at rs=1. 1185-------------------------------------------------------------------------------- 1186rel1_17_0 2004-08-11 1187Fixed buffer overrun in Context::string_bcast. (note: bug #27 is not fixed). 1188Fixed bug 26: inconsistent qbox xmlns namespace in species file and qbox output. 1189Modified qb.C: use new qbox_xmlns() function to print qbox XML namespace. 1190EnergyFunctional.C:909: change lda argument to 14. This was done automatically 1191by the BLACS lib, but corrected line conforms to the BLACS convention that 1192lda >= m. 1193ChargeDensity.C: changed the <total_electronic_charge> output to an XML comment. 1194SaveCmd.C: include proper xmlns info when saving a sample. 1195blas.h: reorganized using the ADD_ macro, removed all platform-specific ifdefs. 1196XCPotential.C: removed underscores in blas calls. 1197SlaterDet.C: removed underscores in blas calls. 1198FourierTransform.C: use ADD_ for zdscal. 1199NonLocalPotential.C: removed underscores in blas calls, removed old OSF ifdefs. 1200-------------------------------------------------------------------------------- 1201rel1_16_0 1202Implemented l=1,l=2 non-local stress. 1203Modified UnitCell.C to avoid infinite loops due to finite precision. 1204Modified mcr.mk to use library mkl_p4 instead of mkl: -lmkl_p4 before -lmkl 1205This change of library accelerates the si432 calculation by 12% overall 1206-------------------------------------------------------------------------------- 1207rel1_15_2 1208Ecut.h: exit without any action if ecut is same as current value. Caused 1209zeroing of the wavefunction velocity when used in a redundant manner. 1210Added isodate.C/isodate.h and release.C/release.h files for separate access 1211to these functions from any other object. 1212Added a <description> tag in the sample file in SaveCmd.C including the ISO 1213date and the release of qbox that was used to create the file. 1214(Note: the tag was already allowed by the Schema specification, but the 1215save command did not write any description information.) 1216-------------------------------------------------------------------------------- 1217rel1_15_1 1218Fixed bug: stop on assert in NonLocalPotential for l=1 or l=2 projectors 1219Fixed bug: Matrix::init_size() with zero dimensions. 1220Known bugs: 1221-wf_diag variable is ignored. 1222-------------------------------------------------------------------------------- 1223rel1_15_0 1224Added R option for the cell_lock variable: rescale unit cell only (keep aspect 1225ratio fixed). 1226Corrected inconsistencies between BOSampleStepper.C and CPSampleStepper.C as 1227regards printing of variables during a run: unit_cell, enthalpy, pv. 1228Known bugs: 1229-Stops on assert in NonLocalPotential::update_twnl() for l=1 or l=2 projectors 1230-Crashes when defining a species without using it, i.e. species with zero atoms. 1231 Traced to incorrect initialization of sizes in Matrix.C 1232-------------------------------------------------------------------------------- 1233rel1_14_3 1234Fixed calculation of ekin_e in MDWavefunctionStepper.C to avoid small error 1235when printing ekin_e at the first iteration of a CP run. Note: only the printed 1236ekin_e is affected, wfs are correct. 1237-------------------------------------------------------------------------------- 1238rel1_14_2 1239Fixed Thermostat in MDIonicStepper.C 1240Fixed use of ionic velocities at first step in CPSampleStepper.C 1241Known bug: 1242Calculation of ekin_e in MDWavefunctionStepper.C at the first iteration uses 1243the wf velocity at t=0. This is second order accurate, but not consistent with 1244the calculation of ekin_e in subsequent steps, which introduces a small error 1245in the printed ekin_e at the first iteration. Ekin_e should be calculated using 1246a first-order expression, i.e. (wf0-wfm)/dt to be consistent. 1247-------------------------------------------------------------------------------- 1248rel1_14_1a 1249SlaterDet.C: Removed messages from resize about cell being enclosed by ref_cell. 1250Note: the determination of enclosure should be revised. 1251SDCellStepper.C: removed debug messages. New "S" option to preserve shape. 1252BOSampleStepper.C: Added output of cell dimensions if cell is not locked. 1253Preconditioner.C: bug fixed in calculation of diag when using confinement. 1254Known bugs: 1255Does not use ionic velocities at first step of CP dynamics. 1256Thermostat is not working. 1257-------------------------------------------------------------------------------- 1258rel1_14_0 1259Matrix.h Reintroduced the clear() call in the resize members. Working with 1260uninitialized matrices sometimes caused NaNs to appear in the padding areas 1261of the danl matrix in NonLocalPotential.C. 1262Note that the clear() call was removed in rel1_13_0, which invalidates 1_13_0. 1263Matrix.C: Added assert()'s in clear() functions. 1264BOSampleStepper.C: Introduced quadratic wf extrapolation option (switch 1265in BOSampleStepper.C, currently disabled in 1.14.0. Added calculation 1266of forces at each electronic step to monitor the convergence 1267of the force x velocity product (disabled in 1.14.0). 1268PSDWavefunctionStepper.C: Fixed bug in usage of preconditioner (affects 12691_13_0 only). 1270PSDAWavefunctionStepper.C: Fixed bug in usage of preconditioner (see above). 1271SlaterDet.[Ch]: introduced new orthogonalization and alignment functions:lowdin, 1272align, ortho_align. 1273 1274 1275-------------------------------------------------------------------------------- 1276rel1_13_0 (not for release, see Matrix clear() bug above) 1277Implemented stress calculation and variable cell dynamics 1278Stress calculation works for s-only non-local potentials. 1279Changed names of variables tau0, taup to r0, rp (and keep tau for scaled 1280coordinates) 1281modified the IonicStepper interface to separate the calculation of rp 1282from the update r0 := rp. 1283Removed a bug in calculation of eta at first step (was not using ekin_ion) 1284-------------------------------------------------------------------------------- 1285rel1_13_0a1 (not for release) 1286Computes stress tensor. Limited to s-only n-l projectors. No cell eq. of 1287motion yet. Tested against GP results in cubic Si64. 1288-------------------------------------------------------------------------------- 1289rel1_12_1 1290Removed Gauss-Jacobi quadrature in NonLocalPotential introduced in 1_12_0 and 1291reverted to trapezoidal rule. Tests on Si64 using trapezoidal, Simpson, Gauss- 1292Legendre and Gauss-Jacobi show that the trapezoidal rule is more efficient 1293than all other methods, likely because higher order methods are not in their 1294asymptotic range for reasonable numbers of points (4-32). A possible 1295improvement over the present status may be to use Richardson extrapolation, 1296since the trapezoidal rule has O(h^2) error. However, from the example of Si64, 1297little improvement is obtained for small (4-8) numbers of points. 1298-------------------------------------------------------------------------------- 1299rel1_12_0 1300Replaced trapezoidal integration in NonLocalPotential.C by a Gauss-Jacobi rule. 1301(Note: this was later found to be less efficient. Removed in 1.12.1e) 1302Fixed bug in PSDWavefunctionStepper.C and PSDAWavefunctionStepper.C: 1303use ecut if ecutprec==0.0. Caused nans when ecutprec was not set. 1304-------------------------------------------------------------------------------- 1305rel1_11_2 2003-12-04 1306Fixed Elan allocation failure problem in SlaterDet::write. Problem was caused 1307by the large number of buffers allocated by the elan lib for all the messages 1308sent to task 0 before a receive was posted for them. A barrier in the loop 1309over states limits the number of messages sent to task 0 at any given time. 1310Note: the problem likely doesn't show on AIX since the MPI implementation may 1311limit the number of buffered received messages. The barrier is added in the 1312AIX code too (SlaterDet::print). 1313-------------------------------------------------------------------------------- 1314rel1_11_1 1315Fixed uninitialized var eta_ in MDIonicStepper that caused errors on AIX. 1316Known problem: elan allocation failure on mcr when writing h2o64 sample 1317using 140 tasks on 70 nodes. Possible bug in SlaterDet::write if USE_CSTDIO_LFS. 1318When running on AIX (using SlaterDet::print instead of SlaterDet::write) no 1319problem arises, and memory usage per node is low.(fixed in 1_11_2). 1320-------------------------------------------------------------------------------- 1321rel1_11_0 1322BOSampleStepper: changed to do extrapolation and ionic move before 1323smoothing the wf. Uses wf velocity if available. Creates it if not 1324Included PSDA wf stepper. 1325BO dyn Works ok with PSDA. 1326Tested 1pe vs 2pe on CH4 BO dynamics: ok. 1327Corrected bug in SampleReader.C (reading of wf velocity, appeared in rel1_9_0) 1328-------------------------------------------------------------------------------- 1329rel1_10_0 (not for release) (Note: known bugs) 1330Complete rewrite of the SampleStepper hierarchy, including CP and BO classes. 1331Runs CP and BO/SD and BO/PSD tests ok. No extrapolation of wavefunctions yet. 1332Note: wavefunction extrapolation implemented yet. 1333Planned work: interchange nite iterations and ionic move in BO stepper, so 1334that the wfs at the end of the iteration are the GS for the ionic positions at 1335the same time. 1336XML files: a <description> tag (optional) was added to the sample.xsd file 1337to include any information about the sample that a user may want to include. 1338Bug in SampleReader.C: reading of density matrix for the 1339wavefunction velocity is incorrect. Causes out of sync MPI call and Abort 1340Corrected in 1_11_0 (candidate) 1341-------------------------------------------------------------------------------- 1342rel1_9_0 (Note: known bugs) 1343Large file support (LFS): LFS is not correctly supported for <iostream> in icc 1344and gcc 3.2. A workaround is implemented in SaveCmd.C, Wavefunction.C, 1345SlaterDet.C. The workaround is activated by defining the USE_CSTDIO_LFS macro. 1346Occupation numbers moved to the SlaterDet class. 1347Occupation numbers are not fixed to 2.0 anymore. Load and save now use 1348occupation numbers. 1349Bug in SampleReader.C: reading of density matrix for the 1350wavefunction velocity is incorrect. Causes out of sync MPI call and Abort 1351Corrected in 1_11_0 (candidate) 1352-------------------------------------------------------------------------------- 1353rel1_8_2 1354Fixed problem with wfvtmp in SampleReader: cannot load sample and 1355then set wf_dyn = MD. Also, the order of load and set wf_dyn MD commands causes 1356a seg fault. 1357-------------------------------------------------------------------------------- 1358rel1_8_1 1359Fixed large-file support for sample files. 1360Fixed resizing of wf velocity. 1361Defined Matrix::operator=. 1362Fixed bug in PSD optimization: G=0 coefficient had zero value of dt2bye. 1363Fixed bug in UnitCell::fold_in_ws: caused infinite loop in some cases. 1364-------------------------------------------------------------------------------- 1365rel1_8_0 1366Added thermostat, th_time, th_temp. 1367Added emass variable. 1368-------------------------------------------------------------------------------- 1369rel1_7_1 1370Fixed bug in calculation of electronic kinetic energy in SampleStepper.C which 1371caused small kink in ekin_e and spike in econst during restarts. 1372-------------------------------------------------------------------------------- 1373rel1_7_0 1374Implemented parallel XML preprocessing of sample files for fast loading. 1375Added -serial option to the load command to force serial parsing of files. 1376Fixed a bug in Context::string_send: was incorrect for zero-length strings 1377initialization of ibuf removed, unnecessary. 1378Added new dmin,dmax,imin,imax,string_send,string_recv,pmap functions in 1379Context 1380Changed the block distribution of the slabs in FourierTransform to adopt the 1381same type of block distribution used in ScaLAPACK. Needed in SampleReader.C 1382in order to read from the gfdata matrix. 1383-------------------------------------------------------------------------------- 1384rel1_6_1exp 13852003-07-24:modified initialization of coefficients in SlaterDet constructor 1386to choose g vectors of minimum norm. Note: reverting to -O2 option using xlC 1387after noting that -O3 causes incorrect results in testBasis. 1388Added dmin,dmax,imin,imax members to Context.C to identify process where 1389max or min is found. 1390-------------------------------------------------------------------------------- 1391rel1_6_0 1392Implemented MD for wfs and for ions, print ekin_e, ekin_ion, econst in 1393SampleStepper.C. Use Stoermer's rule to compute velocities at end points. 1394Implemented load/save of wf velocities 1395-------------------------------------------------------------------------------- 1396rel1_5_1 1397Fixed bug in NonLocalPotential that appeared for Contexts having multiple 1398columns and rows (computation of enl and fion incorrect). 1399-------------------------------------------------------------------------------- 1400rel1_5_0 May 21, 2003 1401Parallel Load/Save XML sample implemented. 1402Load is slow: 6 min for 256 states on a (64)^3 grid (H2O_64, 73 Ry) 1403Save is ok (30 s.) 1404Bug found in NonLocalPotential::update_eigr: index of first local atom not 1405computed. Fixed. Affects calculations with multiple-column contexts. 1406Fixed bug in NonLocalPotential::energy: loop over fnl elements. 1407Added optional ecut attribute to the wavefunction element to insure 1408restart from xml sample in exact same conditions. If ecut attribute is not 1409specified in xml file, a value is inferred from the grid size. 1410Fixed incorrect initialization of dwf in EnergyFunctional::energy. 1411Implemented PSD for wf_dyn. 1412-------------------------------------------------------------------------------- 1413rel1_4_0 May 16, 2003 1414Implemented byteswap operation when loading/saving XML base64 file on AIX. 1415(affects WavefunctionHandler.[Ch] and SlaterDet.[Ch]. 1416All storage in base64 form is therefore little-endian. No byte swapping needed 1417on Intel platforms. 1418-------------------------------------------------------------------------------- 1419rel1_3_0 May 16, 2003 1420New features in rel1_3_0: 1421XML input for sample, species, atomset, wavefunctions, including base64 1422encoding. New hiearchy of XML handler classes for parsing of sample files. 1423Parsing done with Apache xerces-c, 2.2.0. 1424Clean-up of FourierTransform: convention of fwd/bwd changed to be consistent 1425with usually accepted definitions (forward transform is Fourier analysis, i.e. 1426computation of Fourier coefficients, and involves scaling by 1/np012. Backward 1427transform is computation of real-space function from coefficients, no scaling). 1428Load command includes reading of sample and appropriate resizing of all 1429variables. 1430The "locked" variable is removed from the Atom class, and will be reintroduced 1431as a Constraint. This is to avoid having to include the "locked" info in the 1432XML definition of an Atom. 1433Builds and runs on Linux (linuc-pc_mpi_icc.mk) and on snow (aix_mpi.mk) 1434-------------------------------------------------------------------------------- 1435Jan 24, 2003 1436Checking in release rel1_0_0: 1437performs SD steps with LDA and local pseudopotentials 1438-------------------------------------------------------------------------------- 1439Thu Jan 9 16:01:42 PST 2003 1440Added UserInterface to qb. Commands Set, Print, Quit, Atom. 1441TO BE FIXED: SlaterDet cannot resize if more states than basis functions. 1442This forces the order in which ecut, cell and atoms are defined. 1443-------------------------------------------------------------------------------- 1444Tue Sep 24 14:19:13 PDT 2002 1445SlaterDet can exist on any context. 1446Developing a Wavefunction that can hold any number of SlaterDets for 1447arbitrary nspin and nkpoints. 1448-------------------------------------------------------------------------------- 1449