Changelog bug-fixes Amsterdam Modeling Suite 2022
Revision | Description |
---|---|
r116452 | AMS2022.105 released (Dec 13, 2023) This is the final release of AMS2022. |
r116020 | e-Reaxff: undo the previous change because it was apparently a bug present only in the version used to obtain published results in the 2016 paper. More recent e-Reaxff version do not contain it. |
r116005 | e-Reaxff: fix a bug that resulted in the electron-electron repulsion being underestimated. With this fix it should now be possible to reproduce the literature e-Reaxff results. |
r113500 | ADF NCI: Fixed a bug that resulted in DenSigned having the wrong sign. |
r112003 | AMS: Fix the FIRE optimizer so it can work with a fully frozen system. |
r111534 | BAND: Repair restarts using the NOCVdRhoPlot and NOPRdRhoPlot blocks”. |
r111504 | AMSInput: Bug with slicer tool “Rotation of normal to z-axis is not accurate”. |
r111330 | ADF: bug fix (RA-)X2C spin-orbit coupling with the SAOP model potential. |
r111074 | AMS2022.104 released (Mar 13, 2023) |
r111073 | AMSpectra: bugfix for crash when exporting x-y values. |
r110787 | AMSjobs: dynamic queues did not update automatically. |
r110783 | APPLE&P: Fix a bug introduced with r110482. |
r110519 | AMS: Fix a bug that could lead to premature termination of MD with molecule gun. |
r110482 | APPLE&P: Improve performance for systems with lone pairs. The speed-up can be up to factor 2. |
r110335 | APPLE&P: Avoid memory leak caused by a compiler bug. |
r110232 | AMSinput: fix the previous fix for .in files import: was broken on anything but Windows after the fix… |
r110148 | AMSinput: Avoid cmd window popup when using .in files in builder on Windows. Filter all supported file types for the builder molecule selection. |
r110141 | AMSinput: improve handling of .in files in builder, taking the bonds into account instead of auto-generating bonds. |
r110109 | APPLE&P: Avoid printing a warning about missing repulsion if both ions are positively or negatively charged. |
r110034 | BAND: with print totalenergy also print a list with the energies in Hartree in full precision. |
r109999 | AMSinput: map to unit cell no longer automatically calls Guess Bonds |
r109994 | GUI: CONECT records from PDB will now create bonds to closeds image for periodic systems. |
r109988 | AMSinput: Overwrite the atom forcefield types when using the APPLE&P atom typer. |
r109986 | APPLE&P: Remove unintentional output printed by the induced dipoles module. |
r109964 | GUI: option to optionally import CONECT records from PDB files. Note: only bonds with explicit CONECT records created, and bond orders will be set to 1. |
r109774 | AMSinput: fix EXCITATIONS, SELECTEXCITATIONS and MODIFYEXCITATIONS keys in fragment runs (incorrectly present). |
r109670 | Fix APPLE&P to ignore angles whose initial value if off by 60 degrees or more. This is necessary to avoid problems with molecules like PF6 |
r109572 | AMSinput: Add Contact distance option to the Molecule Gun options |
r109530 | AMSinput: fix ETSNOCV input for the fragment runs (incorrectly contained the ETSNOCV block) |
r109417 | AMSinput: generate proper run script for fragment calculations with atoms not assigned to a fragment (like the Ni(CO)4 tutorial). |
r109325 | ADF: bugfix POLTDDFT basis sets icw PBE. |
r109211 | AMS MD: Do not remove motion in directions where ApplyVelocity is zero. Prior to this fix, ApplyVelocity cannot coexist with ApplyForce on the same region because it overrides all components of the net velocity, effectivelly cancelling any effect of ApplyForce. If you need to recover the old behaviour (forcing a particular velocity component to zero), just set it to a tiny value like 1e-300. |
r109186 | AMSmovie: Show the FFMPEG error information when exporting a movie failed. |
r108958 | AMS MD: Fix a bug which caused the force applied by ApplyForce to be 904 times smaller than requested. |
r108926 | ADF: restore the ALLOW BADSCF option, which was no longer having any effect. |
r108919 | BAND: bug with EngineRestart. This would delete the restart file, and also occasionally crash. |
r108797 | PLAMS: bugfix reading System blocks with [Units] in the header of the Atoms or Lattice subblocks |
r108691 | AMSinput: Autografs less strict topology and sbu matching. |
r108682 | GUI: Fix scmd startup on Windows with named pipe, related to bugfix revision 108550. |
r108661 | GUI: Fix the autografs builder for certain topologies and centers. |
r108639 | AMS driver: fixed a crash with the combination of PES scan gap filling and GeometryOptimization%KeepIntermediateResults |
r108638 | AMS driver: fixed a bug where KeepIntermediateResults only worked with the FIRE optimizer |
r108618 | AMSinput: do not prevent the user from combining IQA and D4 dispersion correction. |
r108550 | GUI: On Windows, use the named pipe method for inter GUI communaction by default (previously via SCM_SCMD_PIPE). Use SCM_SCMD_NOPIPE to use the old method with sockets and a random port, or use SCM_SCMD_PORT to set a fixed port. |
r108388 | AMSinput: Fix enabling input options for the COSMO-RS Compound template. |
r108131 | OLED workflows: Fix issue when using @include with relative paths in the input. |
r107932 | GUI: Load the environment variables SCM_SCMD_PORT and SCM_SCMD_PIPE from the preferences before starting scmd. |
r107924 | AMSlevels: Fix SFO coefficients display in MO tooltip for frozen core results. |
r107677 | GUI: Unit conversion between energy (or wavenumber) and wavelength is less strict now, in particular for relative energy spectra (FCF). |
r107600 | AMS MD: Fix hang in fbMC molecular translation for molecules with vanishing net force. |
r107599 | AMS: Fix parsing input containing @include with inline blocks and fix performance issues handling big @include files. |
r107597 | AMSinput: Fix displayed default mass for eReaxFF El/Eh particles. |
r107554 | PLAMS: Fix crash when calling AMSJob.load_external() after finish(). |
r107334 | AMS: Fix a performance bug in APPLE&P, which may make it much faster (up to a few times).
Fix an issue in Reaxff that could result in NaNs in the VdW energy when using atoms with incompatible inner core parameters. |
r107334 | Molecule gun: fix a bug that caused AMS to ignore atomic info (such as the region) in the added system if the main system did not have any. |
r107326 | AMSinput: when updating the geometry after opening a finished job, proplery update the lattice vector panel. |
r107316 | AMSinput: avoid crash trying to select atoms with same PDB name if no atoms selected. |
r107303 | BAND: Prevent error VOCOrthogonalizerModule: no current k point. This could happen for MetaGGAs with more than one k-point. |
r107180 | AMS: Solved performance issue for geometry optimizations of very large systems with a fast engine. |
r107062 | AMSinput: added extra checks for a correct setup for a CPL calculation. |
r107032 | AMSinput: fix NEB with atom constraints. |
r107010 | AMSinput undo fix for NEB with atom constraints, it breaks handling of multiple molecules. |
r106986 | AMSspectra: handle reading results with no spectra info without crash.’ |
r106978 | AMSinput: fix NEB with atom constraints. |
r106965 | AMS: prevent symmetry error: no equivalent error found opratm, in case of a numerical phonon run on a linear 1D periodic system. |
r106929 | AMSinput: Fix the pre-optimizer when dependencies (such as forcefield files) are needed. |
r106905 | ForceField engine: Fix crash when running with only one CPU per (NUMA) node. |
r106801 | Bugfix: Fixed an issue with python modules being unusable together with a cloud license. |
r106799 | amsbashrc.sh: Check if /scratch is also writable, to use as SCM_TMPDIR. |
r106641 | AMSmovie: Fix exporting ffmpeg movies when -loop is used (e.g. normal modes or phonons). |
r106600 | ADF: bug fix velocity-gauge oscillator strengths icw TDA for LDA or GGA XC functionals. The velocity-gauge oscillator strengths icw TDA for hybrids does not produce correct results. |
r106318 | AMS: In the file amsresults/output.xyz element names were truncted to two characters. |
r106112 | ADF: Fix calculating bond energy in case of 1 SCF iteration for range separated and hybrid XC functionals. |
r106048 | AMSinput: Fixed the runscript generation for hyperpolarizability calculations with ADF. A space was missing between the frequencies keyword and the value. |
r105877 | AMS input reader (fortran) can now do conversion between inverse units. E.g. it can convert 1/bohr to 1/angstrom. This solves a bug where AMSInput would write the unit [1/angstrom] for the option BandStructure%DeltaK (Band and DFTB) if you changed the default unit. |
r105796 | AMSinput: On Linux, fix window popup on startup. |
r105759 | AMSjobs: Fix ‘kill’ to kill jobs that are incorrectly marked queued. |
r105719 | Bugfix reaxff: avoid calling Frotran write() statement by more than one thread at once. |
r105521 | ParAMS: Update ReaxFF parameter database |
r105444 | BAND: for spin unrestricted calculations sometimes a band gap was reported whereas there was none. This happened when only one the spins had fractional occupations. |
r105211 | AMSspectra: fix markers for NMR spin-spin splitted spectra |
r105172 | ADF: fix input handling CDFT in case of ChargeAndSpin. |
r105018 | BAND: Voronoi (deformation) charges now relative to fragments used. They used to be independent of the choice of fragments. |
r104886 | AMS2022.103 released (Jun 21th, 2022) |
r104885 | ReaxFF: Fix incorrect formatting of the force field FeOCHCl.ff |
r104851 | ParAMS GUI: fix clearing ref values (which should revert to proper ref value if defined), and adding reference values if they don’t already exist. |
r104821 | AMS MD: Fix a bug in the molecule sink (RemoveMolecules) where a successful removal of some molecules caused other MD features executing during the same timestep (other sinks, CVHD, fbMC molecular moves) to work with invalid bonds and molecule data. |
r104792 | Params: Replace non-ASCII characters with ASCII equivalents in the ReaxFF forcefield files, to avoid encoding issues on Windows. |
r104666 | QE: Fix parsing the Fermi energy from spin polarized QE results. |
r104660 | AMSinput: Avoid crash with Fuse Atoms if no atoms selected. |
r104656 | AMSdos: Fix reading spin polarized QE results. |
r104499 | AMSinput: Store the molecule styles used when saving. |
r104472 | AMSmovie: Fix “Select atom close to origin” in AMSmovie. |
r104470 | QE: Fix parsing the QE CP results and converting to a kf result file that is readable by the GUI. |
r104458 | BAND,DFTB: Allow for EngineAutomations during an NEB calculation. This can help to avoid non converging SCF (or SCC) problems while the gradients are still large. The error message was “ERROR: Energy calculation failed. Stopping optimization” |
r104456 | BAND: Error when calculating the band structure if the parent directory name had a space in it. The error message contained “-c: error: unrecognized arguments:” |
r104440 | AMSpackages: Fix potential gui hanging issue when installing an invalid package. |
r104339 | Densf: bug fix: avoid reading number of atoms from CUBINPUT file if it is only used for grid definition. |
r104318 | ParAMS GUI: avoid hanging gui when job crashes/finishes very quickly. |
r104206 | AMSinput: fix saving regions if defined for multiple molecules, and handling of All region for multiple molecules. |
r104158 | AMSinput: fix reading multiple molecules from .ams files (the order was not guaranteed, causing issues for jobs pointing to multiple molecules_. |
r104123 | AMSinput: fix importing multiple system blocks from run script |
r103978 | BAND: Changed the default value of Dependency%Core from 0.98 to 0.80, thus making the “FROZEN CORE SEEMS TO BE TOO LARGE” error much less likely. |
r103939 | Params gui: fix saving adjusted AMS settings for a job in a jobcollection (the change was ignored). |
r103907 | Params: Fix importing VASP OUTCAR jobs on Windows. |
r103881 | Params: Fix losing atom selection on menu use on Windows. |
r103878 | AMSinput: fix capitalization of Params -> ParAMS. |
r103851 | Update the download_repository.py utility script. |
r103841 | GUI: Fix the initial value for the periodic cell range view, after having it changed once. |
r103794 | AMSjobs: Add extra check for jobs marked killed by the GUI, in case they were finished correctly. This can happen on slower Networked File Systems. Requires the env var SCM_CHECK_JOBSTATUS to be set. |
r103784 | ADF/BAND: QTAIM might warn “PSearch37 WARNING: More critical points found than can fit”, especially when using many cores. Potentially critical points could be missed. |
r103759 | AMSmovie: change startup to avoid wrong window layout in rare cases. |
r103709 | PLAMS: Fix a filename clash for jobs whose name explicitly ends with the counting suffix (e.g. “.002”). A manually supplied counting suffix is now ignored. Also fix a potential race condition in the JobManager class. |
r103692 | AMS (Linux/Mac only): Try to delete temporary files when the program receives a segmentation fault (program bug) or a bus error (full disk) signal. |
r103678 | AMSinput: if opening very old AMS jobs, result dialog asking what to import was misbehaving (prog not found). |
r103675 | AMSinput: avoid unneeded traces (effect was to slow down AMSinput). |
r103649 | Add a possibility to use InfiniBand adapter info for licensing as the last resort. |
r103598 | AMS2022.102 released (Apr 26th, 2022) |
r103597 | BAND: All printed band energy ranges would be zero if using a finite electronic temperature and a single k-point. Energy and gradients not affected, only the table was wrong. |
r103468 | AMSinput: pre-optimizers now run in serial mode by default. This can be adjusted with a toggle in AMSpreferences. |
r103443 | ADF: BugFix: LibXC meta-GGA and meta-hybrid functionals can not be combined with analytical frequencies. |
r103363 | AMSbands: Fix display of bonds and lattice for phonons results. |
r103346 | AMSjobs: fix display of warning messages in the two-line summary below the jobs |
r103342 | AMSspectra: Fix opening BAND phonons results in AMSspectra. |
r103324 | ParAMS GUI: properly resize graph windows when changing window size. |
r103281 | AMSinput: fix graphics glitch in Notebook toolbar on some machines. |
r103270 | GUI: Fix issue with certain GUI modules (mainly AMSinput) where the window was previously on a second monitor which is no longer connected. |
r103258 | AMS MD: Make Deformation raise an error when TargetLength contains negative components instead of ignoring them silently. |
r103257 | ReaxFF: Fix handling lattices with negative diagonal components. |
r103230 | Documentation: Fix broken link in AWS parallelcluster installation guide. |
r103223 | BAND: When using band as an engine for the MD molecular gun the program woudl stop with NAODataModule: unexpected spherHarmonicType. |
r103217 | MOPAC: Remove MOPAC single-point output files created by slave processes (there can be many of them during MD). |
r103143 | AMSinput: Fix the fuse atoms function when there are bonds across the periodic boundary. |
r103137 | AMSinput: Fix issue with the atom fftype check when not all atoms have an fftype, when switching forcefield types. |
r103128 | ParAMS GUI: do not save empty training or validation set. |
r103096 | GUI: Fix issue with nested busy indicators. |
r103090 | AMS MD: Fix several bugs in the ReflectiveWall feature. |
r103071 | AMSreport: Fix exporting history frame in SDF format. |
r103066 | AMSreport: Fix using history frames where atoms were added or removed, where e.g. a molecule gun or sink was used. |
r103061 | Fix bug in configure script for source distribution. |
r103043 | AMS: Added a check for normal modes in symmetric displacements in combination with periodicity (which is not supported). |
r102975 | BAND: Improved handling of Frozen Core too large error. There was a bug that the core/core overlap matrix was tested for the first k-point, rather than the intended gamma k-point. Also the feedback has been improved about which element is causing the problem. |
r102971 | AMSmove chemtrayzer2: fix chemtrayzer2 name (no space in front of 2) |
r102966 | AMSinput fix coloring of some fields on the Hybrid panel. |
r102963 | AMSmovie chemtrayzer2: proper coloring of input fields. |
r102949 | AMSmovie: Enable Select Molecule and Select Connected from the Selection menu. |
r102943 | AMSmovie: fix the Details panel for atom details (colors, selecting atoms, …) |
r102932 | 3D-RISM: Correct an error in the coordinates of the 3D-RISM methanol solvent. |
r102865 | AMSpackages: GUI now refreshes the package overview after aborting an operation. |
r102813 | AMSmovie: Fix periodic view change for trajectory files in xyz format. |
r102808 | BAND: clean up semaphores at the end of a calculation. On the MaxOS this could lead to problems when running many calculations, because the system would run out of available semaphores/mutexes. This does not affect any of the results. |
r102797 | BAND: prevent warning “WARNING: Nr of shared arrays …” when doing a restart (orbital/densityplot). |
r102768 | AMSinput: Add busy indicator for saving larger jobs. |
r102751 | Tutorials: Make stress_strain_curve.py compatible with AMS2020. |
r102734 | ADF: bug fix single point gradients dDsC dispersion correction. |
r102719 | OLED material DB updated to v2022.1-1: Switched alpha- and beta-MADN. Our names were wrong for these compounds. Outliers with a modified Z-score >16 are now removed prior to calculating mean and sigma of IP/EA. |
r102701 | AMS: Fix a few minor bugs in the shared array locking. |
r102700 | AMScrs: Fix default window size with screen scaling or larger font, and remember the size on startup like the other GUI modules. |
r102672 | AMSinput: fix 3d-rism when the path to AMSHOME contains capitals. |
r102658 | GUI: Avoid issue caused by rapid inputs that create or update a graph window. |
r102655 | AMSinput: Avoid All-type constraints as scan coordinate. |
r102653 | AMSinput: Preoptimize with GFN1-xTB now also clears the parameter directory. |
r102649 | IRC: Converged IRC geometries, including end points, should now have all the requested properties calculated. |
r102631 | AMSinput: Clarify and add import button for the OLED materials database. |
r102629 | ParAMS: Handle PES reference values with incorrect shape inside Logger. |
r102626 | AMSmovie: Better handling of incorrect kf files from remote running jobs. |
r102606 | ParAMS documentation: Fix some incorrect parameter names starting with ‘n/a’. |
r102604 | AMSinput: remove Excellent numerical quality option from menu as it is ridiculously expensive and sometimes has numerical issues. |
r102597 |
AMS: Work around a bug in Cluster MKL on AMD Zen that may cause AMS to stop with an “Error exit from DistributedMatrix%SVD” message.
IRC: save engine rkf files for the final points on the IRC path; enable calculation of optional properties for the initial IRC point. |
r102585 | AMSinput: reading in older input files (ADF2019 or earlier) could have duplicate periodic bonds. |
r102580 | AMSjobs: transfer of all job files from remote systems would fail if the path contains spaces. |
r102575 | AMS: fix a bug related to a combination of frozen atoms and RemoveMolecules during molecular dynamics. |
r102550 | ACErxn: Make it work for very simple systems very simple reactions. |
r102536 | AMSinput: improve stability of feedback of pre-optimization. |
r102533 | ParAMS GUI: improve stability for monitoring running jobs. |
r102512 | AMS: Speed up calculation of the induced dipole moments by caching moment-independent pre-factors in the electric field calculation. |
r102487 | AMSinput: when parsing input (for example via the ParAMS GUI), switch to proper engine before handling AMS options as the engine switching could reset AMS options like the task. |
r102484 | BAND: small performance improvement for calculation with Core=None. |
r102481 | ReaxFF: fixed a bug that resulted in errors in the valence angle contributions to the force and the virial for a small unit cell. The bug existed in all ReaxFF versions and it was the main reason for not allowing small unit cells. With this fix, also the way the stress tensor is calculated has been changed, so the size of the unit cell should be less likely to cause issues. |
r102478 | Params GUI: click in scatter plot on distances, angles and dihedrals: give feedback by selectig proper atoms. |
r102475 | BAND: reduce memory footprint for metaGGA (force) calculations, and also make them faster for bigger systems. |
r102465 | Quantum Espresso: Fix bug in qerc.sh. |
r102457 | ParAMS GUI: avoid change of shape for 1-d array values when editting. |
r102452 | GUI: Show Selection Only would fail to show anything if all atoms selected. |
r102447 | ParAMS GUI: Show correct atom when clicking in a scatter plot with forces. |
r102427 | AMSinput: Better handling of explicit forcefield files for the ForceField engine (expert option). |
r102424 | ParAMS GUI: try to fix stability issues while monitoring running calculations. |
r102411 | ParAMS: Fix bug when reading an lgDispersion ReaxFF force field |
r102395 | ParAMS GUI: fix stability issue when displaying results of a running optimization (error about xxmin). |
r102387 | Update repository download python script for AMSpackages offline usage. |
r102382 | AMS: Add a possibility to configure the KF page size (=file size increment) via the SCM_KF_PAGESIZE environment variable (which is in kB). |
r102373 | AMS: fix a bug in IRC that prevented it from computing optional properties (such various ADF chanrges) at converged reaction path points. |
r102364 | AMSinput: fix the Add Region In Same Way To Identical Molecules option on the Regions panel. |
r102341 | ParAMS GUI: validation set results should be from training set best parameters, and make that explicit with the label. |
r102338 | ParAMS GUI: fix/change color for errors and validation set in table |
r102315 | AMSinput: fix ‘Add region in same way to identical molecules’ command on Regions panel. |
r102305 | Fix division by zero in WallPotential when an atom is located exactly at the origin. |
102269 | Update AWS parallelcluster guide with information about multi-node jobs. |
r102258 | AMSinput: fix protein controls on Regions panel. |
r102257 | AMScrs: fix: .multipleform file should not contain density in one of its forms, only at the compound level. |
r102252 | AMSdos: show bonds for BAND results. |
r102245 | Tutorials: Fix images for Young’s modulus, yield point, Poisson’s ratio. |
r102235 | AMSinput: fix several screen dpi issues related to the notebook bar and dropdown menubuttons. |
r102231 | AMSjobs: update Job Clean Up menu comamnd to handle Params results, and show only files present. |
r102226 | Fix Fukui tutorial: AMSoutput command is now called Conceptual DFT (FDL) |
r102220 | AMSmovie: fix 2D Histograms for MD, with non-symmetric bin counts. |
r102215 | Remove old ReaxFF ParAMS example. |
r102208 | Add a description for the APPLE&P force field file format. |
r102207 | The pressure should now always be calculated during MD when using the ForceField engine. |
r102192 | Update versions of QE in documentation for GUI. |
r102189 | Update default value of QE_VERSION to 7.0 in qerc.sh. |
r102186 | Remove pytest-qt from amspython environment. |
r102176 | AMSinput: fix graphic glitch in notebook bar. |
r102170 | Params gui: avoid crash by using Graph menus when no results are present. |
r102167 | AMSinput: fix writing the ElectricField input key to the run script. |
r102163 | AMSinput: fix unit conversion when the user changed the default and opens a .ams file without explicit units set. |
r102158 | AMSjobs: avoid crash when the modification time of a job cannot be determined. |
r102146 | AMSmovie: display molecule for non-trajectory results. |
r102135 | Only compute transition dipole moments between excited states if asked for. |
r102128 | Update the reference for Quantum Espresso 7.0 in documentation |
r102085 | AMS2022.101 released (Feb 28th, 2022) |