Releases: mlund/faunus
Version 2.10.0
What's Changed
- Support for Gibbs ensemble for studying phase equilibria.
- Updated documentation about multipole analysis by @IVinterbladh in #432
- Add Random slump parameter for tempering by @IVinterbladh in #434. This should make it possible to use grand canonical schemes with parallel tempering.
Full Changelog: v2.9.1...v2.10.0
Version 2.9.1
Changes
- Fix "progresstracker" compilation error on some systems, incl.
Full Changelog: v2.9.0...v2.9.1
Version 2.9.0
What's Changed
- Add a
preface
section to the user input which may contain a list of actions to be run prior to simulation. The discrete angle
analysis below is an example of this. - Added
--norun
command line option to skip simulation. - Discrete angle integration by @mlund in #427
This allows for explicit angular integration of two rotating, rigid bodies. Energies and quaternions are streamed to disk and free energies and thermally averaged energies are reported. - Update minimum cmake version stated in manual by @SHervoe-Hansen in #424
- Add VSC dev container config files by @mlund in #425
- Enabled OpenMP by default
Full Changelog: v2.8.0...v2.9.0
Version 2.8.0
What's Changed
- Fix speciation moves for parallel tempering by @mlund in #419
- Custom group to group pair-potential by @mlund in #421
- Rename namespace and base classes by @mlund in #422
- Upgrade nanobench from v3 to v4 by @mlund in #423
- Table cleanup by @mlund in #420
- Fix GCC warnings
Full Changelog: v2.7.0...v2.8.0
Version 2.7.0
What's Changed
- Add minimum distance option to
groupmatrix
analysis by @MarcoPolimeni in #418 - Add scipy clustering example for
groupmatrix
analysis - Add optional particle-particle threshold to cluster move (
com
keyword) - Add PSC patchy particle unittest (964948e)
- Upgraded external libraries (0bf11d2)
- Updated installation instructions (8bc5f8c)
- Updated TBB instructions for when using C++17 parallel algorithms and certain GCC installations.
- Fixed a potentially dangling reference (7d168cb)
- OpenMP is now disabled by default (62036d3)
- Fixed GCC 12 warnings
- Clarify positions keyword in manual (6da63c8)
- Fix plot pair-potential notebook
Full Changelog: v2.6.1...v2.7.0
Version 2.6.1
What's Changed
- Add policy to swap PSC patch information only by @mlund in #415
- Group matrix analysis by @mlund in #416
Full Changelog: v2.6.0...v2.6.1
Version 2.6.0
Since the last release, 139 commits have been added. Full list of changes from v2.5.0 to v2.6.0.
New features
- Generalised smart Monte Carlo move with support for regions (#405)
- Quadratic charge move (#406)
- Debye screening length can be deduced for complex salts
- MPL is used for all MPI calls (#398)
- Faster, internal SASA hamiltonian with support for periodic boundary conditions (#397)
- Separate density analysis for atoms and groups
- Reduce parallel summation memory requirement by a factor of two
- Patchy sphero-cylinders (PSCs) are back! (#408)
- PSC trajectory format
- VMD visualiation
- Ancients tests are back
- Enhanced translational displacement analysis for Dynamic MC (experimental)
- Ability to save penalty function tables to disk during simulation
Compatibility
- Compilation defaults to C++20 standard meaning that a pretty modern compiler is needed.
- A
scripts/Dockerfile
is provided for creating containers with JupyterLab and Faunus preinstalled.
Partial list of fixes and enhancements
- Refactoring of speciation move to allow for future expansion (#410)
- Fix
dirrot
mass center error (bf088b4) - Update manual for units on angles (
dprot
) - Fix potential harmonic bond breaking (#403)
- Fix time stamp in output (#395)
- Fix ion-ion Ewald energy (28f0c41)
- Enable C++20 mode compilation (a975438)
- Correctly handle inactive particles when using group output with xtc files (#402)
- Fix 'constraint' energy when loading state (0d27a68)
- Safer xtc file handling
Version 2.5.0
Since 2.4.0 we have included more than 120 commits and closed 59 pull requests. While mostly invisible for users, we have made lots of internal refactoring and unit-testing to make Faunus more robust and maintainable.
Enhancements
- Electric potential analysis (#379)
- Constant interval moves (#381)
- Force calculation and LD propagation of angular potentials (#368)
- FASTA file loader
- Improved structure file I/O (#376)
- PCG Random number generator supprt (#369)
- Add Langevin Dynamics that can be mixed with Monte Carlo moves (#298). It works with common pair-potentials and a subset of bonded interactions.
- Add particle energy histogram to
transrot
move (#358) - Add shape anisotropy analysis to
clustermove
(#336) - Add
copy_policy
to conformation swap move (#354) - Support mapping of hexagonal cells to cuboids (#328)
- Allow cuboids and hexagons in scattering analysis (#327)
- Enable Intel Threading Building Blocks (#325)
- Polymer shape analysis and tensor file output (#323, ea7c628)
- Enable integer selection for inactive molecules (#317)
- Introduced overwrite option to
savestate
(#309) - Allow Gzip compression a number of streaming outputs (#289, #308)
- Add replay move (52e5364)
- Optional to perform filename increment for
savestate
(#292) - Anisotropic volume perturbation (#291)
- Optimize cluster search in
ClusterMove
(#285) - Relax Gouy-Chapman input (#248)
- Add conformation id analysis (2e929b8)
- Add python support module (555e8e7)
- Add Apple M1 support (#355)
- Store simulation duration in json output (#290)
- Added example:
bend
for mixed MC/MD - Handle build dependencies using CPM (#380)
Bug Fixes and Optimisations
- Fix empty error message from Hamiltonian (#360)
- Use CPM.cmake for CMake dependency handling (#380)
- Respect active particles in Langevin Dynamics (#324)
- Refactor widom analysis (#310)
- Splining of Yukawa fails bug (#301)
- Fix speciation check for full or empty groups (#300)
- Enable JSON output for nonbonded energy (#288)
- Fix
moltransrot
if atomic group (#287) - Dipole mass center fix (#334)
- Parallel Temper fix (#333)
- VirtualVolumeAnalysis (Space is a reference) (#330)
- AAM file reader (atoms begin at line 2) (#329)
- Update crankshaft in
schema.yml
(#307) - Fix numeric overflow in energy (4f57102, 9b42049)
- Fix possible infinite math in Release mode (b0859d9)
- Clarify keyword
inactive
in manual (#316) - Optimisation: binary search in
Bonded
energy (#335) - Upgrade external libraries (#297)
Version 2.4.2
Changes
Bug fix
-
Fixes a bug (#300) where products or reactants could be wrongly (de)activated even if the groups are full/empty. This is a problem important only when simulating at extreme conditions where the groups are nearly empty or nearly full. The issue could appear as a non-zero net-charge upon restarting an equilibration run which would typically be initiated with all molecules in their active state. The issue would not appear with swap moves and only if multiple products or reactants were involved in the reaction.
-
Fixed broken Eigen library URL
Nothing else has changed since v2.4.1
Version 2.4.1
In v2.4.0, the make install
target is broken. This patches (#281) this problem, while the main Faunus code is unchanged.