Single version of energy minimization for argon is working, need to add constraint
Add one more file into Makefile.in
Add some lines into global.cpp to make it work with energy minimization
Fix a bunch of bugs :-) Single version of conjugate gradient with golden search linesearch pass a couple of functions test. Brent's algorithm is still broken
NLModel0, NLModel1 pass uit test
*** empty log message ***
to avoid cyclic depency, refactory constraint class
Functor.hpp pass unit test
begin unit test of minimizer
Adding GoldenSection and Brent LineSearch Method
using class Minimizer1D derived from MinimizerBase instead of a functor to do line seach
Substantial changes. OOPSE now has a working WATER.cpp forcefield and parser. This involved changes to WATER.cpp and ForceFields amoung other files. One important note: a hardwiring of LJ_rcut was made in calc_LJ_FF.F90. This will be removed on the next commit...
member list fixes for rigid bodies
Adding MinimizerParameterSet class.
Revision of NLModel0 and NLModel1
revision of NLModel
More BASS changes to do new rigidBody scheme a copy of WATER.cpp from this morning
revision of constraint for Nonlinear Optimization Model
Convert Eulers in degrees into radians
Changed default orientation in BASS to use Euler angles in the following order: phi, theta, psi Removed the ability to set orientation using a unit vector
Corrected spelling in several directories, and stated WATER.cpp
constraint class in energy minimization
Adding energy minimization
Energy Minimizer
Made some error messages more user-friendly
Adding warning if sample time, status time, thermal time and reset time are not divisible by dt
BASS changes to add RigidBodies required a change in how the MoleculeStamps are used by divideLabor in mpiSimulation.cpp
fix a bug in creating eor file
fixed a bug where only MPI jobs could write eor files
fixed an struct mismatch error in the mpi initialization of the AtomStruct
Fixes for Dumps
Documented the Spud Toss
changes for charge charge interactions
More work for adding charges
Changes for adding direct charge-charge interactions (with switching function)
Some changes for new MPI organization and direct charge-charge interactions
*** empty log message ***
open and close the eor file whenever it is used instead of rewinding it
change the interface of writeFrame
Merge the code of writeFinal and writeDump; Adding sortingIndex into DumpWriter; Fix a bug of writing last frame twice in integrator
fix a bug in copying string
Dumpwriter only write out the atoms on master nodes
tagub is not a bug. Just roll it back fix a bug of copying string to a pointer Still have Seg fault, it looks like a random MPI seg fault in totalview
Fix a bug of declaration of tagub
New DumpWriter (Attempt #4)
A work in progress...
null terminate some strings just in case
refixed the NVT readin XS state bug.
added strncpy to DumpWriter
fixed the restart from NVT exstended state bug
First Stab at fixing DumpWriter
added support for the ignore XS state info flag
Fixed a bug of sending message from master node to itself in DumpWriter.cpp and InitializeFromFile.cpp
performance fixes in the dipole dipole and reaction field code
Making do_Forces a little more sane
Attempting to increase performance by reducing spurious function calls
mangling forces even further
mangled do_forces...
Added bitmask to do_forces property lookup
Fixes to profile code.
More profiling fixes.
Another change for MPI in timing.
Small update to timing in MPI
the profiling commands work now. Will start adding PROFILE ifdefs into the code
added some profiling routines
added some profile functionality
Added functions for simple profiling in fortran.
Fixed bug in parallel EAM. rho_row and rho_col were scattered into the same array. Unfortunately, MPI zeros the array between scatters so half of the sum was being lost. Fixed by added a temp array for column scatter, then sum loop over nlocal.
Changes for gradients (to do minimizations)
added a more verbose error message in SimInfo. Added a more informative error message in InitializeFromFile
Fixed a bug in SimInfo ordering of radii
added a routine to SimInfo.cpp to inline a min function.
reordered the rcut/ecr/boxSize initialization removed the rcut/ecr shrink and grow algorithm. the simulation will now exit when it runs into rcut or ecr.
Added support for compiling fortran without use of mpich modules. We use mpif.h instead.:
moved the velocity scale matrix calculation outside of the atom loop in the NPT family of integrators.
added the following parameters to BASS: * useInitialExtendedSystemState * orthoBoxTolerance * useIntiTime => useInitialTime
did a merge by hand from the new-templateless branch to the main trunk.
bug Fixes include:
* fixed the switching function from ortho to non-ortho box.
!!!!! THis was responsible for all of the sudden deaths we saw.
* some formating in the string when we write out the extended system state.
* added NPT.cpp to the makefile.in
added template stuff to the Maikefile template little changes to some printf format statements
bug fixes for rList problems
Fixed bug that size(q0) was being queried before q0 was allocated.
fixed a stdlib.h include error in bass.l fixed a little bug in the first time step, regarding the setting of ecr and est in fortran
som efixes to the way rcut is setup, as well as additional debugging comments.
C++ compatibility for templates
add chi and eta to the comment line of dump file.
did a complete overhaul of how c calls fortran. All function pointers and fortran calls are rigidly typecast now.
Compatibility fixes
started trying to understand extern "C" stuff for pointers
fixes for compatibility
replace c++ header stuff with more portable c header stuff Also, mod file fixes and portability changes Some fortran changes will need to be reversed.
Stuff for MOD support in other compilers
added routines for the sysbuilder to work with simSetup
added eam ForceField files to the init fixed an eam mpi parmeter setup bug
added a new NPT integrator, NPTxyz. It scales the x, y, and z direction sepeartely. no box skew allowed.
added useInitTime to the BASS syntax. * useInitTime = false: sets the origin time to 0.0 regardless of the time stamp in the .init file * default=> useInitTime = true;
Changed DumpReader to use linked lists instead of a vector. Fixed the makefile to build DumpReader.cpp Removed a comment output in Exclude.cpp Modified DumpWriter and Integrator to write an eor file every time a frame is written. This lets the .eor file represent the last written frame of a simulation.
Contiuned work on staticProps. should be in a position where it will compile and run first runs.
Fixed bug in calc_eam.
removed entahlpy from the statwriter and thermo.
changed the formating ogf the error statements in simError added a function to get the maxCutoff
added mpif90 mod check back same for conifig.in fixed wrappers to extern "C"
fixed a lot of warnings and errors found with SUN's SUNWspro.s1s7
light change in syntax. no signifigant change.
fixed some additional remarks from icc -w3 (extra verbose output)
cleaned things with gcc -Wall and g++ -Wall
fixes for configure
moved readyCheck in the integrator so that it is called before the first Statistics are written.
Removed NPTfm from Integrator.hpp. Some small syntax cleaning in NPTfm and SimSetup
fix bug in calculating maxCutoff
Converted NPTf to work with the NPT base class. Removed NPTfm and NPTim from cvs
added NPT base class. NPTi is up to date. NPTf is not.
fixed bugs in NPTf, found (nearly) conserved quantities for both NPTi and NPTf
*** empty log message ***
*** empty log message ***
fixed NPTi to now work with constraints.
fixed ecr grow in SimInfo fixed conserved quantity in NPT (Still some small bug) NPTi appears very stable.
add conserved quantity to statWriter fix bug of vector wrapping at NPTi
Added integrators to Makefile.in
updated the ChangeLog added two new NPT integrators, they still need work.
dependency on config.h
Changes to autoconf / configure method of configuring OOPSE
added resetTime to the Global namespace. added ability to reset the integrators in the NVT and NPT family.
fix a bug at MPI version of PolicyByMass
Added: check uniqueness of molIndex
fix bug of MPI_Allreduce in ZConstraint, the MPITYPE is set to MPI_DOUBLE, however, the corret type is MPI_INT. Therefore, when we turn on the optimization flag, it causes a seg fault
More fixes for stress tensor parallel bug.
fix bug in calc_dipole_dipole.F90 and calc_stikcy_pair.F90 molMembershipList use global index instead of local index
set default force substraction policy to PolicyByMass
*** empty log message ***
added define statemewnt to Statwriter and Dumpwriter to handle files larger than 2 gb. commented out some print statements in Zconstraint
Use make_sprng_seed() to generate seed and check the seed which is specified by user at least contains 9 digits
changed the Makefiel a litle. Fixed a bug in MPI_DUFF. The atom block type was not being properly constucted in MPI. (The MPI struct had 6 doubles declared versus the actual 11)
*** empty log message ***
small bug fix on frequency of output dumps.
user can setup seed in bass file now, if he does not specify any value for seed, oopse will take the value of seconds of system time as seed
updated the Changelog. added some bug fixes for setting the random number generator seed value. fixed a bug where ghostbend atom b was not being set. ( recent bug from SimState conversion)
bug fixed in ghost bend class
*** empty log message ***
reformmating ZConstraint and fixe bug of error msg printing
Tested MPI version of Z-Constraint Method
Stable ZConstraint with average force substraction strategy
Added some profiling code -DPROFILE.
harmonic potential & z-contraint method
fixed a really annoying bug in Directional Atom, where mu was getting written to pseudorandom memory location.
debugging globals
added harmonical potential to z-constraint method
added method of moving zconstraint molecules to specified positions
*** empty log message ***
Added zConstraint into the BASS language syntax.
changed the number of degrees of freedom to account for zConstreints
EAM works...... Neighbor list also works.....
fixed a deprcated instance of Atom::setZ and Atom::getZ in ZConstaint.
switched SimInfo to use a system configuration from SimState rather than arrays from Atom
Bug fixes for eam...
stable version of Z-Constraint
add index range checking into ZConstraint
Added Z constraint.
More bug fixes for eam.
working on the props code
Added eam to simSetup and added changecutoffeam.
module use fixes for eam and do_forces.
Finished most code for eam....
added the scan function to the DumpReader. It should now save the start of each frame in a vector.
making some changes to read dump files
*** empty log message ***
Fixed a current time initialization bug in InitFromFile.
some initial changes to Dumpwriter and friends to accomadate random file access
Initialized currentTime to 0, in case no one sets it.
fixed Initializefrom file to start the simulation from the time specified in the init file.
Started work on a DumpReader
Fixes for SSD/E
Changes for SSD/E
commented out an eam line
fixed spelling issue
added info module
Added massive changes for eam....
More up to date version of EAM_FF
Changed how cutoffs were handled from C. Now notifyCutoffs in Fortran notifies those who need the information of any changes to cutoffs.
Fixed bug in updating mixing lists
more fixes for box changes
removed some debugging print statements.
fixed a long lived bug in do_forces. Rrf was not being used in the neighborlist correctly. rcut was conssistently being set lowere than Rrf causing the dipole cutoff region to be to small. Also led to the removal of the taper region to buffer the dipole cutoff.
working on fixing ssd bug
Fixes for the NPT ensembles
cleaned up simSetup
fixed some bugs, Changed entry_plug to info where appropriate
Fixing pressure tensor
more archaic code fixes
removed old outdated code
fixes to get rid of get_vx and set_vx
removing get_vx
Added NPTfm
Bugfix in NPTim, fixes for NPTfm
Checking in changes for NPTim
Removed some debugging write statements
Fixes for get and set routines in Atom and DirectionalAtom
added get and set routines to Atom and DirectionalAtom
found a bug. Unit vectors were not being updated
Working on NPTim
working on som integrator bugs
Starting to worry about all the strtok() calls in our code
Fixed hmat in DumpWriter (MPI) and eor.
fixed some bugs
Bunch of 1-d array -> 2-d array stuff
Bug fixing NPTi and NPTf. there is some error in the caclulation of HmatInverse.
adding in dan's NPT stuff
Fixes and merging NPTf
Fixes for both NPTf and NPTi
Fixes in NPTi migrated into NPTf
*** empty log message ***
fixed box scaling
NPTi
cleaned up the dependecy scripts in the makefiles
fixed the bugs introduced by switching the periodic box to a matrix
Fortran flexi-BOX
Fixes for flexi-BOX
working on adding the box matrix to everything.
Updated the ChangeLog, and Converted most of the SImInfo to use non-Isotropic boxes. wrapVector needs to be finished.
Changed over the bonds to Harmonic bonds in the DUFF frc file fixed constraints.
Fixes to NVT. Check them!
added Harmonic bod into the DUFF forcefield and BondExtensions.cpp
Doing some work to debug the constraint code.
NPT fix
Most of the integrator and NVT seem to be working now.
NVT additions
slowly converting to new integrator and forcefield names.
finished the basics of the integrator and SimSetup.cpp
doing some work on SimSetup to clean it up / get it to work with the new Integrator.
minor changes in an attempt to fix output times.
Added Teng's parmeters fro the ghost Bend in TraPPE_Ex some work on the integrator. ( incomplete)
added constrainA and constrainB to the Symplectic integrator
currently modifiying Symplectic to become the basic integrator.
added some member variables for position, velocity, etc.
changed how NVT is now derived from Integrator
fixed an a mismatched Ghostbend bug.
optimized the ssd calc loop
added a new test for constraint failure
fixed up sysBuild to where it should now build our systems
Added eam force files...
More eam work.
working on the system builder
Fixed ordering on NVT calculation in integrators.
added Ghost bends to the TraPPE_Ex forceField
Added first mangling of EAM.
fixed a memory bug in Fortran, where molMembershipArray was declared nLocal instead of nGlobal.
Bug fix in progress for NPT
added a globalIndex counter to Molecule
Working on ConstantStress
fixed a n mpi init bug in SimSetup. caused a miscalculation of nLocal.
fixed a bug in symplectic, where presure was only being calculated the first time through.
Added volume and enthalpy to status file
fixes for NPT and NVT
It works (kinda)...
Fixes for NPT / NVT
Moved expand neighborlist to init_FF.
fixes for NVT
fixes for nvt / npt
dt/2 fix in nvt
Fixes for affine transform
Changes to integrate the NVT and NPT ensembles
Fixes for NPT and NVT
fixed a sign error in the radial portion of SSD.
doing some testing in sticky through Symplectic.
Working on NVT
Fixed transpose bug in mpi reduce for tau and virial.
bug fixes
Many fixes to add extended system
Added ExtendedSystem infrastructure for NPT and NVT calculations
Bug fixes for simulation module rewrites
bug fixes for compilation
bug fixes to fortran wrappers
Breaking c and fortran, c gets smarter, fortran gets dumber...
fixed a memory read bug in neighborlist
Changes for Extended System
Fixes for ExtendedSystem
Added extended system header
changes for extended system code
renamed nvt to extendedsystem
added some little fixes here and there.
fixed a possible call before initialize bug.
just little things like deleteing unused variables and such.
a few fixes to simError.h also some fixes to Molecule.hpp
fixed some small things with simError.h
Starting work on NPT
Added NVT file (very broken for now)
Changed Readme, added some files
dipoles mostly work, but there is a memory leak somewhere.
Fixed a bug where MPI was not getting the proper atomIdents.
more bug fixes....
Fixed DumpWriter to be more robust to errors. also added a little namespace to InitFromFile to wrap it's helper functions in MPI
Fixed bug with pot_local not zeroed.
Fixes in MPI force calc and in Trappe_Ex parsing.
bug fix in DumpWriter.cpp
Bug fixes in read-write routines.
fixed a bug where the Excludes were not being created properly
mpi fixes and debugging mpi read write from file.
fixed long range interactions in Trappe
MPI buggy, fixed mpiRefresh issue.
fixed a bug where excludes were not being initialized
Fixed a single processor segfault bug.
fixed the compile time bugs, Source builds and links
fixed a few more bugs.
added the Molecule.cpp file
fixed the makefile
fixing some compile time bugs
I have implemeted Molecules everywhere I could remember to. will now attempt to compile.
finished updating SimSetup to initialize and use the new MPI division of labour, and Molecule class
finished conversion of TraPPE_ExFF to use Molecule
LJ_FF has been converted to the new Molecule model. TraPPE_Ex is currently being updated. SimSetups routines are writtten, but not yet called.
fixes for local->global atom numbering in MPI
little bug fixes here and there.
Fixes to fileio for MPI
bug fixes many bug fixes
Making DumpWriter less dependent on sequence of atoms on the other processors. Node 0 now fires potatoes at other processors to get them to send french fries back.
the skeleton for making the molecules is in place. ForceField needs to be updated next.
fixed a couple of the "static" bugs in Atom and Exclude
still working on the SimSetup routine. also fixed some things in Exclude.hpp
Added Atom.cpp and Exclude.cpp
Fixes in Atom and Exclude list
Make Atom.cpp able to add and delete ranges of atoms
added an Exclude class with static arrays similar to the Atom class
I'm overhauling the molecule class to contain it's own bonds, bends, and torsions. may god have mercy on my soul.
Finished globalIndex.
MPI stuff for passing out molecules
working on load balancing
Fixes for Parallel thermalization
fixed an mpi include bug in THermo.cpp
Fixed bugs with calculation of potential energy and temperature.
*** empty log message ***
electrostatic changes for dipole / RF separation
little bug fixes here and there
fixed bug where short range interactions were not being calculated. removed some debug print statements
various write statements for debugging
Fixed F_MACH_DEP bug
Makefile fixes, directory re-org, autoconf fixes
Makefile fixes for depends
This commit was generated by cvs2svn to compensate for changes in r377, which included commits to RCS files with non-trunk default branches.