This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "MoleCuilder". The branch, Fragmentation_Automation_wo_JobMarket has been created at ef0dc6aec0daf62326d90b941f9d8b2e0cd8abfb (commit) - Log ----------------------------------------------------------------- commit ef0dc6aec0daf62326d90b941f9d8b2e0cd8abfb Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Sun Apr 27 16:36:58 2014 +0200 Removed (very) old joiner and analyzer programs. - they are not needed anymore when FragmentationAutomation works also without JobMarket now. commit 21854c7e3d111962fab06041bc7dab5f959cccee Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Apr 23 13:57:20 2014 +0200 Using MPQCCommandFragmentController in FragmentationAutomationAction. - FragmentationAction does not lament missing JobMarket and output-types. commit b84db242bb23b3f493f94c9fd568ac2d27ea1d19 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Apr 23 13:57:57 2014 +0200 Added MPQCCommandFragmentController. - MPQCCommandFragmentController mimicks JobMarket's FragmentController or SpecificFragmentController, respectively, without relying on JobMarket. - modified FragmentJob to grant MPQCCommandFragmentController access to to setId(). In JobMarket access is granted to FragmentController which we do not inherit from here. commit 7f8c9ad6477b6f1c9412bc613327680267c2d70f Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Sat Apr 26 19:19:09 2014 +0200 FIX: Need to purge FragmentationResultContainer. - otherwise molecuilder stalls on quit. - requires JobMarket includes for pyMoleCuilder. commit 786d28e4533ffcfc1d63d95f08f65c92dd946946 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Apr 23 13:24:32 2014 +0200 Lib..FragmentationAutomation and lib..Jobs now also compiled without JobMarket. - FragmentationAutomationAction, MolecularDynamics.., and ParseFragmentJobs.. are also compiled, makes sense with upcoming MPQCCommandFragmentController. - must not compile SpecificFragmentController or MPQCFragmentController without JobMarket. - now requires libboost_iostreams. - this is due to MPQCCommandJob fiddling with streams to convert results. commit 2cc7faa758cecbb1c378dc72c67f25dfd7cbd323 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Apr 23 13:25:58 2014 +0200 LibJobMarket only compiled in if CONDJOBMARKET. commit 1dfe00d53becf9875c49a1c974659affcd083a2b Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Apr 23 13:43:17 2014 +0200 Added inclusion switch on HAVE_JOBMARKET for files depending on JobMarket. - this is preparatory for using MPQCCommandJob without JobMarket. - this is possible with several basic files taken over from JobMarket project. - changes to ExportGraph_ToJobs, needs to export MPQCCommandJob when no HAVE_JOBMARKET is given. commit c3de511c88d74c7dd3ee27b035ff60e0ecf0b720 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Apr 23 13:19:14 2014 +0200 Added FragmentJob, ..Result, JobId, SystemCommandJob, and types from JobMarket. - taken over from version v1.1.4 with disclaimer adapted. - changed include paths to JobMarket -> Jobs/JobMarket. - these files are only used when no HAVE_JOBMARKET is defined. - SystemCommandJob had Info, causing extra Chronos entry. commit a498c1b15125817769fd9b27a366acd471b6b413 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Tue Apr 22 18:11:59 2014 +0200 Extracted ResultContainer from ReceiveResultContainer. commit d4eaf11f136baa85a4bdfa982418399b22a44110 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Tue Apr 22 17:59:51 2014 +0200 Renamed ResultContainer -> ReceivedResultContainer. - this belongs to SpecificFragmentController. - renamed _ResultContainer_impl -> _ReceiveResultContainer_impl. - FIX: also renamed inclusion guard. commit a42feea2484491fdb28f92c73043956e00a2d922 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Apr 24 23:56:03 2014 +0200 FIX: lib...FragmentationAutomation was linked in outside CONDJOBMARKET. commit 37ef6d9c81dd05d5433115e175bac28b62275986 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Mar 20 18:29:00 2014 +0100 Changed element class and periodentafel to have and use functions from ion. - If ionization is zero, fall back to other function. commit 31bd9c87bb06e45f77fe0e6e5b0d8ca6a9751ae7 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Mar 20 18:28:44 2014 +0100 Added ion class, derived from element. commit 9b68fc02f36ed091c9bb88104aedd4c88c66f0dd Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Mar 5 23:12:01 2014 +0100 FitParticleChargesAction now averages over all fitted partial charges. - i.e. we iterate over all fragments and not just fit the first. commit 550f2a98ea5244ef1fc747f6debbc4635e941f7c Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Mar 5 22:36:44 2014 +0100 Enhanced error messages in FitParticleChargesAction. commit e1fe7e9e5ab6e5acab3403d31d4139a02a1c1754 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Feb 27 21:15:41 2014 +0100 FunctionModel now uses list_of_arguments to split sequence of subsets of distances. - this fixes ambiguities with the set of distances: Imagine the distances within a water molecule as OH (A) and HH (B). We then may have a sequence of argument_t as AABAAB. And with the current implementation of CompoundPotential::splitUpArgumentsByModels() we would always choose the latter (and more complex) model. Hence, we make two calls to TriplePotential_Angle, instead of calls twice to PairPotential_Harmonic for A, one to PairPotential_Harmonic for B, and once to TriplePotential_Angle for AAB. - now, we new list looks like A,A,B,AAB where each tuple of distances can be uniquely associated with a specific potential. - changed signatures of EmpiricalPotential::operator(), ::derivative(), ::parameter_derivative(). This involved changing all of the current specific potentials and CompoundPotential. - TrainingData must discern between the InputVector_t (just all distances) and the FilteredInputVector_t (tuples of subsets of distances). - FunctionApproximation now has list_of_arguments_t as parameter to evaluate() and evaluate_derivative(). - DOCU: docu change in TrainingData. commit 16227a32089e461901d97290d3e2fb626919cf26 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Feb 27 22:44:28 2014 +0100 Removed FunctionModel::getSpecificExtractor() that is not needed anymore. - it was only used in FitPotentialAction generating WorstFragmentMap. - strangely required to change the order of some libraries (libMolecuilderFragmentation_Summation needed further down). commit f4496dfa7f4fdd324a537113b0d2419c2bc7496f Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Feb 27 22:26:39 2014 +0100 Extracted associating each configuration with its L2 error into TrainingData. commit 4e009d7565cb91ce80fa628407fbf1936e23d12c Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Feb 27 18:46:49 2014 +0100 Extracted function from PotentialFitPotentialAction::performCall(). - new function getNumbersFromElements(). - also added warning when take-best-of and set-threshold have both been used. commit 1ba8a154f0c3a28438bcfafdbb90182004c65f55 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Feb 27 20:07:56 2014 +0100 DOCU: Extended documentation on FunctionApproximation and potential fitting. commit 1b1fa5704cd7e88e0af6f128b9e17b69d9aac345 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Jun 25 00:16:55 2014 +0200 TESTFIX: Reduced size of loaded file for regression test InvertMolecules to speed up. commit 89604a46c1db5a85643b922ce06b5f4b5f0c5345 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Jun 25 00:14:38 2014 +0200 TESTFIX: Changed size of loaded file for unselection AllMolecules regression test. commit f678fc41737d988593634f12c8a2a6dbb76c282f Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Jun 25 00:06:02 2014 +0200 TESTFIX: Reduced box size in FillRegularGrid to speed up regression test. commit e4157b9b9699796a9b05157d81e33042b7670e3b Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Wed Jun 25 00:02:45 2014 +0200 TESTFIX: Reduced box size in DipoleCorrelation-DiscreteAngle regression test to speed up. - the histogram files have not yet been changed. commit 6d3cb81a37b7c7089163a6bbf61f4c8e6874cc81 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Tue Jun 24 23:57:37 2014 +0200 TESTFIX: reduced dbond file size in regression test Molecules/BondFile. commit 2cc1cbe98dd0605266ded20ff583bfaf6b64fb77 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Tue Jun 24 23:29:38 2014 +0200 TESTFIX: FillSphericalSurface regression now with just 20 instead of 200 fill points. commit 1dbbebbd0643e4b6a024acbe95232cacda8a7a10 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Tue Jun 24 23:41:19 2014 +0200 TESTFIX: reduced required threshold for fit-potential regression tests to speed up. commit 0932c2cf081d696ef9cc91e7a5183c13064994e0 Author: Frederik Heber <heber@ins.uni-bonn.de> Date: Thu Jun 26 09:41:59 2014 +0200 FIX: RandomNumberEngine_Encapsulation::clone() did not set the seed obtained from its prototype. - this caused all engines to always start with default seed of 1, irrespective of the set parameter. The engine is always cloned from prototypes in the factory that can be manipulated to change their parameters. Hence, clone needs to set the seed (the engine's only parameter). - EmpiricalPotentials now use RandomNumberGenerator to obtain random starting values. - TESTFIX: fit-potential regression tests now use fixed seed in order to always take the same amount of time (some test runs in debug mode take very long because of ill-chosen random values). ----------------------------------------------------------------------- hooks/post-receive -- MoleCuilder