[SCM] MoleCuilder branch Candidate_v1.4.9 created. v1.4.8-54-g0516fd9

20 Oct 2014


      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, Candidate_v1.4.9 has been created
        at  0516fd9f440d63cdfbb63b946c9933f28f304b68 (commit)

- Log -----------------------------------------------------------------
commit 0516fd9f440d63cdfbb63b946c9933f28f304b68
Merge: 3b17982 33ee996
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Sun Oct 19 19:13:32 2014 +0200

    Merge branch 'Fix_ConvexizeNonconvexEnvelope' into Candidate_v1.4.9

commit 33ee9960e3c188eaa1d7490c6b14994fbd779863
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 9 12:00:11 2014 +0200

    TESTS: Extended Tesselations tests to convex case.
    
    - moved all present tests to NonConvex folder, added same set but including
      amylose in Convex folder.

commit 565695bc8823d4df417f859cee34ed9b17e3b0b8
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 16 16:34:13 2014 +0200

    TESTFIX: Regression test Tesselation/BigConvex is now working.

commit 5f7b95f1f29d4009517dae2f5ef5ae993e846ce8
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 9 11:47:03 2014 +0200

    Convex-envelope has debug parameter DoOutputEveryStep.
    
    - this writes for each convexization step a distinct tesselation file.
    - added parameter to ConvexizeNonconvexEnvelope().

commit bcb83114d92eb11cccb67975f4967b7fb251de3d
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Sep 25 18:24:30 2014 +0200

    Added SphereRadius parameter to ConvexEnvelopeAction.

commit 991c35d557cd7eb1196fd569a74860662f03eeb6
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 9 11:34:55 2014 +0200

    Added tcl scripts for viewing and debugging convexized surfaces.

commit 8544b32bd227c145f76c3bf5a30ca16c01e9a7d9
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 9 10:55:44 2014 +0200

    DOCU: Added extensive explanation how convexization works.
    
    - updated userguide for convex-envelope.

commit e2396e6d7030fd404f25afa9b5e8d1e1bc72bdb8
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Sat Oct 4 10:44:54 2014 +0200

    Added check whether all points remain inside surface and whether it is convex to ConvexEnvelopeAction.

commit aac19f2bb7cdef195d60c2e090de27fd193f1ee9
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 9 10:27:36 2014 +0200

    ConvexizeNonconvexEnvelope() uses now RemoveFullConcavePointFromTesselatedSurface().
    
    - with this convexization of a nonconvex tesselated surface seems to be
      robustly working as we can now deal also with points with degeneracies.

commit 2ef2cc3b3e76c8ecb94ddaddfb1203081ea07eb0
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Wed Oct 8 18:14:39 2014 +0200

    FIX: ConvexizeNonconvexEnvelope() needs to purge degenerate triangles before convexization.

commit 81ffd87c0c1926d2a925cdbb620380e16e421cbe
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Wed Oct 8 15:08:11 2014 +0200

    FIX: GainMap is used to store the concavity angle, we flip now always most concave.
    
    - this overrules prior flipping those that gives greatest gain in volume.

commit d289c64b86bc28a435d03cdf4f3ece926c6debf6
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Sat Oct 4 16:34:41 2014 +0200

    Rewrote ConvexizeNonconvexEnvelope().
    
    - now works for some sphere radii on amylose (e.g. 40 and 50).
    - we first remove all points that are only attached to concave lines.
    - also, we always flip the baseline (i.e. add a general tetraeder to the
      non-convex volume) for concave lines when the tetraeder does not intersect
      with present triangles.

commit 6da9e995abeca059d65efd473cd2ac3eacf8279d
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 9 12:31:18 2014 +0200

    FIX: Tesselation::RemoveDegeneratedTriangles() ended up in endless loop.
    
    - even if a triangle pair is not removed, it must still be removed from the
      degenerated list or we run into infinite loops.

commit 30ce7d21a0bc0d344058826d689837a313b01efd
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 9 09:37:06 2014 +0200

    Extracted RemovePointSurroundedByPolygon() out of Tesselation::RemovePointFromTesselatedSurface()
    
    - also added new functions
      Tesselation::RemoveFullConcavePointFromTesselatedSurface() and
      Tesselation::CheckAllConcaveInPolygon().

commit 77eec57c5ebcc041fb7d83112c887c95824f977d
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Tue Oct 7 09:38:17 2014 +0200

    FIX: Removed flipping for maximal gain in Tesselation::RemovePointFromTesselatedSurface().

commit e26c3533deedb0298b184037987b587e4d718562
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Wed Oct 8 14:46:36 2014 +0200

    FIX: Tesselation::RemovePointFromTesselatedSurface() uses wrong NormalVector.
    
    - with the point to be removed we have a very good measure for a vector
      pointing inside. Before we used the average of all normal vectors.

commit d45a5bfcf5b03614dc189256f1861f31e819f103
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Tue Oct 7 09:55:26 2014 +0200

    FIX: Tesselation::getVolumeOfConvexEnvelope() calculated volume wrong.
    
    - replaced code by using present function for tetrahedron calculation with
      center of gravity as fourth point when going over all boundary triangles.
    - TESTFIX: volume change in convex regression test.

commit 32d7e4e99ad240050c6936e8be2d5450c409f996
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Tue Oct 7 08:57:30 2014 +0200

    FIX: Tesselation::FlipBaseline() caclculated NormalVectors wrong.
    
    - before we just took the average of the normal vector of both present
      triangles as guide. Now, we check the intersection of the fourth endpoint
      on the plane of the first triangle and decide by on which side of the
      three boundary lines the intersection lies whether to flip the normal
      vectors sign for the associated (new) triangle or not.

commit d33bb8a2c2523a42b0a9e36768e551c57c340cbc
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Tue Oct 14 18:37:03 2014 +0200

    Added Tesselation::IsPointBelowSurroundingPolygon().
    
    - IsPointBelowSurroundingPolygon() creates all triangles as planes and checks
      whether point is below.

commit f1ae15bf748d2c4746bb9ed901593b75d8a44b1f
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Fri Oct 3 12:27:57 2014 +0200

    FIX: Filling of non-convex polygon with only minimal angles did not work.
    
    - when removing points from non-convex boundary and we encounter non-convex
      polygons, we must pick smallest angles but close to the non-convex site!

commit c2c786e40198b3b955ca9b66593dd013bd531b2e
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Fri Oct 3 11:43:03 2014 +0200

    FIX: RemovePointFromTesselatedSurface() used Start,Middle,EndNode confusingly.
    
    - FIX: Very last node was removed twice due to StartNode = EndNode setting.

commit 6d52807f7b413a18b8d814ac5501fdce735fcdea
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Fri Oct 3 11:38:16 2014 +0200

    FIX: Reduced verbosity noise from Line::getIntersection().
    
    - removed use of info and toned down verbosity levels (which are so low
      because of now missing info).

commit b8d21539ab1899769efb705f4de59ef647f234b1
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Sep 25 20:26:03 2014 +0200

    Changed some verbosities in Tesselation code.
    
    - modifications in BoundaryPointSet, BoundaryLineSet, BoundaryTriangleSet.
    - also to CandidateForTesselation, TriangleIntersectionList, tesselation,
      tesselation helpers and boundary code.

commit dfe0549d4ce8433fde1acf2f7886d20fc1f0174a
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Tue Oct 7 09:28:44 2014 +0200

    DOCUFIX: Corrected formula for general tetrahedron in docu.
    
    - also code now matches with formula to avoid further confusion, although
      the result does not change by that.

commit 75d42ad5b96258a67afde9aa02cf39b04a4f6059
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Sep 25 20:23:50 2014 +0200

    FIX: RemovePointFromTesselatedSurface() needs to pick smallest angle when polygon is not convex.
    
    - AddTesselationLine() now returns whether new or present line was used.
    - FIX: Last inner line in closed polygon has been added by one but last
      triangle already, we added it twice.

commit c2f30642d960bd67dc5b7bca66d267525a8f58ab
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Sep 25 20:24:52 2014 +0200

    FIX: RemoveTesselationLine() needs to erase line from OpenLines if present.

commit bb86bd850f990066e79d37a776337c627624df65
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Sep 25 18:25:23 2014 +0200

    FIX: AddTesselationLine() would only work in initial tesselation phase.
    
    - the CandidateForTesselation instance is NULL when we convexize and this
      was not correctly processed by AddTesselationLine().

commit 613790861011df25a40877f5c6bfb2156def7a8c
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Sep 25 20:26:20 2014 +0200

    FIX: Changed all performCriticalExit() into ASSERT() in tesselation.

commit cead3fe0679f2e6b3b53a7300f42efde4b3cc552
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Oct 9 10:56:24 2014 +0200

    FIX: Raster3D needs header.r3d, which had been missing so far.

commit cfb9c59bbf7a92a53173e5044b42303e74361093
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Tue Sep 23 18:14:18 2014 +0200

    FIX: Setting ActionQueue::_lastchangedaction without heeding whether Action failed is bad.
    
    - causes segfault when notification informs about change and _lastchangeaction
      is still NULL (as the very first Action failed).
    - we now only notify when action succeeded.

commit 1b6cec8a7217342cc34f00da7084bcec75e948a3
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Mon Sep 22 22:10:11 2014 +0200

    FIX: QSeisPlotPage's dstor calls deinitializeLayout() only when initialized.

commit 7d320c29fb3cf5f18317f1e2370bbf89802448e4
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Mon Sep 29 16:40:10 2014 +0200

    MEMFIX: Static coordinator instances of specificv potentials need to be Memory::ignore()'d.

commit 4da4defddffa2503040fe329681f55b1f6f8c856
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Mon Sep 29 16:35:04 2014 +0200

    FIX: Added Boost::MultiIndexContainer check for MemDebug.
    
    - this requires CodePatterns 1.2.10 and allows finally enable-debug compilation
      with Ubuntu 14.04.

commit 3b179825d4e6f17bd22fbd7f880f59cc7fd8924e
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Fri Sep 19 13:19:18 2014 +0200

    FIX: QtHomologyList's plot widget was too high (and fixed).
    
    - also the sizes of splitter's children in QtHomologyList are now stored to
      settings (as QsizePolicy does not have an effect on a splitter).

commit 5c8807e908af52740778a67bac4173280b6d141b
Merge: b119a09 c7fe90c
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Sun Oct 19 19:12:11 2014 +0200

    Merge branch 'Experiment_Amylose_Water' into Candidate_v1.4.9
    
    Conflicts:
    	doc/userguide/userguide.xml
    	tests/regression/Atoms/testsuite-atoms.at
    
    - mirror-atoms and translate-to-origins occupied the same places.

commit b119a092aa1848936f8c80974f857ded3c098885
Merge: 9e1bfbc 46ce1c8
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Sun Oct 19 19:10:16 2014 +0200

    Merge branch 'DynamicMolecules' into Candidate_v1.4.9

commit 9e1bfbc27ee42cdb562491c7e51fdb9da26faaf0
Merge: 2ed3bf3 0ea063b
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Sun Oct 19 19:09:30 2014 +0200

    Merge branch 'Refactoring_PotentialActions' into Candidate_v1.4.9

commit 0ea063beb1f1de6945df69f74ce759aab7643539
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Fri Sep 12 09:54:12 2014 +0200

    Extracted SavePotentialsAction.
    
    - PotentialTrainer::StorePotentials removed.
    - Fit...Actions do not store anymore.
    - TESTFIX: potential regression tests now use save-potentials and compare
      against stored file.

commit f1eabd03d00ab44fb84ebc758d0ebfc85d6dee6d
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Fri Sep 12 09:39:58 2014 +0200

    Extracted ParsePotentialsAction from FitCompoundPotentialAction.

commit 98d166e8fbc83bccb5843ecc5d4296f2c25488c0
Author: Frederik Heber <heber@ins.uni-bonn.de>
Date:   Thu Sep 11 22:38:05 2014 +0200

    Split FitPotentialAction into two actions.
    
    - FitPotentialAction fits just a single potential.
    - new FitCompoundPotentialAction fits via a potential file.
    - PotentialTrainer contains all helper functions and common code and is only
      compiled when CONDLEVMAR.

-----------------------------------------------------------------------


hooks/post-receive
-- 
MoleCuilder

git＠ins.uni-bonn.de

tags

participants (1)