Merge pull request #366 from RosettaCommons/vmullig/genkic_sampling
Vmullig/genkic sampling
Updating the ABBA.bin_params database file with new stats generated by Gaurav. Thanks, Gaurav!
This should cause integration test changes in all tests that use mainchain torsion bin transition probabilities. This includes bin_initialization, bin_perturbation, genkic_bin_perturbing, genkic_bin_sampling, and genkic_bin_setting.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #367 from RosettaCommons/kwikwag/peptiderive-talaris
Fixing some issues in PeptiDerive:
- fixed flags mix up between dump_cyclic_poses and dump_peptide_pose (PeptideDeriverPoseOutputter)
- fixed markdown report listing cyclic peptides as linear peptides
- remove redundant separators after tables in markdown
- fixed not choosing best cyclizable peptide correctly
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #364 from RosettaCommons/vmullig/valgrind_issue
Vmullig/valgrind issue
I think I've fixed the major issue causing many of my unit and integration tests to fail the Valgrind tests on the test server. It was an issue with order of initialization of variables in the Conformation assignment operator and copy constructor (I think). In the middle of the copy constructor, I was calling a function to which I was passing the new, partially-initialized Conformation object, which was problematic. This has been cleaned up. Testing now.
(This does fix all of the unit and integration tests of mine that were being flagged by Valgrind. No unit or integration test changes.)
notify author
notify list [rosetta-logs@googlegroups.com]
Update ScoreAnalyticEtable.bench.hh
Reversed reduction in number of trials--we were actually encountering a significant figure issue with these. I didn't know because before these always had at least one outlier!
notify author
notify list [rosetta-logs@googlegroups.com]