Merge pull request #858 from RosettaCommons/vmullig/tweak_helixfit
Adding an error message in a case of a cryptic failure.
Adding a proper error message when the ResidueTypeSet tries to create a residue type for which no params file exists.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #795 from RosettaCommons/jadolfbr/mpi_dbout
This merge fixes 2 FeaturesReporter MPI Issues:
Bugs Fixed
---------------
1) Allows the use of the command-line flag for separate_db_per_mpi_process to work in with the XML.
2) Fixes a nasty bug that was sending the MPI JOB_FAILURE_TAG instead of the REQUEST_MESSAGE_TAG when requesting batch and protocol ids for MPI. This fixes sqlite3 MPI and may fix MySQL MPI if it was broken as well.
Code and Test
-------------------
Reorganized basic/message_passing into basic/mpi and moved enums out of MPIJD and into this namespace. This allows use of the enums instead of ints, which should help prevent something like this in the future (Thanks Vikram!)
Add MPI integration test for this (Again, thanks Vikram!).
Expected Test Failures
-------
Some cosmetic changes in integration tests that used the message_passing namespace. MPI integration test will fail for being new.
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #832 from RosettaCommons/JWLabonte/sugars/scoring
Carbohydrates: sugar_bb scoring term completion
All unit tests pass.
carbohydrate and vancomycin test changes expected, because of improvements to sugar_bb score term>
c-terminal_conjugation test changes expected, because I corrected a wonky pdb input file.
notify author
notify list [rosetta-logs@googlegroups.com]
15/10/13 updated the Xcode template project.pbxproj.template file to remove links to protocols_f.5 and add libz to protocols_a.5; current master now builds correctly in Xcode using this new template file. Because the only change is this template file, I am pushing this file to master without a pull request
notify author
notify list [rosetta-logs@googlegroups.com]
Merge pull request #840 from RosettaCommons/vmullig/asford_multibody_copy3
Add support for non-pairwise decomposible score terms to the packer
This merges Alex Ford's recent changes to the packer into master, adding support for score terms that are not pairwise decomposible but which can be made fast to compute or update by other means. (An example is the aa_composition score term, which imposes a nonlinear penalty for deviation from a desired amino acid composition. It's very fast to count amino acids of a given type.)
Tasks:
- Fix issues introduced by recent changes to disulfide code.
- Add full Doxygen documentation for new classes.
- Re-beautify.
- Check the (many) integration test changes.
- Fix PyRosetta issues.
- Re-beautify.
- Fix failing unit test.
- Rename variables with leading underscore to have trailing underscore instead.
- Figure out PyRosetta unit test issues.
- Re-beautify.
For a future pull request:
- Add RosettaScripts-accessible means of setting up the new score terms.
- Add a better base class for packer-compatible non-pairwise-decomposible score terms.
- Revise the aa_composition energy, adding more user controls (e.g. layers, regions of the protein, etc.)
Integration test changes expected:
- Many tests have small cosmetic changes (because some packer-related namespaces and corresponding tracer names change).
- bundlegridsampler_composition_energy actually changes (because the aa_composition energy now works with the packer, so the output from this test is much better)
notify author
notify list [rosetta-logs@googlegroups.com]
Remove protocols_f.5 library from devel unit test src.settings (we still have that?) and core. This actually does what my previous commit should have done.
notify author
notify list [rosetta-logs@googlegroups.com]