[med-svn] [SCM] ray annotated tag, v2.1.0-rc0, created. v2.1.0-rc0
Sébastien Boisvert
sebastien.boisvert.3 at ulaval.ca
Sat Nov 3 20:44:40 UTC 2012
The annotated tag, v2.1.0-rc0 has been created
at 9425959e55f65d265dcdcb0487261a5a4ca95279 (tag)
tagging 80c83ca3fcbdcb41a2cd049981d0e13c96fb0022 (commit)
replaces v1.6.1-rc1
tagged by Sébastien Boisvert
on Thu Oct 4 11:48:49 2012 -0400
- Shortlog ------------------------------------------------------------
v2.1.0-rc0
David Eccles (gringer) (1):
fixing a segfault when no contigs are found
Eric Normandeau - Bioinformatics (8):
Updated script
Updated script
Updated script
Updated script
Updated script
Updated script
Updated script
Updated script
Pier-Luc Plante (3):
Scaffolder is not required when using unpaired reads.
Patch Koala: Added an option (-use-maximum-seed-coverage) so that higly-covered seeds can be ignored.
Corrected the tet that determines the quality control results.
Sébastien Boisvert (1073):
Added the processor name for debugging purposes.
Added option -show-memory-allocations
Added an indicator for completion.
Changed ifdef linux for ifdef __linux__
* Fixed the showMemoryUsage function.
Added a unit test for uniformity.
Added a unit test for uniformity of vertices on ranks.
Fixed a bug for the seed length distribution written to a file.
Fixed a bug for the seed length distribution written to a file.
Changed MAXKMERLENGTH=128 for MAXKMERLENGTH=64 in examples.
Changed MAXKMERLENGTH=128 for MAXKMERLENGTH=64 in examples.
Moved the allocator in its own file.
* Fixed a communication problem for k>32.
Now Ray writes observations for libraries in a file.
Added (some) code documentation.
Optimised run()
Added balance for the profiler.
* Added a progression indicator for scaffolding.
Merge branch 'master' of git at github.com:sebhtml/ray
Modified the code to reduce the number of calls to fopen.
Modified the code that writes scaffolds in FASTA format.
To use 454 mate-pairs in an SFF file, you must extract them
Fixed a compilation error for BzReader.
Added some comments in the code.
New Ray option: -write-kmers write k-mers, coverage values and arcs to PREFIX.kmers.txt
Adding experimental support for color-space reads.
* New experimental option: -color-space for .csfasta files (unstable)
* Added an assembler panic when no k-mers are found in reads.
* Ray can generate an assembly in color space (the resulting assembly is in color space).
Added a small comment to document call_RAY_MPI_TAG_ASK_VERTEX_PATHS.
Corrected a typo in the eagerness explanation.
Documented color-space mode.
Edited unit tests to consider changes to the code base.
Added DEBUG=n as default in the Makefile.
An additional round of smoothing is now the default.
-write-kmers don't write k-mers with coverage of 1
Fixed: Sending RAY_MPI_TAG_GET_CONTIG_CHUNK messages fails
Fixed smoothing bug in CoverageDistribution.
* New options: -minimumCoverage, -peakCoverage, -repeatCoverage -- provides coverage values automatically.
Fixed a bug in a script that generate Ray commands
Corrected script.
New C script.
Created a program to filter qseq files.
filter-qseq.c is working properly now.
A read can actually be empty after trimming.
Merge branch 'master' of git at github.com:sebhtml/ray
KmerAcademy is a place where all k-mer must transit before making it to the GridTable. Those observed only once remain in the KmerAcademy. This academy is then destroyed.
Memory for low-coverage k-mers (the KmerAcademy) is freed after graph construction.
The memory usage for optimal read markers is now reported with -show-memory-usage
Merged VertexData with Vertex and merged VertexTable with GridTable
* Reduced the memory usage during the building of the graph.
Added DynamicVector to reduce memory fragmentation
Fixed a free problem in SeedExtender.cpp when no seeds are available.
Added parenthesis for clarity
Removed segmentation fault
Edges pointing to non-existing vertices are now purged
VirtualCommunicator statistics are now reported for BufferedData
Reinstated code for -write-kmers
New outputted file: PREFIX.degreeDistribution.txt contains the distribution of degrees in the graph.
Cleaned the code by removing MemoryConsumptionReducer (~ -1000 SLOC).
Fixed a bug in KmerAcademy
The hashing functions now only take an object, the reverse-complement logic is outside.
The hash tables are now implemented using open addressing with quadratic probing.
Now using MyHashTable for GridTable
Implemented a sparse hash map (following the description of Google sparse hash; see structures/MyHashTable.h)
Increased the maximum number of reads per message-passing interface rank in Ray
Improved the performance of MyHashTable (running time)
Removed data type attribute in Message because it is always 64-bit integers anyway.
Added statistics for the new hash table.
Added some documentation in the source code of Ray.
In MyHashTable, the bitmap is now static instead of dynamically allocated. Furthermore, a group of buckets now contains 64 buckets instead of 48.
Changes: MyHashTable now automatically grows and uses double hashing for open addressing.
Fixed a compilation error for MAXKMERLENGTH>32
Removed a useless prototype.
Fixed compilation errors for unit tests.
Fixed a boundary issue with the second hashing function
Now, only the MessagesHandler is aware of the message-passing stack (mpi.h and MPI_*). The other folks use the inbox and the outbox to communicate -- or the virtual communicator if they are intelligent.
Fixed an infinite loop when a file contains no sequences at all.
Investigating possible memory fragmentation -- added printing of debug information.
Renamed a directory
Updated unit tests.
Created a framework to launch systems tests in parallel on a compute grid
printStatistics should be called on the k-mer academy, not on the grid table.
Merge branch 'master' of git at github.com:sebhtml/ray
Continued work for the system tests. Mostly script editing.
Fixed a segmentation fault when one rank has no seeds.
System tests can now be done with Sun Grid Engine. Entry point is system-tests/main.sh
Removed examples directory -- system tests now replace them.
A system test can be run outside of a compute grid (cloud) using run-test.sh.
New output file: PREFIX.MessagePassingInterface.txt contains the number of sent messages for each type.
Implemented SmartPointer objects to allow defragmentation of memory.
Allocating a number of elements, not a number of bytes & fixed a bug for a SmartPointer after deallocate
Changed the default number of processors for system tests to 30
Fixed a bug that reported 0 sequence reads for the first file while otherwise handling everything correctly.
Unstable version of an allocator with real-time defragmentation.
Removed memmove.
Added documentation for getIngoingEdges and getOutgoingEdges
Fixed a buffer overflow in getKmerAtPosition, reported by David Eccles (gringer).
Restored the correct number of cores
Removed some output in print().
Fixed a bug in deallocate and added defragment method.
Added defragment calls when a core is waiting, testing it.
Modified defragment to be coherent with the low-granularity of Ray.
Renamed the Sun Grid Engine runner script for system tests and added the script for testing on a symmetric multiprocessing processing machine.
Edited an error message to make it more explicit.
Added a method to obtain the allocated size for a SmallSmartPointer.
Added a method to obtain the allocated size for a SmartPointer.
Implemented incremental resizing to allow low latency.
When all message-passing interface ranks are done computing k-mers (or vertices), the incremental resizing is completed if necessary.
Changed the default number of buckets for MyHashTable and changed parameters for incremental resizing.
Added m_firstGapStart to speed up the lookup.
Cleaned scripts for Sun Grid Engine
Fixed compilation warnings.
Changed the rules for allocation, now more real-time.
Working on the speed.
Fixed a bug that I introduced in ChunkAllocatorWithDefragmentation
defragment() now is O(65536).
Added DefragmentationLane class.
Fixed a compilation warning.
The allocation process is now a little bit faster -- only defragmenting when there are enough gaps.
Fixed a bug (KmerAcademy::insert) that effectively doubled the memory usage of the k-mer academy.
Fixed main-sge.sh.
Added a to-do task.
Added a task to do.
Documented the main loop of Ray.
Added a to-do task, I think there is memory fragmentation during the selection of optimal read markers.
Documented insert().
Added time complexity.
Changed maximum execution time.
Fixed 2 segmentation fault bugs in this newly crafted memory allocator.
Corrected 2 dead symbolic links.
Fixed a display bug.
Added a new output file called PREFIX.NetworkTest.txt which contains average latency including software overhead.
By default, don't pack structures in memory (FORCE_PACKING=n).
Fixed a display problem.
Removed a displayed message.
Fixed a problem that caused contigs within scaffolds to be 2 times shorter when MAXKMERLENGTH=2.
Fixed the computation of the destination for the network test.
New thing to do.
Added -O3 when running the profiler.
Removed calls to defragment() outside of DefragmentationGroup.
Added a message indicating that network testing is to be done.
Using cout instead of printf to print uint64_t values.
deallocation is faster, so the overall process is faster.
Fixed a compilation warning.
Added host name in network test message.
Added a system test for the Illumina MiSeq.
Renamed MiSeq system test.
Updated version to 1.6.1-rc3
Updated R scripts.
Fixed the range.
Pacific Biosciences long reads now load correctly into Ray systems.
Fixed an R script for the plotting boundaries.
Moved symbolic links.
Corrected the number of pairs.
Updated the change log, version 1.6.1 should be released today.
Added the symbolic link for data-for-system-tests.
Ray v1.6.1 is code named 'sunray'.
Using png instead of pdf.
Now using png instead of pdf.
Added log10 scale to library distribution plots.
Changed a constant in a R script that controls the boundaries.
Moved this script because it interferes with system tests.
Added a working prototype in Python for calling library peaks.
Merge branch 'master' of github.com:sebhtml/ray
Library caller prototype works well on all 3 Assemblathon 2 datasets without worrying about persky adaptors, thanks to Optimal Read Markers (TM).
Fixed the command in a system test.
Now working on v1.6.2
Added an entry in the change log for 1.6.2.
Ported the prototype for finding peaks from Python to C++.
Ray can find more than one peak in any paired library.
Now selecting the correct peak to choose the next vertex.
Choosing the good peak for a paired library if a mate is already available.
Now Ray uses the maximum peak of a paired library to compute the expiry position of a read.
Fixed some compilation warning with gcc 4.1.2.
Fixed a system test.
Don't set NSLOTS if already defined.
Changed the behavior for repeats.
Merge branch 'master' of git at github.com:sebhtml/ray
Corrected the peak finder.
Changed to 64 slots.
Don't print the tree.
When the extension is finished, show library peak usage.
Don't compute or update peaks for libraries with manually-provided information.
Added a coding style file.
Updated the coding style.
Corrected a code comment.
Moved the configuration of the virtual communicator in Machine.cpp
Fixed a compilation error.
Implemented parallel file partitioning.
File partitioning is now performed in parallel.
Improved the peak finder when there is no deviation.
Created an heuristics module and moved related bits in it.
New experimental heuristics: The Ray NovaEngine.
Added a unit test for the NovaEngine.
Improved the NovaEngine, but not using it yet. It needs more testing.
Removed assertion.
Removed some messages.
Improved the unit tests for the NovaEngine.
Improved the NovaEngine according to unit tests.
Simplified the algorithm that finds peak and added 35 unit tests to test it (for various datasets).
Added 35 unit tests.
Added a symbolic link.
Added an entry in the changelog.
Moved Kmer routines in the class Kmer.
Fixed a typo.
Unit tests are now files named test_<test_name>.sh
Added a file describing how to submit a patch.
Improved the document about patching.
Only show nova choices if -show-extension-choice is provided.
Improved the README for system tests.
Added an abstraction layer for the operating system.
Added a TODO item.
* Added information on unknown nucleotides in the instruction manual.
Added a TODO item.
Added a unit test and modified CoverageDistribution.cpp to handle low-coverage datasets.
Removed roughly half of the messages with the MPI tag RAY_MPI_TAG_KMER_ACADEMY_DATA.
Fixed a compilation warning.
Disabling by default the experimental NovaEngine. Results so far are promising !
Added 2 scripts for code editing.
Added comments at random places.
Removed options in 2 system tests.
Fixed a bug in the recently introduced peer-to-peer parallel Partitioner.
Option use:NovaEngine enables experimental NovaEngine.
Removed by default the reporting of libraries in stdout.
Fixed a bug in the NovaEngine.
Now using the NovaEngine.
New output files: PREFIX.SequencePartition.txt and PREFIX.NumberOfSequences.
Fixed compilation errors with HAVE_LIBZ=y and HAVE_LIBBZ2=y
Added options -use-NovaEngine and -show-NovaEngine for debugging purposes.
Only use NovaEngine when paired information is available.
New option: -write-seeds which is useful for debugging the code.
Added some unit tests for the Ray NovaEngine (for mate-pair reads). Seems to work quite well so far.
Fixed a compilation warning.
Added a section on how to launch Ray in the manual.
Improved the manual.
Added a unit test and fixed NovaEngine to handle it.
Modified the extension algorithm to avoid collapsing of repeated k-mers that are near each other in the genome.
Added read placement freezing.
Merging of similar paths has been modified.
Fixed an implementation bug for double hashing.
Fixed comments and assertions for new correct code.
Added a test for open addressing.
Fixed a bug in the incremental resizing algorithm of MyHashTable.
Merge branch 'master' of github.com:sebhtml/ray
New development option: -show-distance-summary.
Added debugging option -show-distance-summary.
Added David Eccles in the README
Removed email of contributor.
Modified the selection engine to that the new unit tests also pass.
Removed the coverage threshold from the algorithm that finds seeds in the distributed graph.
Added N50, median, average and largest contig and scaffold lengths in PREFIX.OutputNumbers.txt
Fixed a compilation errors due to the algorithm library.
Fixed the maximum length of input reads to 65535.
Updates in the README.
Fixed a recently introduced regression in heuristics (should not choose an invalid choice).
Added a few things in the README.
Implemented a Bloom filter to reduce the memory usage. This is ridiculously good and
Fixed a memory problem in the computation of optimal read markers.
ExtensionElement objects now contains reads in 2-bit format.
Fixed a compilation warning with Intel compiler.
Fixed a compilation warning with gcc.
Fixed a integer overflow.
Fixed a compilation Warning.
Flag invalid choices before doing the actual selection.
The paired read simulator is now a separate project, see https://github.com/sebhtml/paired-read-simulator
Added list of working C++ compilers.
Changed the default number of persistent requests.
Modified NovaEngine to pass a new unit test (as well as the old unit tests).
Extension of seeds if done endlessly until growth stops.
Fixed a bug.
Added some documentation files.
Input opcodes are now shuffled before being utilised.
Fixed which arguments are picked up by opcodes -p and -i.
Changed the precision of things that go together.
Fixed a bug in the scaffolder, now more vertices should be investigated.
Fixed a segmentation fault that occurs in rare cases.
Ray merger will merge more things now.
RayCommands file is written correctly now.
Fixed N50 when there is only 1 scaffold.
Added skipping events.
Don't fetch read markers when not needed, use less memory to know is a vertex was assembled.
Only send a RAY_MPI_TAG_ASK_IS_ASSEMBLED message if starting on a seed on flow 1.
Changed the prototype of VirtualCommunicator::getMessageResponseElements.
Modified the order of the steps performed when merging identical paths.
Fixed a display problem.
Added tag counts for option -run-profiler.
Reduced the number of messages with tag RAY_MPI_TAG_REQUEST_VERTEX_COVERAGE in SeedExtender.cpp
Preparing code for a change.
Added options -read-checkpoints and -write-checkpoints, this is still in development.
Option -write-checkpoints writes all checkpoints.
-read-checkpoints works with checkpoints <CoverageDistribution>, <GenomeGraph> and <Seeds>.
Checkpoints are now operational.
Checkpoint files are now written in a binary format.
Added a MANUAL file.
Fixed a messaging bug occuring very rarely.
Improved checkpointing messages.
Added option -test-network-only to only to test the network and return.
Improved checkpointing message.
Fixed a bug when no sequence files are provided.
Added checkpoint Partition.
Added checkpoint Extensions.
Fixed a hanging problem.
Changed 1 hash function because it was a copy.
Read checkpoint before writing it.
Added a file describing checkpoints.
Don't write a checkpoint if it was just read.
Limiting seeds to probably unique vertices.
Skip a seed if within it during flow 1 and a vertex is already processed.
Removed unused scripts.
Added gmane link in the README
Added -read-write-checkpoints in the changes.
Updated the MANUAL.
Added option -debug-fusions.
Updated the ouput of -help option.
Added checkpoint Sequences.
Added an explicit flush.
Restored original state.
Added additional debugging information.
Testing symbolic links.
Removed logo from source.
Added a Documentation directory.
Removed the manual target from the Makefile.
Changed where is written the binary Ray.
Added a function to create directories.
Fixed compilation warnings.
Added documentation for the virtual communicator.
Added documentation for the virtual processor.
Added documentation for the network latency.
Fixed an argument name.
Fixed compilation warnings for 32-bit systems.
Introducing the VirtualProcessor class.
Removed MyForest and its iterator minion.
Added VirtualProcessor initialization.
Updated Documentation files.
Added \author tag to all classes.
Migrated some code only utilised by the scaffolder.
Added INSTALL.txt.
Adding new files in Documentation/.
Merge branch 'master' of github.com:sebhtml/ray
Fixed compilation warning.
reset() must be called in the constructor.
Fixed a compilation warning.
Updated the path to Ray in system tests.
Modified the behavior of Ray when fusions are generated.
Added option -version to Ray.
The ChangeLog file will not be maintained anymore, use ./scripts/dump-ChangeLog.sh
Added scaffolder cases in Documentation/
Merge branch 'master' of git at github.com:sebhtml/ray
Changed the maximum number of cycles to 16 in merging code.
Removed hard-coded parameter -debug-fusions.
Added TaskCreator and Merger classes.
Added debugging messages in FusionData.
Added method hasWorkToDo to VirtualProcessor.
Added interface Worker for worker classes.
Restored worker codes.
Added some debugging information for fusions.
Using the VirtualProcessor for edge purge.
Fixed some unit tests by moving scaffolder methods.
Implemented a new better and simpler merger module -- FusionTaskCreator/FusionWorker.
Removed OperatingSystem dependency in unit tests.
Added some debugging information for FusionTaskCreator.
Changed the default algorithm in VirtualProcessor -- now using a minimum work unit.
Added an AUTHORS file.
Fixed a machine-state bug in FusionWorker.
Fixed a state-machine bug in TaskCreator/FusionTaskCreator.
Workers push virtual messages, not real messages.
Disabled the reverse-complement copies of extensions.
Added Joiner code.
Added debugging information in JoinerWorker.
Updated a threshold in FusionWorker.
Added selected hit in standard output for JoinerWorker.
Now printing hit information in JoinerWorker.
Joiner software stack now joins otherwise un-joined paths in the distributed graph.
All output files are written in a directory provided with option -o.
The default is now ASSERT=y in the Makefile.
Corrected positions in JoinerWorker when on the other strand.
Added 3 unit tests for NovaEngine and improved the heuristics.
Implemented the reverse strand case in JoinerWorker.
Improved heuristics for selection.
Corrected the number of flowed vertices in the seed extension.
Fixed a bug in the virtual processor wherein it was not force-flushing messages when needed.
Cleaned some code in the task creator routines for edge purging.
Cleaned some code in the task creator routines for edge purging.
Changed --oneline to --pretty=oneline for compatibility with older versions of git.
Modified the Sun Grid Engine job template to erase the directory before running the whole thing.
Fixed the content of a displayed text in the fusion task creator.
Modified the scaffolder routines to check the vertex coverage values in paths.
Removed the minimum number of raw scaffolding links.
The checkpoint ContigPaths is now written on demand.
The checkpoint ContigPaths is now fully operational (read and write).
Added option -write-contig-paths to write contig paths with coverage values. This is enabled by default.
Fixed the code that counts the number of extended seeds.
RayVersion and RayCommand are now written (a bug was introduced).
Limiting scaffolding links to vertices that have one parent, one child and a coverage value near the peak.
Changed the default message size for network testing.
Added information in ScaffoldLinks.txt
Added non-persistent MPI communication just to compare.
Added the standard deviation in ScaffoldLinks.txt
Implemented a new greedy scaffolding algorithm as discussed with François Laviolette.
Modified some scaffolding code to obtain the correct side of a contig when it allows both.
Restored the default number of words in the network test to 500.
Added some documentation for Infiniband.
Updated for ScaffoldLinks.txt format to v2.0.
Added more debugging information in the scaffolding test.
Fixed scripts to accomodate new prefix directory option.
Fixed an integer overflow in the computation of standard deviations.
Merge branch 'master' of git at github.com:sebhtml/ray
The number of enabled MPI ranks can be changed during the network test by changing a variable in the source code.
Fixed a communication problem in MessageProcessor.
Added some debugging information for -debug-fusions.
Fixed a bug in JoinerWorker in which two overlapping paths would not be joined together.
Regression bug on phix system test fixed.
Fixed some divisions by 0 in the scaffolder.
Added assertions and fixed a bug in GridTable.
Added option -debug-scaffolder.
Removed dependency for clock_gettime.
Added more details in the output of -run-profiler.
Added option -show-read-placement.
Added option -show-communication-events.
Added overlays for option -show-communication-events.
Added a communication optimizer with urgent messages.
The option -show-communication-events now shows all messages with overlays too.
Enabled the communication optimizer for the network test too.
Fixed a bug in the code that loads the checkpoint GenomeGraph.
Added a time period during which other messages are more important than urgent messages.
Added option -write-network-test-raw-data.
Write raw data for network tests if -test-network-only is provided.
Implemented a round-robin algorithm for the reception of messages.
Merged the persistent communication layer with the round-robin reception.
Changed mpirun to mpiexec as mpiexec is in the standard.
Removed inline code because compilers optimize the code anyway.
Disabled persistent communication in the round-robin reception.
Added comments in the communication layer of Ray.
Added timer warnings with -run-profiler.
Reduced the computation granularity for the code that computes reverse-complement extensions.
Reduced the granularity in call_RAY_SLAVE_MODE_EXTENSION() by cleaning expiration positions.
Remove expired reads from the list of unmated reads to reduce the computation granularity.
Added the compilation option CONFIG_CLOCK_GETTIME for the profiler.
Added granularity summary for option -run-profiler.
Migrated the version in the Makefile.
Fixed some PATH issues in system tests.
Simplified the release procedure.
Fixed PATH problems in unit tests.
Updated the manual page.
Removed data files in unit tests.
Added option -write-read-markers.
Added option -show-consensus to display read sequences and consensus during the extension.
Added option -write-marker-summary.
New option -write-marker-summary which writes marker files for further inspection.
Added some profiling points inside the source code of Ray.
Added a new output file called ElapsedTime.txt.
Added some code to store agreement values during read threading.
Refactored some code to ease further enhancements.
Corrected the accuracy of expiry positions.
Added memory usage reporting in the fusion step.
Some code refactoring for the extension of seeds.
Improved the algorithm that compute seeds, less (none) assembly errors
Added the Open-MPI option -output-filename in system tests.
Changed the threshold for the repeated vertices.
Removed some debugging messages in the creator of fusion tasks.
Merge branch 'master' of git at github.com:sebhtml/ray
Added a working prototype router for message routing.
Don't append message statistics when the count is null.
Added automatic route generation.
Added option -cores-per-node to help creating better routes with option -route-messages.
Added option -connection-type to specify the type of connections to use with -route-messages.
Added some type definitions and enhanced the routing algorithms.
Assertions are not compiled by default now. (ASSERT=n)
Added relay event counting to avoid hanging code with routes generated
The route from B to A is now computed independently from the route from A to B.
Routes are now computed by minimizing the relay events for each rank.
Connections not present in routes are removed automatically by the message router.
In the network test, the length of the vector is reserved once if raw data are to be stored.
The order in which the routes are computed is now random.
Don't show the final message if the option -test-network-only was provided.
Improved the algorithm that generates the random graph for message routing.
Fixed the number of edges in the complete graph.
Decoupled the creation of the connection graph from the actual message routing.
The route are now computed with more verbosity.
Implemented de Bruijn message routing.
Implemented various types of graphs for message routing.
Changed the code so that no routes are computed with de Bruijn routing.
The code for de Bruijn routing is now working just fine when
Optimized isConnected and getNextRankInRoute for de Bruijn routing.
Simplified the code that finds the next vertex in the de Bruijn routing.
Simplified the implementation of de Bruijn message routing.
Fixed a bug that created files with strange names (unexpected behavior).
Corrected the number of edges in Routing.Summary.txt.
Implemented message routing with Kautz graphs.
Modified de Bruijn and Kautz graph implementations to avoid copying bytes.
Added an experimental graph implementation.
Removed a comment in the seed extension code.
Added speed indicators and some estimation of remaining time to
Added software components to get speed readouts throughout the seed extension iterations.
Added option -routing-graph-degree to change the degree of the routing graph.
The mode latency is measured instead of the average during the network test.
The slave mode is now shown with the speed of Ray.
Added average values for each profiled steps.
Added a class for assembly seeds.
The seed extension is now done independently from
The peak coverage is not utilised in the read k-mer indexing.
The k-mer peak coverage is no longer an issue.
Removed the peak coverage dependency in the vertex messenger.
Removed the dependency on the peak coverage from the OpenAssembler engine.
Removed the repeat coverage from the seed extender.
Removed the peak coverage dependency from the bubble objects.
Removed the restriction about peak coverage values within the scaffolding bits.
Added a script to compile Ray with fancy options.
Added a maximum number of flow cycles
Created the SwitchMan module and ported the network test code.
Added a scripting directory and some methods to the switchman.
Added design blueprints for Ray Biological Abundances.
Updated the README.
Enhanced the layout
Added master switches in the SwitchMan.
New structure for scripting-related files.
Migrated some scripting code in the ScriptEngine.
Updated the design ideas for biological abundances
Added 3 new list for compilation (empty tags, slave modes & master modes).
Implemented contig biological abundances (enabled by default).
The frequency table must be cleared regularly
Moved the license in the main directory.
Implemented entry counting for biological abundances.
Disabled detailed reports and added an option to enable them.
Added option -minimum-contig-length.
Counting sequences to search is now implemented.
Almost done implementing biological abundances.
Biological abundances are now operational.
Files are now written correctly to BiologicalAbundances.
Fixed directory names for option -search.
Added speed during the biological abundance computation.
Switched to BufferedData for buffered messages.
Prototyped a use-case for buffering messages in the search engine.
Communications are now buffered in the search engine.
Added the mean k-mer coverage in the list of metrics.
Ray is now tested with these C++ compilers too: pathscale and pgi.
Now using the BufferedData object to compute contig abundances.
Added new MPI tags to eventually distribute more the inputs and outputs.
Added comments in the virtual communicator.
Added some comments in the switchman
Each rank now writes their own files instead of sending data to master.
Fixed the contig mean coverage and an assertion.
Documented the memory management code.
Fixed a compilation warning with fprintf()
K-mer coverage depth proportion values are now computed by Ray.
Normalised method names for master modes.
Normalised method names for slave modes.
Renamed the script engine and the nova engine.
Changed column names to more verbose ones.
Changed the default to DEBUG=n to disable debugging symbols.
Added a machine helper class to delegate things.
Moved the code of call_RAY_MASTER_MODE_SEND_COVERAGE_VALUES() in the machine helper.
Added slave switches for RAY_MPI_TAG_WRITE_KMERS, RAY_MPI_TAG_EXTENSION_END and RAY_SLAVE_MODE_EXTRACT_VERTICES.
Ported the scheduling of the k-mer academy step to the switchman framework.
Added a tick logger that outputs in RayOutput/Scheduling/.
Added handlers for slave modes, master modes and message tags.
Removed the empty tag list.
I separated application code from platform code.
Ported the code to use slave and master handlers.
Now using setObjectHandler() for master and slave methods too.
I removed reference to Ray in type names in the platform.
I added a new class representing a compure core.
Added comments for the compute core.
Moved all handlers in Machine to MachineHelper.
Added a file describing the handler design.
Ray is now distributed as Ray Application and Ray Platform.
Ray Platform is now licensed under the LGPLv3.
Ray Platform is now licensed under the LGPLv3.
Now converting bases to upper case if necessary.
Changed the maximum line length for fasta files.
Now adding the license to installations.
Fixed the EXTRAVERSION for beta5.
We don't need to configure the virtual communicator twice.
Started to implement contig identification.
I created a working prototype for contig identification.
Moved Ray Platform to RayPlatform.
Moved code/Application/* to code/*.
I removed license notices in scripts.
Fixed paths in the README.
Added a README for the Ray Platform.
Added more documentation for the Ray Platform.
Added an error message in the switchman is not configured properly.
I explained why latency data files can not be written before the end for everyone.
Fixed greater-than-1 ratios for contig identifications.
Cleaned the sorting code.
Contig identification is now working properly.
I added threshold to remove some false positives.
Modified the version to 2.0.0-beta6.
Fixed install.sh script.
I fixed a fatal bug in the search engine for k-mers.
Merge branch 'master' of git at github.com:sebhtml/ray
I fixed a bug in the file creation.
I fixed a compilation error with gcc (GCC) 4.6.2 20111027 (Red Hat 4.6.2-1).
I created a new interface called CorePlugin for modularity.
I removed ratio for ticks counts.
I ported the scaffolder, the network test and the message processor to the plugin architecture.
I ported MachineHelper, Amos and FusionTaskCreator to the CorePlugin architecture.
I ported JoinerTaskCreator, Library, Partitioner, SeedingData, and CoverageGatherer to the CorePlugin architecture.
I simplified the way the ComputeCore is utilised.
I started to work on graph coloring with Ray.
I removed the mean coverage in the search results.
I started to implement color virtualization.
I changed the threshold to report something.
K-mers with Ns are not counted anymore by the search engine.
The GenBank identifiers are now detected in files.
I added a slave mode for coloring the graph.
I implemented virtual colors in Ray.
Unidentified sequences are not colored.
I added documentation for virtual colors.
Virtual colors with 0 references are recycled.
The index in the color set has been removed.
Physical colors are not embedded in namespaces.
Biological abundances are now computed with colors.
I removed useless code in allocateVirtualColor().
A set of available handles reduced the time complexity to constant time.
I corrected performance issues in the graph coloring algorithms.
Handlers now have only one method called call().
The Library plugin now uses adapters.
The Searcher plugin now utilises adapters to register itself.
The plugin Partitioner now uses adapters (a design pattern) to register itself
The plugin NetworkTest now utilises adapters to register with the core.
The plugin MessageProcessor now utilises adapters to register with the core.
The Makefile was updated to add adapters.
The plugin Scaffolder now utilises adapters to register with the core.
The plugin MachineHelper now uses adapters to register with the core.
The plugin Amos now uses adapters to register with the core.
The plugin FusionTaskCreator now uses adapters to register with the core.
The plugin SeedingData now uses adapters to register with the core.
The plugin JoinerTaskCreator now utilises adapters to register with the core.
The plugin CoverageGatherer is now using adapters to register with the core.
I fixed the Makefile for adapters.
I removed the script engine from the RayPlatform.
A plugin now must ask the core to allocate handles.
Things in the Makefile file in code/ are now grouped per plugin.
The plugin SeedExtender is now using adapters to register with the core.
Registered plugins are now printed in <RayOutput>/Plugins/*.Plugins.txt.
I added the rank in the SwitchMan.
Ratios for slave mode ticks are not printed anymore.
The plugin SequencesIndexer now utilises adapters to register with the core.
The plugin FusionData now utilises adapters to register with the core.
The plugin SequencesLoader now uses adapters to register with the core.
The plugin KmerAcademyBuilder now uses adapters to register with the code.
The plugin EdgePurger now uses adapters to register with the core.
The plugin VerticesExtrator now uses adapters to register with the core.
The handle RAY_SLAVE_MODE_COUNT_FILE_ENTRIES is now in the plugin Partitioner.
A system of symbol registration was added.
The macros for listing slave modes were removed.
Plugins now register their own master modes.
Message tags are now allocated by the core. Plugins have to resolve symbols.
The memory allocation types were removed.
RayPlatform now compiles independently without knowledge of an application.
I updated the makefiles.
I fixed a bug in setPluginDescription().
I added a cmake file for RayPlatform.
I added a CMakeLists.txt file for Ray, but it does not work yet.
Master switches are now registered by the plugins using the ComputeCore API.
MessageTag-to-SlaveMode switches are now registered by plugins using the ComputeCore API.
Added VirtualCommunicator and VirtualProcessor in the plugin list.
Plugin descriptions are now written in the directory <RayOutput>/Plugin/.
Reply MessageTag handles are now registered by plugins.
MessageTag sizes are now registered by plugins.
Plugins now register their next master modes.
I updated plugin descriptions.
Don't register plugins before the first call to registerPlugin().
I fixed a bug in the Makefile files.
Fixed the Makefile, yet again.
I moved the RayPlatform to another git repository.
The version of the RayPlatform is now written in a file called RayPlatform_Version.txt.
The version of RayPlatform is now printed with option -version.
I updated Ray to updates to Ray Platform.
I removed desired values from call to allocate*Handle().
I created application_core/
I created directories plugin_Amos and plugin_CoverageGatherer.
I removed an extra new line.
I created plugin_EdgePurger.
Moved files in plugin directories.
Moved license.
I added a CMakeLists.txt to compile Ray on Microsoft Windows.
I added the code to print virtual color summary.
v2.0.0-rc4
I fixed the install script again.
The maximum coverage is not set to 2^sizeof(COVERAGE_TYPE)-1
Builds are now performed with 4 slots for g++.
The virtual color algorithms should now be faster.
The indexing strategy is now better for virtual colors.
I modified the virtual color handle code.
I simplified the indexing algorithm.
Unit tests test_invert.sh, test_coverage_distribution.sh and scaffolder_test.sh are now updated.
I added a progression indicator.
A progression indicator was added for sequence abundances.
I removed calls to seekg() and tellg().
I modified the coloring code to fully exploit message multiplexing of RayPlatform.
I added a script to automate product shipment.
I refactored the unit test framework.
Two bugs were fixed for message multiplexing when coloring.
The coloring code is now using stdio.h instead of iostream.
A header in ScaffoldLinks.txt was added.
I added uniquely colored assembled k-mer readouts.
I added in-place operations for virtual colors with 1 reference.
Colored assembled k-mers need to be in contigs.
I changed the minimum threshold for nicely assembled.
I added a simple test to weed out false positives.
I updated the test test_limit_kmer.sh.
Fixed test_novaengine.sh.
I fixed the test for test_peakFinder.sh.
I fixed unit-tests/test_uniformity.sh.
Memory usage is now reported by default.
Entries with no colored and assembled k-mers are not written at all.
Outer distances (insert sizes) are not estimated if there are no paired reads.
I changed the threshold for strangely assembled but colored k-mers.
The plugin plugin_Searcher now writes distributions too.
I added unit tests for the algorithm that finds peaks.
A more robust algorithm to detect library peaks is now shipped with Ray.
I added a configuration command for buffers.
New script to plot colored distributions.
I fixed headers of distributions.
I removed the threshold for uniquely colored and assembled k-mers.
I resolved a compilation warning in plugin_Searcher.
Code for flagging false positives have been added.
Buggy correlations are now printed in color plots.
I updated the routing tests.
The CMakeList.txt has been updated to use the new files in RayPlatform.
I fixed double-to-int conversions.
The script that ships the product is now enhanced.
mpiexec and exit() are not friends after all.
HAVE_LIBBZ2=y works as expected now.
The callback call_RAY_MPI_TAG_SEARCHER_CLOSE() was added.
I added some documentation for people who want to submit changes.
Merge branch 'master' of git://github.com/enormandeau/ray
Writing files can not be distributed yet because only rank 0 has some of
Sequence abundances are now written in a single file per
The new directory _Coloring contains a coloring summary.
I modified the instructions to require signed-off patches.
Ray now writes all coloring distributions in XML.
Ray now generates a file per directory containing sequence counts.
I fixed the quality values.
PeakFinder now supports simulated data too.
I added placeholders to transmit pointers.
I modified another part that may cause segmentation faults.
I added pack_pointer and unpack_pointer functions.
A file containing directories is now created.
Files containing virtual colors are just too large to be written.
Observations are 0 if some quality values are 0.
I removed outdated scores.
K-mers with a coverage < 2 are not valid.
I reduced the complexity of QualityCaller.
ColoredPeakCaller is now utilised to weed out false positives.
The test suite has moved to a different project.
Modified the colored peak finder to pass tests.
A bug was fixed for call_RAY_MPI_TAG_REQUEST_VERTEX_READS().
This commit finally fixes a very critical bug that happens once in a while.
I modified the peak caller to pass new tests.
Added a new plugin for phylogeny analyses.
Phylogeny colors are forwarded to call_RAY_MPI_TAG_ADD_KMER_COLOR().
The PhylogenyViewer core plugin is now wired.
Proportions are automatically calculated from now on.
Physical colors for the phylogeny are fetched from the graph.
Taxon are now fetched using the phylogeny colors.
Added some code for syncing taxons across the tribe.
The phylogenetic tree is now loaded properly.
Ray now loads taxon names too.
K-mer observations are placed in the tree of life.
Only consider assembled k-mers for unknown observations.
Taxon observations are synced with master.
Taxon hits are written in a file.
I updated the manual.
The number of responses initially has to be m_size, not 0.
Taxon observations are now reported along with a proportion and in XML.
Ray writes SequenceAbundancesRaw.tsv too.
call_RAY_MASTER_MODE_COUNT_SEARCH_ELEMENTS is now the entry point of plugin_Searcher.
The number of colored k-mers is reported in _DeNovoAssembly for each contig.
The code that computes the total number of k-mer observations has been fixed.
I fixed 3 compilation warnings.
Discovered k-mers in the graph should be sufficient to cover a search entry.
A XSL sheet was added for Taxons.xml.
I edited the solutions section.
The code contributed by Éric Normandeau is broken and does not work.
I added a simple test to filter out some false positives.
The symbols '<' and '>' are not directly allowed in XML files.
The sample name was added in _Phylogeny/Taxons.xml.
Biological abundances are now produced in XML.
The XSL sheets were updated.
I added file names too.
The format of Taxons.txt was updated to include the taxonomic rank too.
I added a colored ratio for taxonomy outputs to consider only conserved sequences.
The XML files with distributions now contain directly the directory and file names.
Ray now issues a warning (and an error) is the taxonomic tree contains loops.
The list of pairs from Genome-to-Taxon should not be printed with warnings.
A new chunk that prints a warning if a taxon has no parent was added.
I fixed the XSL sheet.
The <coloredProportion> node was added for unknown stuff.
Recursive counts are now computed too.
All taxons with a non-null recursive count are reported.
Only 1 XML file is necessary for each search directory. The transformation is done by XSLT.
Two new XSL sheets were created to generate visually appealing tables.
A blueprint for l-mers was authored.
The sequence names can not contain special characters reserved for XML.
Entries are now sorted in the taxon tables.
A taxon not being in the tree should not be fatal.
The colored proportions should be constrained to a given rank for better
Reading reference frequencies and color frequency is useful.
<MPIrank>.CoverageData.xml contains contig coverage values.
More details are required for not-a-tree errors.
Paths should not be tested in production settings.
1-based positions are easier to understand with a proper file header.
Only the printed position should be 1-based.
The estimated genome length is not very accurate sometimes.
Sorted entries are easied to read.
The code that fetches the base name of a virtual file system path is buggy.
Two new XSL sheets can fetch the most abundant phylum and family taxons.
The plugin GenomeNeighbourhood (compatible with RayPlatform) produces
The plugin GenomeNeighbourhood is not linked with the rest.
The design for computation of neighbourhoods is written.
The CMakeList.txt file is now correct.
Exploration of the graph highlights path relationships.
Fetching the number of paths is a good step for implementing neighbourhood discovery.
Detailed information is not required in the standard output.
Paths are fetched with message passing to get neighbourhoods.
The code that gets neighbourhoods is almost operational.
Nearby paths are detected automatically.
The exploration stops locally when something gets in the way.
Left neighbours are communicated to master.
Right neighbours are communicated too.
The final list for neighbours is agglomerated on master.
A new output file called NeighbourhoodRelations.txt is created.
Position is strand-specific.
New XSL sheets are necessary to convert taxonomic profiling XML files.
This is Ray release candidate 5.
The CMake file was updated.
GC content can be useful when plotting contig coverage depths.
The new option -one-color-per-file will force Ray to use one color per file.
The manual page was updated.
Input/output file operations are faster in "append" mode than in normal mode.
Using append mode makes operations faster.
The XML files for contig coverage are generated with a few input/output operations.
The sequence abundance XML files are now written with buffered input/output
The input/output operations for contig identifications are grouped
I added a comment for getWriter()
More documentation was added.
The buffer size for input/output operations on the file system is 32 MB.
File system operations are buffered for colored coverage distributions.
File intput/output operations for taxons are buffered.
Contig lengths are written to disk using buffering.
Contigs.fasta is written using buffering.
Scaffolds.fasta is now written with buffering in mind.
Scaffold links are written to disk with buffering.
Scaffold lengths and scaffold components are written to disk with buffering.
I changed CONFIG_FILE_IO_BUFFER_SIZE to 16 MB.
Memory usage will be reported during the joining of paths.
The number of assembled k-mers was added.
A progression indicator was added for the network test.
The total number of colored k-mers and k-mer observations are now computed too.
I added a XSL sheet to fetch proportions.
Progression indicators were added for scaffolding.
A integer overflow was fixed in the scaffolder.
The buffering code was optimized.
The neighbourhood file is generated with buffering.
I added a guide for message routing.
I created another XSL sheet for sequence abundances.
I fixed a integer overflow in the profiler.
Total counts for parts of the distributed graph are shared correctly.
I increased the sensitivity of the method while not changing the specificity (which is very high).
I added <totalColoredKmerObservations> and <totalColoredKmers> to XML objects.
I disabled the plugin for neighbourhoods since it is buggy.
This is an experimental patch to accelerate one of the slowest computational
I added the Number of k-mers in the distributed de Bruijn graph.
Support was added for a EMBL_CDS namespace.
The backbone for a GeneOntology CorePlugin is ready.
The CorePlugin GenomeNeighbourhood is reinstated, but it is still disabled.
4 compilation warnings were weeded out.
A variable shadowing caused a compilation error on OS X using openmpicxx.
Ray now counts colors related to EMBL_CDS objects.
Two bugs were fixed. Hangs could be encountered depending on
The period for reporting to the user the progression of loading sequences
Gene ontology annotations are loaded by Ray.
The code path should continue as usual for de novo assemblies.
Biological signal for ontology terms are extracted from
I ported the plugin GeneOntology to the new macro commands.
A compilation warning was fixed regarding a comparison between two
Ontology terms are counted and synchronized.
The documentation of Ray was updated to include gene ontology profiling.
Ray now produces gene ontology profiling.
Gene ontology profiles are now generated in a compact format too.
A missing header was added.
The switchman handle should not be in the plugin MessageProcessor.
A missing header was added.
Callback methods are public by convention.
Plugins interact indirectly with the compute core using adapters, which
Old declarations and implementations of adapters were removed from the tree.
I fixed a linking error when ASSERT is not provided.
The CMakeLists.txt file was updated.
This is Ray version v2.0.0-rc6 (release candidate 6).
I forgot to update the manual.
This patch fixes a bug that makes Ray hangs.
Any k-mer should only be counted at most once for each gene
I added proportions for gene ontology terms.
I fixed a division by zero.
I updated the formula for computing proportions of DNA sequences.
Relationships are now loaded from the gene ontology file.
The paths to the root are written to the XML file too.
Proportions are now available automatically in the
I created a XSL sheet to convert the gene ontology Terms.xml file
Domains are loaded from gene ontology, and depths are computed too.
Gene ontology information is extracted for each domain.
The total number of observations used for computing g ne ontology proportion
A bug was fixed for the synchronization of counts. The master rank does not need to send
I fixed a bug occuring when -gene-ontology Operand1 Operand2
K-mer observations are not required to be assembled
Profile files for taxonomy are automatically generated.
Some code was added to derefence alternate identifiers for gene ontology analysis.
Profile files are also produced with -search option.
The number of dereferenced handles was added.
I added the brandname to profile files.
I removed the superfluous branding.
This change set fixes a bug introduced in the commit 19c1838382cda4bcc19dffcdc25745a076ffc38a.
A critical bug was fixed in the gene ontology plugin.
I added the percentage of assembled k-mers for each
A new XSL sheet was created to obtain the assembled proportions
A bug concerning the file name for -write-kmers was fixed.
I added more details to an obscure error message in Ray.
The code was polished and more information was added
I added more information to zero in on the bug.
A elusive bug that caused runtime problems was finally solved.
These data types were added to avoid elusive bugs:
The maximum number of message tags was moved in RayPlatform.
This is Ray v2.0.0-rc7.
I fixed again the CMAKELIST file.
Deprecated parameters were removed from the help page.
I added an error message for unknown extensions.
The code that put together paths was modified to avoid creating misassemblies.
ParallelPaths.txt contains parallel path data.
Ray v2.0.0-rc8 has even less misassemblies (Ray v2.0.0-rc7 had so few already).
The options for using checkpointing features requires a
A new option is available to disable read recycling.
The release procedure was updated.
Patch information was updated.
Routing strategies were updated.
A compilation warning was removed for an integer comparison.
The peak finder should detects simulated data as well.
The multiplicator used for spawning read helpers was changed from 2.0 to 1.5.
This change set improves the fidelity of Ray when computing peak
This solves a division by 0.
Floating numbers must not be stored with the integer type 'int'.
This is Ray v2.0.0.
The copyright was updated to add 2012.
When there are 508 reads and 32 MPI ranks, the number of reads
A list of releases was added.
The codename of the next release will be "Ancient Granularity of Epochs".
An assertion was added for the performance scaled messaging related
Two assertions were added to detect possible message corruption.
The help page was update to add the data reliability option.
The peak finder was modified to pass new tests.
I edited the guide to submit changes.
The manual now includes the new option for overly-covered seeds.
A error was fixed in the file that says how to submit changes.
The return statement was misplaced in a recent patch.
I added the names 'Ray Méta', 'Ray Communities', and 'Ray Ontologies'.
An assertion was added to make sure that data is not overwritten.
Searcher: added verbose statements
Searcher: fixed a race condition
Searcher: added a missing value.
SeedExtender: moved system calls inside this plugin
SeedExtender: modified the code for hot skipping
SeedExtender: implemented hot skipping
Parameters: 4 options were added to change distributed storage behavior.
Documentation: Ray can be run with a single configuration file containing options.
The default load factor threshold was changed to 0.75.
The methods setKey() and getKey() were added to KmerCandidate
If the hash table is verbose, ask it to display its status.
NetworkTest: added the option -skip-network-test to skip the network test.
Added a new option to enable genome neighbourhood calculation.
I added some code to detect windows 32 bits and windows 64 bits.
More parameters for compilation can be provided with EXTRA=...
Porting Ray to the new RayPlatform: removed macro calls in .h files.
Porting Ray to the new RayPlatform: removed remaining codes in .h.
Porting Ray to the new RayPlatform: removed token 'generated_automatically'.
Porting Ray to the new RayPlatform: added CreatePlugin and BindPlugin
Porting Ray to the new RayPlatform: updated the macro names in C++
Porting Ray to the new RayPlatform: removed adapter from plugin
Porting Ray to the new RayPlatform: remove calls to setObject.
Porting Ray to the new RayPlatform: Ray compiles with the simplified
I removed handlers from the cmake file.
Updating the manual.
SeedExtender: changed the verbosity period.
Removed some output from the computation of seeds.
The manual was updated to include pointers to documentation.
If you run Ray with a configuration file (mpiexec -n 4 Ray Ray.conf)
Information to compile Ray with gcc was added.
The default number of buckets is now 1048576. The default number of
This fixes a input/output bug for the Ray configuration file.
The code that randomizes the arguments was removed because it can
The edge purging should be done in a massively parallel way unless
Merge branch 'master' of https://github.com/plpla/ray into pl
I added a script to build Ray with link time optimization.
The EXTRA commands are also given to the linking command.
I added -fwhole-program for better optimization.
I added compilation flags for compression.
I added instructions to build Ray with link time optimization.
NetworkTest: the number of test messages is now constant regardless
application_core: added a call to obtain a string configuration
KmerAcademyBuilder: option -bloom-filter-bits can sets the number
KmerAcademyBuilder: Bloom filter has 64 M bits by default.
Merge branch 'master' of github.com:sebhtml/ray
Merge branch 'master' of github.com:sebhtml/ray
SequencesLoader: added a 'please wait' before counting entries in
SequencesLoader: a bz2 file can contain many compressed streams.
application_core: bugs were fixed in the configuration routines.
GeneOntology: removed the use of argv
Merge branch 'master' of github.com:sebhtml/ray
Merge branch 'master' of github.com:sebhtml/ray
Fixed an integer overflow in the distributed storage engine.
A path with 0 k-mers has 0 nucleotides, not 0-k+1.
Merge branch 'master' of github.com:sebhtml/ray
A new routing graph is available: the hypercube.
Documentation: documented the hypercube features of Ray.
core: the default number of buckets is now 268435456 per rank.
scaffolder: it can be disabled with -disable-scaffolder
normalized option names with -enable-* and -disable-*
documentation: moved assembly options up
core: added documentation for class Parameters.
SeedingData: -use-minimum-seed-coverage changes the minimum
documentation: added missing operands in the manual and -help page
core: Ray -version provides more compile flags like popcnt and sse
SeedingData: seeds can not contain k-mers with too low coverage
build: the C++ standard is C++ 1998. gcc -ansi provides that
Searcher: large integer constants needs ULL for portability
SeedExtender: added additional information for an error
MessageProcessor: k-mer data messages should never be discarded
VerticesExtractor: don't flush while waiting for messages
KmerAcademyBuilder: only send the forward k-mer, not the lower
VerticesExtractor: improved the code quality for easier reading
MessageProcessor: don't discard k-mers while receiving messages
VerticesExtractor: store twin edges in a single source
EdgePurger: any edge is removed only if a end is not in the graph
MessageProcessor: removed a call to a private attribute
Documentation: added a document about profiling Ray
Documentation: added information about elapsed time
BuildSystem: added a strip command to reduce the memory footprint
BuildSystem: replaced -ansi with -std=c++98 for more verbosity
Documentation: updated the author file
KmerAcademyBuilder: removed the k-mer academy
VerticesExtractor: this module extracts vertices to add edges
Merge branch 'kill-kmer-academy'
MessageProcessor: new text to show when the Bloom filter is created
KmerAcademyBuilder: added the number of set bits in the Bloom filter
MessageProcessor: added a warning when the oracle is half full
KmerAcademyBuilder: the Bloom filter can have any number of bits
Merge branch 'bloom-features'
MessageProcessor: coverage depth starts at 1 with Bloom filters
MessageProcessor: the thresold is 50.0 (50.0%), not 0.5
KmerAcademyBuilder: added the number of filtered k-mers
Merge branch 'bug-hunting'
application_core: added routing with a convex regular polytope
NetworkTest: the number of exchange can be changed with -exchanges
Documentation: added options for a 64-rank polytope
Documentation: updated the taxonomy documentation
NetworkTest: added average round trip latency
scripts: initial version of a script to create NCBI taxonomy
scripts: download NCBI bacterial genomes too
Merge branch 'master' of github.com:sebhtml/ray
Documentation: added documentation for NCBI taxonomy
Documentation: simplified the usage of the tool to pull NCBI data
scripts: the script that pulls NCBI data is almost ready
scripts: the script that pulls NCBI stuff is ready
Documentation: added information about XML files
Partitioner: also create a file FilePartition.txt
MachineHelper: don't run the AMOS code path if not necessary
Parameters: throw a warning when distances are invalid
Merge branch 'for-seb-September-2012'
Searcher: fixed a race condition where a message was lost
Calls to deprecated methods were eliminated.
This is Ray v2.1.0-rc0 "Ancient Granularity of Epochs"
-----------------------------------------------------------------------
--
Packaging of Ray in Debian
More information about the debian-med-commit
mailing list