v2.1.2

JOSS submission updates

v2.1.1

Bugfix release for documentation.

Added

• Document for submission to the Journal of Open Source Software.

Fixed

• Several typos and outdated content in the documentation.

v2.0

khmer v2.0 release notes

This is the v2.0 release of khmer and the first from our new lab at the University of California, Davis. It features Python 3 compatibility, streaming I/O from Unix Pipes, mixed-pair sequence file format support, and a new parameter to simplify memory usage. We also have a software paper in-press describing the project and the citation reminders have been updated to reflect that.

Overall there are an additional 2,380 lines of Python code (mostly tests) and 283 less lines of C++ (despite adding features). This release is the product of over 1,000 commits to the codebase since v1.4.

Documentation is at https://khmer.readthedocs.org/en/v2.0/

New items of note:

New behavior

Streaming I/O from Unix Pipes

All scripts now accept input from named (like /dev/stdin, or that created using <( list ) process substituion) and unnamed pipes (like output piped in from another program with |). The STDIN stream can also be specified using a single dash: -. #1186 @mr-c #1042 #763 @SherineAwad #1085 @ctb

New parameter for memory usage, and/or tablesize/number of table parameters.

There is now a -M/--max-memory-usage parameter that sets the number of tables (-N/--n_tables) and tablesize (-x/--max-tablesize) parameters automatically to match the desired memory usage. #1106 #621 #1126 #390 #1117 #1055 #1050 #1214 #1179 #1133 #1145 @ctb @qingpeng @bocajnotnef

Digital normalization script now supports mixed paired and unpaired read input

normalize-by-median.py now supports mixed paired and unpaired (or "broken-paired") input. Behavior can be forced to either treat all reads as singletons or to require all reads be properly paired using --force_single or --paired, respectively. If --paired is set, --unpaired-reads can be used to include a file of unpaired reads. The unpaired reads will be examined after all of the other sequence files. normalize-by-median.py now has a --quiet option to reduce the amount of output. #1200 @bocajnotnef

Mixed-pair sequence file format support

split-paired-reads.py --output-orphaned/-0 has been added to allow for orphaned reads and give them a file to be sorted into. #847 #1164 @ctb

Scripts now output columnar data in CSV format by default

All scripts that output any kind of columnar data now do so in CSV format, with headers. Previously this had to be enabled with --csv. (Affects abundance-dist-single.py, abundance-dist.py, count-median.py,
and count-overlap.py.) normalize-by-median.py --report also now outputs in CSV format. #1011 #1180 @ctb

Reservoir sampling script extracts paired reads by default

sample-reads-randomly.py now retains pairs in the output, by default. This can be overridden to match previous behavior with --force_single.

Most input and output files can be compressed

We support gzip and bzip2 input and output file compression everywhere that it makes sense #505 #747 @bocajnotnef

New scripts

Estimate number of unique kmers

unique-kmers.py estimates the k-mer cardinality of a dataset using the HyperLogLog probabilistic data structure. This allows very low memory consumption, which can be configured through an expected error rate. Even with low error rate (and higher memory consumption), it is still much more efficient than exact counting and alternative methods. It supports multicore processing (using OpenMP) and streaming, and so can be used in conjunction with other scripts (like normalize-by-median.py and filter-abund.py). This script is the work of @luizirber and the subject of a paper in draft. #390 #1239 #1252 #1053 #1072 #1145 #1176 #1207 #1204 #1245

Incompatible changes

New datastructure and script names

For clarity the Count-Min Sketch based data structure previously known as "counting_hash" or "counting_table" and variations of these is now known as countgraph. Likewise with the Bloom Filter based data structure previously known at "hashbits", "presence_table" and variations of these is now known as nodegraph. Many options relating to table have been changed to graph. #1112 #1209 @mr-c

Binary file formats have changed

All binary khmer formats (presence tables, counting tables, tag sets, stop tags, and partition subsets) have changed. Files are now pre-pended with the string OXLI to indicate that they are from this project. #519 #1031 @mr-c #1159 @luizirber

Files of the above types made in previous versions of khmer are not compatible with v2.0; the reverse is also true.

In addition to the OXLI string, the Nodegraph and Countgraph file format now includes the number of occupied bins. See http://khmer.readthedocs.org/en/v2.0/dev/binary-file-formats for details. #1093 @ctb @mr-c #1101 #1103 @kdmurray91

load-graph.py no longer appends .pt to the specified filename

Previously, load-graph.py appended a .pt extension to the specified output filename and partition-graph.py appended a .pt to the given input filename. Now, load-graph.py writes to the specified output filename and partition-graph.py does not append a .pt to the given input filename. #1189 #747 @bocajnotnef

Some reporting options have been turned always on

The total number of unique k-mers will always be reported every time a new countgraph is made. The --report-total-kmers option has been removed from abundance-dist-single.py, filter-abund-single.py, and normalize-by-median.py to reflect this. Likewise with --write-fp-rate for load-into-counting.py and load-graph.py; the false positive rate will always be written to the .info files. #1097 #1180 @ctb

An uncommon error recovery routine was removed

To simplify the codebase --save-on-failure and its helper option --dump-frequency have been removed from normalize-by-median.py.

Single file output option names have been normalized

--out is now --output for both normalize-by-median.py and trim-low-abund.py. #1188 #1164 @ctb

Miscellaneous changes

The common option --min-tablesize was renamed to --max-tablesize to reflect this more desirable behavior.

In conjuction with the new split-paired-reads.py --output-orphaned option, the option --force-paired/-p has been eliminated.

As CSV format is now the default, the --csv option has been removed.

Removed script

count-overlap.py has been removed.

Notable bugs fixed/issues closed:

When normalize-by-median.py decides to keep both parts of a pair of reads it was only adding the k-mers & counts from one to the countgraph. #1000 #1010 @drtamermansour @bocajnotnef

The partition map file format was not robust to truncation and would hang waiting for more data. #437 #1037 #1048 @ctb

extract-paired-reads.py and split-paired-reads.py no longer create default files when the user supplies filename(s). #1005 #1132 @kdmurray91

Additional fixes/features

find-knots.py was missing a --force option and unit tests. #358 #1078 @ctb
The check for excessively high false-positive rate has also received a --force option #1168 @bocajnotnef

A bug leading to an infinite loop with large gzipped countgraphs was found #1038 #1043 @kdmurray91

All scripts that create nodegraphs or countgraphs report the total number of unique k-mers. #491 #609 #429 @mr-c

Read pairs from SRA are fully supported. Reported by @macmanes in #1027, fixed by @kdmurray91 @SherineAwad in #1173 #1088

Of interest to users:

Added Hashtable::get_kmers(), get_kmer_hashes(), and get_kmer_counts() with corresponding CPython functions. #1047 #1049 @ctb

The DEFAULT_DESIRED_COVERAGE for normalize-by-median.py is now 20. #1073 #1081 @ctb

FIFOs are no longer seen as empty. #1147 #1163 @bocajnotnef

When the k-size is requested to be larger than 32 (which is unsupported) a helpful error message is reported. #1094 #1050 @ctb

We try to report more helpfully during errors, such as suggesting the --force option when outputs files already exist. #1162 #1170 @bocajnotnef

There is a paper related to trim-low-abund.py: "Crossing the streams: a framework for streaming analysis of short DNA sequencing reads" and it has been added to the CITATION file and program output. #1180 #1130 @ctb

We have dropped support for Python 2.6 #1009 #1180 @ctb

Our user documentation got a bit out of date and has been updated. #1156 #1247 @bocajnotnef @mr-c #1104 @kdmurray91 #1267 @ctb Links to lists of publications that use khmer have been added #1063 #1222 @mr-c The help text from the scripts has also had a thorough cleanup for formatting. #1268 @mr-c

fastq-to-fasta.py's --n_keep option has incorrect help text. We now point out that all reads with Ns will be dropped by default unless this option is supplied. #657 #814 #1208 @ACharbonneau @bocajnotnef

We've updated the URL to the '88m-reads.fa.gz' file. #1242 #1269 @mr-c

@camillescott designed and implemented an optimization for normalize-by-median.py #862

abundance-dist.py can now be used without counts over 255 with --no-bigcount. #1067 #909 @drtamermansour @bocajnotnef Its input file requirement can no longer be overridden #1201 #1202 @bocajnotnef

khmer v2.0 will be released as a package for the Debian GNU/Linux operating system. Big thanks to @kdmurray91 for his assistance. #1148 #1240 The C++ library, now named liboxli, will have its own package as well.

sandbox/multi-rename.py now wraps long FASTA sequences at 80 columns. #450 #1136 @SherineAwad

Of interest to developers:

The khmer project is now a Python 3 codebase with backwards compatibility to Python 2.7. Huge credit to @luizirber #978 #922 #1045 #1066 #1089 #1157 #1191 #1108 Many developer impacting changes including the file khmer/_khmermodule.cc is now khmer/_khmer.cc. #169 #904

@camillescott did an extensive refactor of the C++ graph traversal code which removed a considerable amount of redun...

V1.4.1

khmer v1.4.1 release notes

This is the v1.4.1 release of khmer. Due to the upcoming Python 3 compatibility in both khmer and Screed we need to modify the dependency between khmer and the Screed library to be only the existing version 0.8, and not some future version.

If you have khmer 1.4 installed then there is no benefit to upgrading; this point release is to keep pip install khmer still working when we release the next version of Screed with Python 3 support. The next version of khmer, v2.0, will also have Python 3 support.

Documentation is at https://khmer.readthedocs.org/en/v1.4.1/ (no changes from v1.4)

Known issues:

All of these are pre-existing.

Some users have reported that normalize-by-median.py will utilize more memory
than it was configured for. This is being investigated in
#266

Some scripts only output FASTA even if given a FASTQ file. This issue is being
tracked in #46

V1.4

khmer v1.4 release notes

This is the v1.4 release of khmer featuring the results of our March and April (PyCon) coding sprints and the 16 new contributors; the use of the new v0.8 release of screed (the library we use for pure Python reading of nucleotide sequence files); and the addition of @luizirber's HyperLogLog counter for quick cardinality estimation.

Documentation is at https://khmer.readthedocs.org/en/v1.4/

New items of note:

Casava 1.8 read naming is now fully supported and in general the scripts no longer mangle read names. Side benefits: split-paired-reads.py will no longer drop reads with 'bad' names; count-median.py can generate output in CSV format. #759 #818 @ctb #873 @ahaerpfer

Most scripts now support a "broken" interleaved paired-read format for FASTA/FASTQ nucleotide sequence files. trim-low-abund.py has been promoted from the sandbox as well (with streaming support). #759 @ctb #963 @sguermond #933 @standage

The script to transform an interleaved paired-read nucleotide sequence file into two files now allows one to name the output files which can be useful in combination with named pipes for streaming processing #762 @ctb

Streaming everywhere: thanks to screed v0.8 we now support streaming of almost all inputs and outputs. #830 @aditi9783 #812 @mr-c #917 @bocajnotnef #882 @standage

Need a quick way to count total number of unique k-mers in very low memory? the unique-kmers.py script in the sandbox uses a HyperLogLog counter to quickly (and with little memory) provide an estimate with a controllable error rate. #257 #738 #895 #902 @luizirber

normalize-by-median.py can now process both a paired interleaved sequence file and a file of unpaired reads in the same invocation thus removing the need to write the counting table to disk as required in the workaround. #957 @susinmotion

Notable bugs fixed/issues closed:

Paired-end reads from Casava 1.8 no longer require renaming for use in normalize-by-median.py and abund-filter.py when used in paired mode #818 @ctb

Python version support clarified. We do not (yet) support Python 3.x #741 @mr-c

If a single output file mode is chosen for normalize-by-median.py we now default to overwriting the output. Appending the output is available by using the append redirection operator from the shell. #843 @drtamermansour

Scripts that consume sequence data using C++ will now properly throw an error on truncated files. #897 @kdmurray91
And while writing to disk we properly check for errors #856 #962 @mr-c

abundance-dist-single.py no longer fails with small files and many threads. #900 @mr-c

Additional fixes/features

Of interest to users:

Many documentation updates #753 @PamelaM, #782 @bocajnotnef, #845 @alameldin, #804 @ctb, #870 @SchwarzEM, #953 #942 @safay, #929,@davelin1, #687 #912 #926 @mr-c

Installation instructions for Conda, Arch Linux, and Mac Ports have been added #723 @reedacartwright #952 @elmbeech #930 @ahaerpfer

The example script for the STAMPS database has been fixed to run correctly #781 @drtamermansour

split-paired-reads.py: added -o option to allow specification of an output directory #752 @bede

Fixed a string formatting and a boundry error in sample-reads-randomly.py #773 @qingpeng #995 @ctb

CSV output added to abundance-dist.py, abundance-dist-single.py, and count-overlap.py, and readstats.py #831 #854 #855 @drtamermansour #959 @anotherthomas

TSV/JSON output of load-into-counting.py enhanced with the total number of reads processed #996 @kdmurray91
Output files are now also checked to be writable before loading the input files #672 @pgarland @bocajnotnef

interleave-reads.py now prints the output filename nicely #827 @kdmurray91

Cleaned up error for input file not existing #772 @jessicamizzi #851 @ctb

Fixed error in find-knots.py #860 @TheOneHyer

The help text for load-into-counting.py for the --no-bigcounts/-b flag has been clarified #857 @kdmurray91

@lexnederbragt confirmed an old bug has been fixed with his test for whitespace in sequence identifiers interacting with the extract-partitions.py script #979

Now safe to copy-and-paste from the user documentation as the smart quotes have been turned off. #967 @ahaerpfer

The script make-coverage.py has been restored to the sandbox. #920 @SherineAwad

normalize-by-median.py will warn if two of the input files have the same name #932 @elmbeech

Of interest to developers:

Switched away from using --user install for developers #740 @mr-c @drtamermansour & #883 @standage

Developers can now see a summary of important Makefile targets via make help #783 @standage

The unused khmer.load_pe module has been removed #828 @kdmurray91

Versioneer bug due to new screed release was squashed #835 @mr-c

A Python 2.6 and 2.7.2 specific bug was worked around #869 @kdmurray91 @ctb

added functions hash_find_all_tags_list and hash_get_tags_and_positions to CountingHash objects #749 #765 @ctb

The make diff-cover and ChangeLog formatting requirements have been added to checklist #766 @mr-c

A useful message is now presented if large tables fail to allocate enough memory #704 @mr-c

A checklist for developers adding new CPython types was added #727 @mr-c

The sandbox graduation checklist has been updated to include streaming support #951 @sguermond

Specific policies for sandbox/ and scripts/ content, and a process for adding new command line scripts into scripts/ have been added to the developer documentation #799 @ctb

Sandbox scripts update: corrected #! Python invocation #815 @Echelon9, executable bits, copyright headers, no underscores in filenames #823 #826 #850 @alameldin several scripts deleted, docs + requirements updated #852 @ctb

Avoid running big-memory tests on OS X #819 @ctb

Unused callback code was removed #698 @mr-c

The CPython code was updated to use the new checklist and follow additional best practices #785 #842 @luizirber

Added a read-only view of the raw counting tables #671 @camillescott #869 @kdmurray91

Added a Python method for quickly getting the number of underlying tables in a counting or presence table #879 #880 @kdmurray91

The C++ library can now be built separately for the brave and curious developer #788 @kdmurray91

The ReadParser object now keeps track of the number of reads processed #877 @kdmurray91

Documentation is now reproducible #886 @mr-c

Python future proofing: specify floor division #863 @mr-c

Miscellaneous spelling fixes; thanks codespell! #867 @mr-c

Debian package list update #984 @mr-c

khmer.kfile.check_file_status() has been renamed to check_input_files() #941 @proteasome
filter-abund.py now uses it to check the input counting table #931 @safay

normalize-by-median.py was refactored to not pass the ArgParse object around #965 @susinmotion

Developer communication has been clarified #969 @sguermond

Tests using the 'fail_okay=true' parameter to runscript have been updated to confirm the correct error occurred. 3 faulty tests were fixed and the docs were clarified #968 #971 @susinmotion

FASTA test added for extract-long-sequences.py #901 @jessicamizzi

'added silly test for empty file warning' #557 @wltrimbl @bocajnotnef

A couple tests were made more resilient and some extra error checking added in CPython land #889 @mr-c

Copyright added to pull request checklist #940 @sguermond

khmer_exceptions are now based on std::strings which plugs a memory leak #938 @anotherthomas

Python docstrings were made PEP257 compliant #936 @ahaerpfer

Some C++ comments were converted to be Doxygen compliant #950 @josiahseaman

The counting and presence table warning logic was refactored and centralized #944 @susinmotion

The release checklist was updated to better run the post-install tests #911 @mr-c

The unused method find_all_tags_truncate_on_abundance was removed from the CPython API #924 @anotherthomas

OS X warnings quieted #887 @mr-c

Known issues:

All of these are pre-existing.

Some users have reported that normalize-by-median.py will utilize more memory than it was configured for. This is being investigated in #266

Some scripts only output FASTA even if given a FASTQ file. This issue is being tracked in #46

Contributors

@ctb, @kdmurray91, @mr-c, @drtamermansour, @luizirber, @standage, @bocajnotnef, *@susinmotion, @jessicamizzi, *@elmbeech, *@anotherthomas, *@sguermond, *@ahaerpfer, *@alameldin, *@TheOneHyer, *@aditi9783, *@proteasome, *@bede, *@davelin1, @Echelon9,
*@reedacartwright, @qingpeng, *@SchwarzEM, *@scottsievert, @PamelaM, @SherineAwad, *@josiahseaman, *@lexnederbragt,

* Indicates new contributors

Issue reporters

@moorepants, @teshomem, @macmanes, @lexnederbragt, @r-gaia-cs, @Magentashades

Version 1.3

This is the v1.3 release of khmer featuring a new FAST[AQ] parser from the SeqAn project.

Docs at: https://khmer.readthedocs.org/en/v1.3/

New items of note:

Fixes the two multithreaded reading of sequence files issues: FASTQ parsing and the recently found read dropping issue. Several khmer scripts now support reading from non-seekable plain and gziped FAST[AQ] files (a.k.a pipe or streaming support). @mr-c #642

Notable bugs fixed/issues closed:

restore threading to load-graph.py #699 @mr-c

Additional fixes/features

increase filter_abund.py coverage #568 @wrightmhw
Provide scripts/ testing coverage for check_space_for_hashtable #386 #678 #718 @b-wyss
Use absolute URI in CODE_OF_CONDUCT #684 @jsspencer
give SeqAn credit #712 @mr-c
Added testing to make sure all sandbox scripts are import-able and execfile-able. #709 @ctb
reduce memory requirements to run tests #701 @ctb
Two minor bug fixes to sandbox scripts #706 @ctb
Upgrade of trim-low-abund for better, more profitable streaming. #601 @ctb
Add --force or --expert or --ignore flag to all khmer scripts that do sanity checking #399 #647 @jessicamizzi
Add XDECREF for returned read tuple in ReadParser.read_pair_iterator() #693 @mr-c @camillescott

Known issues:

All of these are pre-existing.

Some users have reported that normalize-by-median.py will utilize more memory than it was configured for. This is being investigated in #266

If your k-mer table is truncated on write, an error may not be reported; this is being tracked in #443. However, khmer will now (correctly) fail when trying to read a truncated file (See #333).

Paired-end reads from Casava 1.8 currently require renaming for use in normalize-by-median and abund-filter when used in paired mode. The integration of a fix for this is being tracked in #23

Some scripts only output FASTA even if given a FASTQ file. This issue is being tracked in #46

A user reported that abundance-dist-single.py fails with small files and many threads. This issue is being tracked in #75

Contributors

@mr-c, @ctb, @camillescott, @b-wyss, @wrightmhw, @jsspencer

Version 1.2

This is the v1.2 release of khmer: minor new features and bug fixes. The start of this release cycle coincided with the Mozilla Science Lab Global Sprint 2014. We honor and thank the 19 new contributors (including four Michigan State University undergraduates) who volunteered their time to contribute!

Docs at: https://khmer.readthedocs.org/en/v1.2/

New items of note:

@mr-c and @ctb are proud to announce khmer's code of conduct http://khmer.readthedocs.org/en/v1.2/dev/CODE_OF_CONDUCT.html #664
All scripts list which files have been created during their execution #477 @bocajnotnef
All scripts now only output status messages to STDERR instead of STDOUT #626 @b-wyss
docs/ a fairly major re-organization and brand new developer docs @ctb @mr-c
load-into-counting.py: --summary-info: machine readable summary in JSON or TSV format #649 @kdmurray91
scripts/extract-partitions.py: added documentation for .dist columns #516 @chuckpr
Makefile: a new target make install-dependencies is useful for developers #539 @mr-c
Sandbox scripts have been cleaned up, or removed (see the sandbox/README.rst for details) #589 @ctb

Notable bugs fixed/issues closed:

do-partition.py's excessive spawning of threads fixed. #637 @camillescott
Fixed unique k-mer count reporting in load-graph, load-into-counting, and normalize-by-median. #562 @mr-c
Clarified and test the requirement for a 64-bit operating system #529 @Echelon9
Removed some of the broken multi-threading options #511 @majoras-masque
Fix table.get("wrong_length_string") gives core dump #585 @Echelon9
filter-abund lists parameters that it doesn't use #524 @jstapleton
Reduction of memory required to run the test suite #542 @leogargu
BibTeX included in CITATIONS #541 @HLWiencko

Additional fixes/features

delete ScoringMatrix::assign as it is unused #502 @RodPic
Root all of our C++ exceptions to a common base exception #508 @iglpdc
deleted KhmerError #503 @drlabratory
normalize-by-median reporting output after main loop exits, in case it hadn't been triggered #586 @ctb
Many issues discovered by cppcheck cleaned up #506 @brtaylor92
Developers have a new Makefile target to autofix formatting: make format #612 @brtaylor92
normalize-by-median.py test coverage increased #361 @SherineAwad
Several unused functions were removed #599 @brtaylor92
Developer docs now link to the stdc++ docs as appropriate #629 @mr-c
Added tests for non-sequential access to input files #644 @bocajnotnef
Removed khmer/theading_args.py #653 @bocajnotnef
Improved test for maximum k value #658 @pgarland
ReadParser no longer crashes if n_threads = 0 #86 @jiarong

Known issues:

All of these are pre-existing.

Multithreaded reading will drop reads. This major issue has been present for several khmer releases and was only found via a much larger test case that we had been previously using. Credit to @camillescott. Workaround: disable threading. The next release will fix this and the other FAST[AQ] parsing issues. #681

Some users have reported that normalize-by-median.py will utilize more memory than it was configured for. This is being investigated in #266

Some FASTQ files confuse our parser when running with more than one thread. For example, while using load-into-counting.py. If you experience this then add "--threads=1" to your command line. This issue is being tracked in #249

If your k-mer table is truncated on write, an error may not be reported; this is being tracked in #443. However, khmer will now (correctly) fail when trying to read a truncated file (See #333).

Paired-end reads from Casava 1.8 currently require renaming for use in normalize-by-median and abund-filter when used in paired mode. The integration of a fix for this is being tracked in #23

Some scripts only output FASTA even if given a FASTQ file. This issue is being tracked in #46

A user reported that abundance-dist-single.py fails with small files and many threads. This issue is being tracked in #75

Contributors

@mr-c, @ctb, *@bocajnotnef, *@Echelon9, *@jlippi, *@kdmurray91, @qingpeng, *@leogargu, *@jiarong, *@brtaylor92, *@iglpdc, @camillescott, *@HLWiencko, *@cowguru2000, *@drlabratory, *@jstapleton, *@b-wyss, *@jgluck, @fishjord, *@SherineAwad, *@pgarland, *@majoras-masque, @chuckpr, *@RodPic, @luizirber, *@jrherr

* Denotes new contributor

version 1.1

This is v1.1, a minor version release; this version adds several new scripts.

Docs at: https://khmer.readthedocs.org/en/v1.1/

Release notes w/links: https://github.com/ged-lab/khmer/releases/tag/v1.1

New items of note:

• removed unnecessary files from PyPI package; distribution is now under 2 MB (#419) @mr-c
• tests are now distributed with package and can be run after 'pip install' (#451) @mr-c
• complain properly on file read failures (#333) @ctb
• Sequence loading scripts will now report total numbers of k-mers if given --report_total_kmers (#491/#429) @mr-c
• added metagenome protocol to acceptance testing (#472) @SherineAwad @ctb

Notable bugs fixed/issues closed:

• removed sandbox/load-into-hashbits.py (superseded by scripts/load-graph.py --no-tagset) (#480, @wrightmhw)
• promoted extract-long-sequences.py to scripts (#461, @wrightmhw)
• promoted fastq-to-fasta.py to scripts (#436, @wrightmhw)
• remove incorrect filesystem space check from abundance-dist.py (#452, @chuckpr)
• when counting hash writes fail, produce error message (#411, @znruss)
• removed a number of memory leaks found by Coverity and valgrind (#451, @mr-c)
• updated reservoir sampling to produce multiple subsamples with -S (#197, @ctb)
• fixed pip2, python2 issues (#428 and #485, @accaldwell @mr-c)
• removed untested/unused code and scripts (#438, @mr-c)

Known issues:

All of these are pre-existing.

Some users have reported that normalize-by-median.py will utilize more
memory than it was configured for. This is being investigated in
#266

Some FASTQ files confuse our parser when running with more than one thread.
For example, while using load-into-counting.py. If you experience this then
add "--threads=1" to your command line. This issue is being tracked in
#249

If your k-mer table is truncated on write, an error may not be reported; this
is being tracked in #443.
However, khmer will now (correctly) fail when trying to read a truncated file
(See #333).

Paired-end reads from Casava 1.8 currently require renaming for use in
normalize-by-median and abund-filter when used in paired mode. The
integration of a fix for this is being tracked in #23

Some scripts only output FASTA even if given a FASTQ file. This issue
is being tracked in #46

A user reported that abundance-dist-single.py fails with small files and many
threads. This issue is being tracked in #75

Contributors

@mr-c, @ctb, @camillescott, @wrightmhw, @chuckpr, @luizirber, @accaldwell,
@znruss

version 1.0.1

This is bugfix release. Note: the installation instructions have been slightly simplified.

https://khmer.readthedocs.org/en/v1.0.1/

New items of note:

This release successfully installs and passes its unit tests on Debian 6.0 "Squeeze", Debian 7.0 "Wheezy", Fedora 19, OS X 7 "Lion", OS X 8 "Mountain Lion", Red Hat Enterprise Linux 6, Scientific Linux 6, Ubuntu 10.04 LTS, and Ubuntu 12.04 LTS. Thanks to the UW-Madison Build and Test Lab for their testing infrastructure.

Notable bugs fixed/issues closed:

fixed thread hanging issue #406 @ctb
Explicit python2 invocation #404 @mr-c
MANIFEST.in,setup.py: fix to correct zlib packaging #365 @mr-c
fixed check_space_for_hashtable to use args.n_tables #382 @ctb
Bug fix: make-initial-stoptags.py error on missing .ht input file, actual input file is .pt #391 @mr-c

Minor updates

include calc-best-assembly.py in v1.0.1 #409 @ctb
updated normalize-by-median documentation for loadtable #378 @ctb
updated diginorm for new FP rate info; corrected spelling error #398 @ctb
Add spellcheck to code review checklist. #397 @ctb

Known Issues

All of these are pre-existing.

Some users have reported that normalize-by-median.py will utilize more
memory than it was configured for. This is being investigated in
#266

Some FASTQ files confuse our parser when running with more than one thread.
For example, while using load-into-counting.py. If you experience this then
add "--threads=1" to your command line. This issue is being tracked in
#249

If your k-mer table (hashfile) gets truncated, perhaps from a full filesystem, then our
tools currently will get stuck. This is being tracked in #247 and #246

Paired-end reads from Casava 1.8 currently require renaming for use in
normalize-by-median and abund-filter when used in paired mode. The
integration of a fix for this is being tracked in #23

annotate-partitions.py only outputs FASTA even if given a FASTQ file. This
issue is being tracked in #46

A user reported that abundance-dist-single.py fails with small files and many
threads. This issue is being tracked in #75

Contributors

@mr-c, @ctb, @luizirber, @RamRS, @ctSkennerton

version 1.0

582 changed files with 40,527 additions and 31,772 deletions.

The team has been hard at work since v0.8 to refine the codebase into a stable product.

https://khmer.readthedocs.org/en/latest/

With the 1.0 release we are making a commitment to using Semantic Versioning[0]: the version number will reflect the impact of the changes between releases. New major versions will likely require you to change how you use the project. Minor versions indicate new functionality that doesn't impact the existing. Patch versions indicate backwards-compatible fixes. Right now we are limiting this promise to the command-line interface. A future release will introduce a stable and mature Python API to the khmer project and at that time we will extend the version system to include that API.

New items of note:

CITATION: Each script now outputs information on how to cite it. There is a new paper to describes the project overall: MR Crusoe et al., 2014. doi: 10.6084/m9.figshare.979190

The documentation for the scripts has undergone an overhaul. The scripts now output extensive notes and the formal documentation website is generated from the scripts themselves and will never be out of sync.

https://khmer.readthedocs.org/en/latest/scripts.html

Notable bugs fixed/issues closed:

git clone of the khmer repo reqs > 0.5 GiB #223 @mr-c
new khmer/file module #357 @RamRS
Floating point exception in count-overlap.py #282 @qingpeng
add documentation for sample-reads-randomly #192 @mr-c
only build zlib and bzip2 when needed #168 @mr-c

Minor updates

khmer tools should output intelligent error messages when fed empty files #135 @RamRS
set IParser::ParserState::ParserState:fill_id to zero at initialization #356 @mr-c
demote nose & sphinx to extra dependencies. #351 @mr-c
CID 1054792 (Medium) Uninitialized scalar field (UNINIT_CTOR) #179 @mr-c
CID 1077117 (Medium): Division or modulo by zero (DIVIDE_BY_ZERO) #182 @mr-c
if --savehash is specified then don't continue if there is not enough free disk space #245 @RamRS
finish fixing implicit downcasts #330 @mr-c
Clean up compile warnings in subset.cc #172 @mr-c
all scripts need to output their version #236 @mr-c
environmental variables need documenting #303 @mr-c
C++ code should be consistently formatted #261 @mr-c
Clean up ancillary files #146 @mr-c
squash option not implemented in abundance-dist-single.py #271 @RamRS
Add documentation on how to tie into a particular tagged version #29 @mr-c
pip install -e fails with compile error #352 @mr-c
remove the unused KTable object #337 @luizirber
zlib 1.2.3 -> zlib 1.2.8 #336 @mr-c
CID 1173035: Uninitialized scalar field (UNINIT_CTOR) #311 @mr-c
CID 1153101: Resource leak in object (CTOR_DTOR_LEAK) #309 @mr-c
remove khmer::read_parsers::IParser::ParserState::thread_id #323 @mr-c
several modifications about count-overlap.py script #324 @qingpeng
fixed runscript to handle SystemExit #332 @ctb
CID 1063852: Uninitialized scalar field (UNINIT_CTOR) #313 @mr-c
[infrastructure] update to new Doxyfile format, make version number autoupdate #315 @mr-c
Removed an extraneous using namespace khmer; in kmer.hh, #276 @fishjord
Minimum and recommended python version #94 @mr-c
KmerCount class appears to be unused #302 @mr-c
If loadhash is specified in e.g. normalize-by-median, don't complain about default hashsize parameters #117 @RamRS

Known Issues

All of these are pre-existing.

Some users have reported that normalize-by-median.py will utilize more
memory than it was configured for. This is being investigated in
#266

Some FASTQ files confuse our parser when running with more than one thread.
For example, while using load-into-counting.py. If you experience this then
add "--threads=1" to your command line. This issue is being tracked in
#249

If your k-mer table (hashfile) gets truncated, perhaps from a full filesystem, then our
tools currently will get stuck. This is being tracked in #247 and #96 and #246

Paired-end reads from Casava 1.8 currently require renaming for use in
normalize-by-median and abund-filter when used in paired mode. The
integration of a fix for this is being tracked in #23

annotate-partitions.py only outputs FASTA even if given a FASTQ file. This
issue is being tracked in #46

A user reported that abundance-dist-single.py fails with small files and many
threads. This issue is being tracked in #75

Contributors

@camillescott, @mr-c, @ctb, @luizirber, @RamRS, @qingpeng

[0] http://semver.org/