pyani

Minor update to increment version number in the code.

This release fixes a deprecated matplotlib call, and deprecated pandas calls.

This release fixes Issue #177: input filenames that contain an open or a closed parenthesis produce the message: 'WARNING: At least one BLAST run failed. ANIb may fail.'

This release fixes issues due to pandas API changes.

• exceptions used in pyani are now found in pandas.errors not pandas.io.common
• changes to the testing API (will not affect most users)

This release fixes the bug reported in issue #178 that affected graphical output if all input filenames could be interpreted as floating point numbers.

This release fixes issue #132 (TETRA analysis failing when one or more 4-mers are missing from input sequences).

Thanks to @ghbio for the bug report!

This release implements several changes:

• fixed warnings about deprecated usage of pandas and other syntax
• ANIb now respects the --workers argument (PR #127 from Özcan Esen)
• distance matrices are converted from square to condensed form for linkage analysis
• matplotlib calls use new API ('false', 'off' -> False; 'true', 'on' -> True)
• deprecate legacy BLAST testing: ANIblastall commands are no longer supported
• update concordance tests (issue #105)
• extend test suites (issue #104)
• modify ANIm concordance test to accommodate new command structure
• add delta-filter wrapper for compatibility with SGE/OGE schedulers

This release fixes issue #97 in which numeric arguments to the GenBank download script were not recognised.

This release fixes a PyPI installation problem, resulting from a badly-built .whl wheel.

• ANIm now uses delta-filter to remove alignments of repeat regions (issue #91)
• added --filter_exe option to specify location of delta-filter utility (issue #91)
• fixed --format option so that GenBank downloads work again (issue #89)
• add --SGEargs option to average_nucleotide_identity.py for custom qsub settings
• README.md badges now clickable
• --version switch added to average_nucleotide_identity.py
• FTP timeouts are now caught differently in genbank_get_genomes_by_taxon.py
• Additional characters in NCBI FTP URIs now escaped in genbank_get_genomes_by_taxon.py - should be fewer failed downloads
• Modified error messaging when NUCmer alignment fails
• average_nucleotide_identity.py argument documentation improvements
• Script now fails immediately if label or class files missing (issue #78)
• Changes to --noclobber log behaviour (issue #79)
• fixed --rerender code (issue #85)

The previous release (0.2.2) on PyPI had script shebangs that pointed to a local development Python, in the wheel and egg distribution files. It is not possible to replace those files on PyPI, so a new release was made to fix this issue, and the old packages withdrawn.

• fix for issue #53 (--maxmatch has no effect)
• fix to genbank_get_genomes_by_taxon.py to account for NCBI FTP location changes
• fixed issue #52 (local variable bug)
• fixed issued #49 (TETRA failure) and #51 (matplotlib bug)
• add several tests and support for codecov.io, landscape.io and Travis-CI
• removed requirement for rpy2
• moved scripts to bin/ subdirectory

v0.2.1 introduces minor changes with respect to v0.2.0:

• genbank_get_genomes_by_taxon.py uses the new NCBI FTP locations, and will attempt to identify and inform the user about failed downloads.
• seaborn output can now handle larger (>500 genomes) datasets
• better handling of missing class information
• adds a Hadamard product output
• you can now rerender graphics without recalculation of alignments
• alignment output is zipped into an archive by default

There are also several bugfixes (see CHANGES.md for more details)

Version 0.2.0 introduces a number of changes with respect to the v0.1 series of releases. Notably:

• v0.2.0 is for use with Python 3
• a longstanding bug in the --fragsize option is fixed
• several pull requests for enhancements are merged
• BLAST/nucmer results are now written to a subdirectory of the output folder, and are compressed to save disk space
• a script to download publicly-available genomes from NCBI is provided

This bump fixes some bugs (issues #5 and #11), and updates installation instructions reflecting the upload of this package to PyPI.

Release made solely to obtain a Zenodo DOI (10.5281/zenodo.31812)