- updated documentation
- adding command line option --precision to round the precision of decimals defining smaller or larger GC buckets
- bugfix when number negatives != 0 but positives > negatives
- bugfix in stdn out (is now tab separated)
- bugfix unordered chromosomes
- output data is again properly ordered by chrom and chromStart
- added seed option to CLI (Issue #11)
- seed is used for reproducing sampling results when debugging
- complete support for bed files (Issue #5)
- added option to specify number of label columns
- use a genome file of the fasta for the contigs by
--genome-file(see #7) to avoid hard-coded contigs. Now this sampler should work with every species. - adding a logger using logging. with --log the log file can be written.
- reworked CLI (Issue #6)
- input is now connected to the -i flag
- refrence is now connected to the -r flag
- added --verbose flag
- added -c (--bgzip) flag
- added bgzip support for -o
- output file is bgzipped if -c flag is set
- output file is bgzipped if filename ends with
.gz, even if -c is not set
- changed how output works
- positive and negative labeled entries are now saved in one file
- without specifying an output file (-o), the results are written to stdout
- minor bugfixes
- debug output is now written to stderr
- rename genome_file to reference-file
Initial negative_training_sampler version.