0.4.0

API changes

• MultiRangeQuery -> MomicsQuery.
• manifest method to output the whole configuration of a Momics instance.
• consolidate method to consolidate a Momics repository.
• No submodules are imported in __init__.py anymore.
• All export functions are now Momics methods.
• All ingestion methods for Momics now start with ingest_*.

New features

• aggregate submodule to merge dictionaries of partial coverage tracks (over pyranges) into genome-wide tracks.
• Relatively basic ChromNN CNN TensorFlow model.
• MomicsDataset class to pass data to TensorFlow models.
• MomicsStreamer class to stream data from Momics repositories.
• Momics can ingest pyranges objects to store genomic features.
• Sequences, features and tracks can be extracted from a Momics repository.

Enhancements

• CLI is now partially based on cloup to improve user experience.
• seq(label = "...) now returns the sequence for an entire chromosome.
• tracks(label = "...) now returns the genome-wide track.
• MomicsQuery queries now rely on pyranges for range queries.
• MomicsQuery queries can extract only a subset of the tracks.
• add_track and remove methods for Momics class.
• to_npz and to_json methods for MomicsQuery class.

Maintenance

• Support jupyter notebooks in documentation.
• Add changelog.
• Improve docs.
• Logging system updates.
• CLI updates.
• Add codecov support.

Bug fixes

• Ensure that queries are done per chromosome, even if provided ranges are stranded.
• All ranges are now 0-based half-open intervals, as in pyranges and BED files.
• Parallelization for queries only relies on TileDB internal system.
• Filters used in Momics tables.
• Removal of Azure-hosted repositories.