This roadmap is a rough plan of upcoming work. Contributions are welcome at anytime, no matter where something is on this list.
- Reduce memory usage of loading and saving by making them zero in-memory copy.
- Document all structs and methods that will be stable and supported indefinitely.
- Remove legacy cruft.
- Return genotypes as 32bit ints.
- Other clean up including sorting bugs and small features. (see milestone issue list)
- Python release to accompany C API 1.0.
- Random access to site genotypes
- Python release to complete features needed for the tskit paper.
- Fast newick parsing
- Improve user experience by adding convenience methods for updating, retrieving metadata and schemas
- Codec performance
- Standardised schemas
- Added stats and options
- Improve errors
- (Fully deal with missing data)[tskit-dev#287]
- Time windowed statistics
- Build indexes to allow performant random tree access