v0.1.12

• Implemented data utils: map_concurrently, iter_transpose, iter_precompute, iter_slices, iter_lines
• Added draft implementation of Molecule and Atom classes

v0.1.11

1. Improved picker performance by utilizing numpy object arrays to quickly update labels.
2. Add support for custom similarity and potential functions for the picker.
3. One more step for meaningful package public API.
4. Switched docs from Quarto to MkDocs with init support of automatic generation.
5. Improved docstrings, names, examples, etc.

v0.1.10

Test Release:

Added extra cpu dependencies for JAX
Added RDKit as an auto-installable dependency

0.1.9

Test Release:

• Added extra cpu dependencies for JAX
• Added RDKit as an auto-installable dependency

v0.1.8

• Simplified the OnlineDiversityPicker for enhanced CUDA performance with JIT compilation.
• Added the ._init() method to initialize the correct size and type for internal data containers.
• Introduced .fast_init() for rapid initialization of internal data, speeding up computations in certain scenarios.

v0.1.7

1. Reimplement .extend() method with better JIT support
2. Reimplement .append() method using .extend() as a backend
3. Add unit tests
4. Bug fixes

v0.1.6

Added power parameter for OnlineDiversityPicker

v0.1.5

OnlineDiversityPicker: Add labels support Improve handling of boundary points

v0.1.4

Changed how OnlineDiversityPicker handles boundary points

v0.1.2

Add OnlineDiversityPicker