1.2.0

A few non-trivial updates:

• (H)CTL model checker has been updated to version 0.3.0 and now supports quantifiers with domain (i.e. V{x} in %domain%).
• PerturbedAsynchronousGraph cannot be created from a network with implicit parameters. This simplifies a lot of correctness reasoning, because the color sets of the internal graphs and the original network are always compatible.
• CI is now testing that example notebooks work.
• Several methods where it makes sense (mostly various mk_* methods) can take a Model object instead of raw values (e.g. VertexModel instead of dict[VariableId, bool]).

Other misc stuff:

• Added Bdd.validate implemented in lib-bdd version 0.5.22.
• Added VertexSet.enclosed_subspace and VertexSet.enclosed_named_subspace.

1.1.2

Mostly provides additional utility methods on top of 1.1.1.

• BooleanNetwork.assign_parameter_name and BooleanNetwork.name_implicit_parameters for "standardized" conversion of implicit parameters to explicit ones.
• BooleanNetwork.is_variable_input, BooleanNetwork.is_variable_constant, as well as BooleanNetwork.inputs (or BooleanNetwork.input_names) and BooleanNetwork.constants (or BooleanNetwork.constant_names) allow to test/retrieve constants and inputs instead of just inlining them, as we supported before.
• In ColorModel.instantiate, it is now possible to automatically infer the regulatory graph that matches the new functions instead of using the original one.
• VertexSet.enclosing_subspace and VertexSet.enclosing_named_subspace allow retrieving the smallest subspace that still contains all vertices in the set.
• SpaceSet.with_all_sub_spaces and SpaceSet.with_all_super_spaces (the same is available for ColoredSpaceSet) allows simpler extension of a set with sub-spaces or super-spaces (this complements SymbolicSpaceContext.mk_sub_spaces and SymbolicSpaceContext.mk_super_spaces).

Full Changelog: v1.1.1...v1.1.2

1.1.1

No functionality changes compared to 1.1.0, but it removes a forgotten print that sometimes appears in tool output.

v1.1.0

Introduces a new phenotype classification method: Now, you can pick between classify_phenotypes and classify_attractor_phenotypes, where the second option actually classifies each attractor in isolation, so you get a better idea of the actual possible behaviors (the first option only detects if a phenotype is possible at all).

v1.0.1

Compared to 1.0.0, updates the lib-bdd dependency to version 0.5.21. This does not add any new significant functionality, but it fixes a performance problem in the implementation of mk_dnf/mk_cnf. It also uses a new, improved to_dnf algorithm. Finally, to_dnf now has a size_limit argument which interrupts the method if the DNF becomes too large.

What's Changed

• Update lib-bdd to 0.5.21 by @daemontus in #24

Full Changelog: v1.0.0...v1.0.1

v1.0.0

This is essentially a complete rewrite of the Python wrappers. Much has been added. Please consult the API documentation regarding the new interface: https://biodivine.fi.muni.cz/docs/aeon-py/latest/biodivine_aeon.html

0.0.9-alpha2

0.0.9a2

v0.4.0a5

0.4.0-alphaa5

0.4.0a5

v0.4.0a4

0.4.0-alpha4

0.4.0a4

Phenotype control alpha release 3

0.4.0-alpha3

versions