Gillespy2

CHANGES
- Auto solver selection process is now a callable function, get_best_solver() which returns a solver based on model configuration and user environment.
- Extended get_best_solver() to consider solver based on supported features and species mode
- added callable function to solvers get_solver_settings(), returns a tuple of relevant settings to model.run() for the solver
- Added badges for conda-forge and download statistics

FIXES
- Fixed issue where seeds were sometimes overlapped for VariableSSACSolver
- Fixed issue where timeout could cause an error for multiple trajectories

CHANGES
- added initial value setter method for Species (myspecies.set_initial_value(val))
- BasicTauHybridSolver is now automatically selected with model.run() for models containing:
- Rate Rules, Assignment Rules, Events, Function Definitions
- axes scaling is now available with matplotlib and plotly plotting

FIXES
- Fixed issue where Tau selection occasionally caused an error
- Assignment rules now work correctly, again
- Rate Rules now work correctly, again
- Default matplotlib y scale now adjusts to simulation
- Events are now comparable
- Models containing Events are now comparable
- coverage badge in README is no longer static, now updates automatically

GillesPy2 v.1.4.0 CHANGELOG

FEATURES

- New Solver : VariableSSACSolver
This solver works identically to SSACSolver, but is
optimized for multiple trajectories using variable input values
for species and Parameters.
- Results and ensembles of Results now share an object type.
- Ensembles can now be created from separate runs using the '+' operator
In order to add Results to an Ensemble, they must share the same model/solver
- Model objects are now comparable
- Added advanced string parsing for user inputs
- Now uses AST parsing, interpretation should now be consistent across solvers
- Results can now be properly pickled and unpickled
- Added getters for all model objects
- Added delete methods for all model objects

FIXES

- BasicODESolver now correctly accepts integrator args.
- Resolved incorrect rounding in BasicTauHybridSolver switching
- All model objects are now checked for unique names when added to model

- removed further uses of isinstance with module objects to avoid potential errors
- Fix for pickling Results objects
- Fixed error where math definitions could overlap with Parameter/Species names in hybrid solver. Parameter/Species names will now take priority in evaluation.
- Added error checking layer for threads
- Re-Ran notebooks and fixed paths for Michaelis Menten notebooks
- added License to Manifest for conda-forge

-This version contains a hotfix for an issue where using module class objects were not passing comparison checks. This was an issue with Python's isinstance method.

- Fix for multiple unnamed Reactions
- Timeouts now work with Windows! Refactored using python threading instead of signals.
- Updated hybrid solver value testing and error messages for float/int initial values