Ideas
Opened this issue · 0 comments
GuettnerBianca commented
Idea collection for future tasks. Items from the task list can be used to create an issue.
For all classes
Style
- #22
- make the interaction parameters a dictionary
- improve encapsulation
- make some properties inaccesible to user
- extra plotting module (LGCA -> data only)
- add an observables module (e.g. calculates entropy)
- #21
Features
- #23
- #24
- #25
- calculate entropy for all classes
- #20
- Support custom interactions more formally with something like set_custom_interaction(fun:callable), set_custom_parameter(name:str, value:object)
- Support stopping the simulation, optionally doing something to the lattice and then resuming the simulation in the same LGCA
- extend it to 3D (can this be done efficiently enough?)
- store and load LGCA objects
Plots
Improvement
- use hypothesis to test (creates different inputs for test cases → robust code)
- write testing functions for all interactions
- #29
- improve style of colourbar striding (Locator, Formatter)
- #12
- #13
- #14
Classical LGCA
- #15
- feature: add channel flexibility within a node for go-or-grow models
- feature: add the option for more lattices for other species
- feature: add different velocities (m) and update constraints on interaction radius r_int
IBLGCA
LGCA without volume exclusion
- #31
- improvement: 1D flux plot with colour gradient instead of fixed colours
- #17
- improvement: correct entropy calculation with flux vector (sum of direction/sum of particles total) and state vector (sum of all incl. rest/sum of particles total) - compute entropy of multinomial distribution for each node and take mean, empty nodes have entropy 0
- improvement: handle rest channels in provided nodes – sum particles in surplus rest channels automatically
- #32