Brain states Simulation - Githubissues

robbisg / sekupy

Detergent for Multivariate Analysis Pipelines of Neuroimaging data in Python

https://sekupy.readthedocs.io/

Other

1 stars 0 forks source link

Brain states Simulation #57

Open robbisg opened 4 years ago

robbisg commented 4 years ago

Brain state simulation is a bit tricky when we have to face with the connectivity simulations.

[ ] Make a general Simulation class also for Kuramoto models
[ ] Make specific subclasses for each simulation
[ ] Make data in a pymvpa dataset
Solve the problem of model for connectivity
- Who is responsible for the generation of matrices
- Who is responsible for the generation of data
- How data can used for output

robbisg commented 4 years ago

StateSimulator vs DataModel

I use a StateSimulator to generate random system dynamics and a AutoregressiveModel to generate data that mimics system dynamics.

Now the problem is that:

I can use the Model inside the Simulator to avoid complicated parameters passing
I can use the Simulator inside the Model as a complex object with all parameters.

Which is the best solution? I use the Simulator inside the Model so that the model can use all information about dynamics to simulate.

TODO:

Change `Simulator` variable names:

[ ] _bs_matrices in states
[ ] _bs_length in state_length
[ ] _bs_sequence in state_sequence
[ ] _bs_dynamics in dynamics

Check

[ ] Check if the Simulator has been trained